EM 1110-2-181031 Jan 95

Chapter 5Coastal Geological Investigations 1

5-1. Introduction

a. Three principal time scales are important inassessing the geologic and geomorphic2 changes of coasts.These include: (1) modern studies, which are basedlargely on field data or laboratory and office experimentsof environmental processes; (2) historic studies, which arebased largely on information from maps, photography,archives, and other sources; and (3) studies of paleo-environments, which are based largely on stratigraphy andassociated geological principles (Figure 5-1). These gen-eral categories overlap. Furthermore, within each of thecategories, certain time scales may be of particular impor-tance for influencing coastal changes. For example, tidaland seasonal changes are significant in modern studies,and Holocene sea level history is important in paleo-environmental studies. Tidal fluctuations are difficult todetect in studies of paleoenvironmental changes, and sealevel typically changes too slowly to be an importantfactor in modern process studies.

b. Several lines of inquiry are available to assess thegeologic and geomorphic history of coasts. One means ofacquiring coastal data is through field data collection andobservation. These data may be numerical or non-numerical, and may be analyzed in the field, laboratory,or office. Laboratory studies are used to collect datathrough physical model experiments, such as in wavetanks, or to analyze geological properties of field data,such as grain size or mineralogy. Office studies includeinterpretation of historic maps, photographs, and refer-ences as well as analyses and numerical simulation offield, laboratory, and office data. Typically, the bestoverall understanding of environmental processes and thegeologic history of coasts is acquired through a broad-based combination of techniques and lines of inquiry.

c. Quality of results depends on several factors,including the use of existing data. If secondary datasources (i.e. existing maps, photography, and literaturesources) are limited or unavailable, assessing the geologic

____________________________1 Chapter 5 is an adaptation of Morang, Mossa, andLarson (1993), with new material added.2 Geomorphic refers to the description and evolution ofthe earth’s topographic features - surficial landformsshaped by winds, waves, ice, flowing water, and chemicalprocesses.

history will be more difficult, more costly, and typicallymore inaccurate. Consequently, before initiating detailedfield, laboratory, or office studies, thorough literaturereview and search for secondary data sources should beconducted. Appendices E and F list sources and agenciesthat can be consulted in searches for secondary data.

d. Quality of research equipment, techniques, andfacilities also influences the quality of the evaluation ofgeologic and geomorphic history. For example, echo-sounding and navigation instruments used to conductbathymetric surveys have recently been improved. Usingthese tools, the mapping of geologic and geomorphicfeatures can be extended further seaward to a higherdegree of accuracy than was previously possible. It isimportant that coastal researchers stay abreast of newtechniques and methods, such as remote sensing andgeophysical surveys, computer software and hardwaredevelopments, and new laboratory methods. For example,recent developments in Geographical Information Systems(GIS) enable the coastal scientist to analyze and interprethighly complex spatial data sets. This report describessome recent developments and techniques that are used inthe analysis of coastal data sets.

e. Scientists must recognize certain problems andassumptions involved in data collection and analyses andmake adjustments for them before attempting an interpre-tation. It is critical to account for various sources of errorin preparing estimates of coastal changes and acknowl-edge the limitations of interpretations and conclusionswhen these are based on data covering a short time periodor a small area.

f. Many of the techniques used to monitor processesand structures in the coastal zone are exceedingly com-plex. This chapter outlines some of the many errors thatcan occur when the inexperienced user deploys instru-ments or accepts, without critical appraisal, data fromsecondary sources. The text is not intended to be sopessimistic that it dissuades coastal researchers fromcontinuing their investigations, but rather is intended toguide them to other references or to specialists whereexpert advice can be obtained.

5-2. Sources of Existing Coastal Information

a. Literature sources.

(1) University and college departments and libraries.In many instances, books, periodicals, dissertations,theses, and faculty research project reports contain data.

5-1

EM 1110-2-181031 Jan 95

Figure 5-1. Techniques for studying geomorphic changes of coasts over various time scales. Arrows indicateapproximate time span during which a particular study technique can be used. X-axis is unitless; width of outlinesrepresent relative importance of general methods for studying coastal changes

This especially occurs when the institutions are in coastalareas, where research is funded by Federal or state gov-ernment agencies (i.e. Sea Grant), where the universityhas graduate programs and faculty active in research inappropriate fields. Major universities also have govern-ment document repositories where Federal and state gov-ernment publications are housed.

(2) Local sources. These can provide detailed andsometimes unique data pertinent to the locale. Such

sources include the local newspaper, courthouse records,historical diaries, lighthouse records, local journals, engi-neering contract records, land transactions, and museums.

(3) Government agencies. Geologic coastal datamay be available from government agencies at the Fed-eral, state, and local level (Appendices E and F). Federalagencies with data archives include the U.S. GeologicalSurvey (USGS), the U.S. Coast and Geodetic Survey(USCGS), the National Oceanographic and Atmospheric

5-2

EM 1110-2-181031 Jan 95

Agency (NOAA), the U.S. Army Corps of Engineers(USACE), (including the Waterways Experiment Station,and USACE District and Division offices), the Depart-ment of Transportation, the Environmental ProtectionAgency, the U.S. Fish and Wildlife Service, and theNaval Research Laboratory (NRL). A geographic list ofCERC coastal geologic and monitoring reports is providedin Appendix G. State agencies with relevant coastalinformation include state geological surveys (or bureausof geology), departments of transportation, departments ofenvironmental resources and/or water resources, and stateplanning departments. Some state health departmentsarchive well logs.

(4) Industry. Energy (oil and gas) companies oftenkeep records, which may be accessible to scientists, ofcoastal processes in conjunction with their offshore drill-ing operations. Construction companies have records infiles on their construction projects. Environmental andengineering firms may also have data from projects thatwere performed for government. Some of these data arein the public domain. Environmental impact reports fromnuclear power plants built in coastal areas contain exten-sive coastal process and geologic data.

(5) Journals and conference proceedings. Most largeuniversity libraries have holdings of national and interna-tional scientific journals. Most of the scientific literatureassociated with the geologic history of coasts is in therealm of geology, oceanography, marine science, physicalgeography, atmospheric science, earth science and polarstudies.

(6) Computerized literature searches. Most majoruniversity and government agency libraries have access tocomputerized literature databases. The databases containinformation that may be acquired by key terms, subjects,titles, and author names.

b. Meteorological and climatic data.

(1) Meteorological and climatic data are often usefulfor characterizing significant environmental processes andfor revealing the characteristics of severe storms. Majorstorms or long-term variations in storminess stronglyaffect coastal morphology (Carter 1988). This is mani-fested, for example, by the changes on barrier beachesassociated with winds, waves, and high water levels,which may cause overtopping and overwashing duringstorms.

(2) Meteorological and climatic data can be compiledfrom secondary sources or through an original data

collection program in the field using instruments andobservations. As with most of the important environmen-tal factors, most existing information pertains to studiesover historic and modern time scales. The NationalClimatic Data Center and the National Hurricane Centerwithin NOAA are important sources of meteorologicaland climatic data.

c. Wave data.

(1) Wave data are required to characterize theprocess-response framework of the coastal zone. Impor-tant wave parameters include wave height, period, steep-ness and direction, and breaker type. Of special interestis the character of waves inside the breaker zone, where itis estimated that 50 percent of sediment movement takesplace, mostly as bed load (Ingle 1966). Wave data canbe: (a) collected from existing sources; (b) estimated inthe office using hindcast techniques from weather maps,shipboard observations, and littoral environment observa-tions; or (c) measured in the field using instrumentedwave gauges.

(2) Wave gauge data are collected by Federal andstate agencies and by private companies. For researchprojects that require wave data, analyzed wave statisticsmay be available if instrumented buoys, offshore struc-tures, and piers are located near the study site. Publisheddata, which are geographically spotty, include statisticsfrom wave gauges, wave hindcasting, and visual observa-tions from shipboard or the littoral zone.

(3) Wave hindcasting is a technique widely used forestimating wave statistics by analysis of weather mapsusing techniques developed from theoretical consider-ations and empirical data. A coastal scientist can usepublished hindcast data or may choose to compute origi-nal estimates for a study area. Appendix D is a list of theUSACE Wave Information Studies reports, which coverthe Atlantic, Pacific, Gulf of Mexico, and Great Lakescoasts. Advantages of hindcasting include the long-termdatabase associated with weather maps and the compara-tively economic means of obtaining useful information.Disadvantages involve the transformation of waves intoshallow water, especially in areas of complex bathymetry.

(4) Visual wave observations from ships at sea andfrom shore stations along the coasts of the United Statesare also published in several references. Although obser-vations are less accurate than measured data, experiencedpersons can achieve reasonably accurate results and thegreat amount of observations available make it a valuableresource. Offshore, shipboard wave observations have

5-3

EM 1110-2-181031 Jan 95

been compiled by the U.S. Navy Oceanographic Researchand Development Activity, (now the Naval ResearchLaboratory (NRL)), in the form of sea and swell chartsand data summaries such as the Summary of ShipboardMeteorological Observations. While geographic coverageby these sources is extensive, the greatest amount ofobservations come from shipping lanes and other areasfrequented by ship traffic.

(5) At the shore, a program sponsored by HQUSACEfor data collection is the Littoral Environmental Observa-tion (LEO) program (Schneider 1981; Sherlock andSzuwalski 1987). The program, initiated in 1966, makesuse of volunteer observers who make daily reports onconditions at specific sites along the coasts of the UnitedStates. Data from over 200 observation sites are availablefrom CERC (Figure 5-2). As shown, LEO data not onlyinclude wave parameters, but also information on winds,currents, and some morphologic features. LEO is bestapplied to a specific site, and does not provide directinformation on deepwater statistics. The biggest disad-vantage is the subjective nature of the wave height esti-mates. LEO data should only be used as indicators oflong-term trends, not as a database of absolute values.

d. Sources of water level data.The NOS of theNOAA is responsible for monitoring sea level variationsat 115 station locations nationwide (Hicks 1972). CoastalUSACE District offices collect tidal elevation data atadditional locations. Daily readings are published inreports that are titled “Stages and Discharges of the (loca-tion of district office) District.” Predicted water levelsand tidal current information for each day can be obtainedfrom the annual “Tide Tables: High and Low WaterPredictions” and “Tidal Current Tables” published by theNOS. A convenient way to obtain daily tides is fromcommercial personal computer (PC) programs. Many ofthese programs are updated quarterly or yearly. Back-ground information concerning tidal datums and tidestations can be found in NOS publications titled “Index ofTide Stations: United States of America and Miscella-neous Other Stations,” and “National Ocean ServiceProducts and Services Handbook.”

e. Geologic and sediment data.

It is often important in studies of the geologic and geo-morphic history of coasts to evaluate existing geologicand sediment data. This type of information is dispersedamong numerous agencies and sources and includes avariety of materials such as geologic maps, soil surveys,highway borings, and process data such as the

concentrations and fluxes of suspended sediment fromnearby rivers. Published data are available from agenciessuch as the USGS, the U.S. Soil Conservation Service, theAmerican Geological Institute, and CERC. Differences ingeology and soil type may provide clues toward under-standing erosion and accretion patterns. Geologic andsedimentologic data are often useful for characterizingsignificant environmental processes and responses, such asthe effects of severe storms on coastlines.

f. Aerial photography.

(1) Historic and recent aerial photographs provideinvaluable data for the interpretation of geologic andgeomorphic history. The photographs can be obtainedfrom Federal and state government agencies such as theUSGS, the U.S. Department of Agriculture, the EROSData Center, and others listed in the Appendices E and F.Stereographic pairs with overlap of 60 percent are oftenavailable, allowing very detailed information to beobtained using photogrammetric techniques. Temporalcoverage for the United States is available from the1930’s to present for most locations. The types of analy-sis and interpretation that can be performed depend in parton the scale of the photographs, the resolution, and thepercentage of cloud cover. The effects of major eventscan be documented by aerial photography because thephotographic equipment and airplane can be rapidly mobi-lized. By such means, the capability exists for extensivecoverage in a short time and for surveillance of areas thatare not readily accessible from the ground.

(2) For modern process studies, a series of aerialphotographs provides significant data for examining avariety of problems. Information pertinent to environ-mental mapping and classification such as the nature ofcoastal landforms and materials, the presence of engineer-ing structures, the effects of recent storms, the locationsof rip currents, the character of wave shoaling, and thegrowth of spits and other coastal features can be exam-ined on aerial photographs. For the assessment of somemorphologic features, photogrammetric techniques may behelpful. It is generally preferable to obtain photographyacquired during low tide so that nearshore features areexposed or partly visible through the water.

(3) For studies over historical time scales, multipletime series of aerial photographs are required. Historicalphotography and maps are integral components of shore-line change assessments. Water level and, therefore,shoreline locations, show great variation according towhen aerial photographic missions were flown.

5-4

EM 1110-2-181031 Jan 95

Figure 5-2. Littoral Environmental Observation forms used by volunteer observers participating in the LEO pro-gram (draft) (Continued)

5-5

EM 1110-2-181031 Jan 95

Figure 5-2. (Concluded)

5-6

EM 1110-2-181031 Jan 95

Therefore, the coastal scientist should account for suchvariations as potential sources of error in making orinterpreting shoreline change maps. Section 5-5 containsa more detailed discussion of aerial photograph analysis.

g. Satellite remotely sensed data.

(1) Satellite data are available from U.S. agencies, theFrench Systeme Pour L’Observation de la Terre (SPOT)satellite data network, and from Russian coverage.1 Inmost instances, the data can be purchased either as photo-graphic copy or as digital data tapes for use in computerapplications. Imagery and digital data may assist inunderstanding large-scale phenomena, especially processeswhich are indicators of geologic conditions and surfacedynamics. Agencies that collect and distribute satellitedata are listed in Appendix E. Numerous remote sensingreferences are listed in Lampman (1993). A listing ofsatellite data maintained by the National Space ScienceData Center (NSSDC) is printed in Horowitz and King(1990). This data can be accessed electronically.

(2) Satellite data are especially useful for assessinglarge-scale changes of the surface of the coastal zone. Inthe vicinity of deltas, estuaries, and other sediment-ladenlocations, spatial patterns of suspended sediment can bedetected with remote sensing (Figure 5-3). In shallownon-turbid water bodies, some features of the offshorebottom, including the crests of submarine bars and shoals,can be imaged. The spatial extent of tidal flows may bedetermined using thermal infrared data, which can behelpful in distinguishing temperature differences of ebband flood flows and freshwater discharges in estuaries. Indeeper waters, satellites can also provide data on oceancurrents and circulation (Barrick, Evans, and Weber1977). Aircraft-mounted radar data also show consider-able promise in the analysis of sea state.

(3) The Landsat satellite program was developed bythe National Aeronautics and Space Administration incooperation with the U.S. Department of the Interior.When it began in 1972, it was primarily designed as anexperimental system to test the feasibility of collectingearth resource data from unmanned satellites. Landsatsatellites have used a variety of sensors with differentwavelength sensitivity characteristics, ranging from thevisible (green) to the thermal infrared with a maximum

_____________________________1Russian Sojuzkarta satellite photographs are availablefrom Spot Image Corporation (Appendix E). Almazsynthetic aperture radar data are available from HughesSTX Corporation.

wavelength of 12 micrometers (µm). Figure 5-4 showsbandwidths and spatial resolution of various satellite sen-sors. Of the five Landsat satellites, only Landsat-4 andLandsat-5 are currently in orbit. Both are equipped withthe multispectral scanner, which has a resolution of 82 min four visible and near-infrared bands, and the thematicmapper, which has a resolution of 30 m in six visible andnear- and mid-infrared bands and a resolution of 120 m inone thermal infrared band (10.4-12.5 µm).

(4) SPOT is a commercial satellite program. Thefirst satellite, which was sponsored primarily by theFrench government, was launched in 1986. The SPOT-1satellite has two identical sensors known as HRV (high-resolution-visible) imaging systems. Each HRV can func-tion in a 10-m resolution panchromatic mode with onewide visible band, or a 20-m resolution multispectral(visible and near infrared) mode with three bands(Figure 5-3).

(5) Several generations of satellites have flown inthe NOAA series. The most recent ones contain theAdvanced Very High Resolution Radiometer (AVHRR).This provides increased aerial coverage but at muchcoarser resolution than the Landsat or SPOT satellites.More information on the wide variety of satellites can befound in textbooks on remote sensing (i.e. Colwell 1983,Lillesand and Kiefer 1987, Richards 1986, Sabins 1987,Siegal and Gillespie 1980, Stewart 1985).

(6) Aircraft-mounted scanners, including thermalsensors and radar and microwave systems, may also haveapplications in coastal studies. LIDAR (light detectionand ranging), SLAR (Side-Looking Airborne Radar), SAR(Synthetic Aperture Radar), SIR (shuttle imaging radar),and passive microwave systems have applications includ-ing mapping of bottom contours of coastal waters. ALIDAR system, known as SHOALS (Scanning Hydro-graphic Operational Airborne Lidar System), is now beingused by the U.S. Army Corps of Engineers to profilecoastal areas and inlets. The system is based on thetransmission and reflection of a pulsed coherent laser lightfrom a helicopter equipped with the SHOALS istrumentpod and with data processing and navigation equipment(Lillycrop and Banic 1992). In operation, the SHOALSlaser scans an arc across the helicopter’s flight path, pro-ducing a survey swath equal to about half of the aircraftaltitude. A strongly reflected return is recorded from thewater surface, followed closely by a weaker return fromthe seafloor. The difference in time of the returns isconverted to water depth. SHOALS may revolutionizehydrographic surveying in shallow water for several

5-7

EM 1110-2-181031 Jan 95

Figure 5-3. SPOT satellite image, Atchafalaya Bay, LA. Suspended sediment from runoff is clearly visible. Dataprocessed by the Earthscan Laboratory, School of Geosciences, Louisiana State University, Baton Rouge, LA

reasons. The most important advantage is that the systemcan survey up to 8 square km per hour, thereby coveringlarge stretches of the coast in a few days. This enablesalmost instantaneous data collection along shores subjectto rapid changes. The system can be mobilized quickly,allowing large-scale post-storm surveys or surveys ofunexpected situations such as breaches across barriers.

Finally, minimum survey water depth is only 1 m; thisallows efficient coverage of shoals, channels, or breachesthat would normally be impossible or very difficult tosurvey using traditional methods, especially in winter.Maximum survey depth is proving to be about 10 m,depending on water clarity.

5-8

EM 1110-2-181031 Jan 95

Figure 5-4. Spectral resolution and approximate spatial resolution of sensors on Landsat, SPOT, and NOAA sat-ellites (from Earth Observation Satellite Company literature and Huh and Leibowitz (1986))

h. Topographic and bathymetric data.

(1) Topographic and bathymetric maps are availablefrom the USGS, many USACE District Offices, and theUSCGS. USGS topographic maps are generally revisedevery 20 to 30 years, and sometimes more often in areas

determined to be of high priority. Nevertheless, the mapsmay be outdated for some studies because of theephemeral nature of many coastlines. The USGS quad-rangles are available in a 7.5’ series (scale 1:24,000) anda 15’ series (scale 1:62,500). The resolution of thesemaps is typically inadequate to provide details of surface

5-9

EM 1110-2-181031 Jan 95

features, but may be sufficient for examining largelandforms and pronounced changes, particularly overlong periods.

(2) Recent and historic hydrographic survey data areavailable from the National Ocean Survey (NOS). Muchof this data can be obtained in the form of preliminaryplots that are of larger scale and contain more soundingsand bottom notations than the published charts made fromthem.

(3) Bathymetric survey maps are sometimes out ofdate because geomorphic changes in many submarineareas occur rapidly. On some navigation charts, the bath-ymetry may be more than 50 years old and the markeddepths may be quite different from actual depths. Thegreatest changes can be areas of strong current activity, ofstrong storm activity, of submarine mass movement, andof dredging near ship channels. The user must also beaware of changes in the datum used in different maps.Annual or more frequent hydrographic surveys areavailable at most Federal navigation projects.

i. Shoreline change maps.

(1) Shoreline changes may be interpreted from navi-gation maps, topographic maps, aerial photographs, andproperty records. In some areas, maps showing shorelinechanges and land loss may have been produced by stateand federal agencies, universities, or engineering firms.However, the user should be aware of potential sources oferror which may not have been adequately corrected whenthese maps were prepared.

(2) Shoreline and coastal change maps that are con-structed from historic maps and photographs are subject tonumerous sources of error. For example, maps may nothave common datums, may have different scales, mayhave variable accuracy due to age or loss of accuracy inpublication procedures, and may be based on differentprojections which in turn cause geometric distortions.Ideally, shoreline change maps constructed from aerialphotographs should be corrected for distortions caused bypitch, tilt, and yaw of the aircraft. Difficulties in identify-ing common points over time, problems in rectifyingscale, and distortions near margins and corners are com-mon. Additional problems include the unavailability ofphotographs of the desired vintage, scale, clarity, or reso-lution. Haze, fog, and cloud cover may obscure groundfeatures. Finally, the water level at the time that thephotographs were taken can greatly influence the positionof the shorelines. Specific data sources and procedures

for analyzing shoreline change maps are presented insection 5-5.

5-3. Field Data Collection and Observation

a. Background.

(1) In order to apply appropriate technologies to afield study, the coastal scientist should know somethingabout the nature of the problem and the expected out-come. For example, if a community is being threatenedby erosion, measurements of processes, topography, andbathymetry are needed to determine storm-induced andlong-term erosion trends. Also, studies of historical datamay be required to determine the rates and spatial vari-ability of shoreline change over time. Studies involvingstratigraphy may be required if the purpose is to find localsources of borrow material for beach nourishment.Design of a research study must include thorough plan-ning of objectives and sampling strategies, given time,logistic, and budget constraints. Much time and effort canbe wasted during a field study if the research objectivesare not well-defined and the sampling plan isinappropriate.

(2) Before undertaking detailed field studies, it isimportant to review all available coastal data pertinent tothe study area and problems. The existing information iscritical to the effective design of field studies and canresult in more cost-effective field work. Often, time andbudget constraints may severely limit data collection,making available data even more important.

(3) While in the field, relevant data and informationshould be meticulously recorded in water-resistant fieldbooks. Details can also be recorded on a tape recorder.Photographs serve as valuable records of field conditions,sampling equipment, and procedures. Video recorders arebeing increasingly used during field reconnaissance.

(4) The type of work conducted in the field may fallinto several categories. It may range from a simple visualsite inspection to a detailed collection of process measure-ments, sediment samples, stratigraphic samples, topo-graphic and bathymetric data, and geophysical data.Studies may include exploring the acting forces, rates ofactivity, interactions of forces and sediments, and varia-tions in activity over time. If the field work will involveextensive data collection, a preliminary site visit is highlyrecommended to help determine site conditions and todevelop a sampling plan.

5-10

EM 1110-2-181031 Jan 95

(5) Spatial and temporal aspects of site inspection areimportant considerations. The spatial dimensions of thesampling plan should have adequate longshore and cross-shore extent and an adequate grid or sample spacing withwhich to meet study objectives. Temporal considerationsinclude the frequency of sampling and the duration overwhich samples will be collected. Sampling frequency andduration are most important in modern process studies,such as monitoring the topographic and bathymetricchanges associated with storms. Studies of paleoenviron-mental or geologic time scales usually do not requirerepetitive visits, but thorough spatial sampling is critical.

(6) A conceptual model is essential before designinga field data collection program. This “model” is a set ofworking hypotheses which use existing knowledge toorganize missing information. As information is acquired,the conceptual model is revised and validated. Additionalobservations may be required to test a wider variety ofconditions, and conceptual models may need to be reviseddepending on the results of the study.

b. Site inspection and local resources.

(1) A general site inspection can provide insightstoward identifying significant research problems at a studyarea, in verifying and enhancing data from aerial photo-graphs and remote sensing sources, and in developingsampling strategies for more rigorous field work. Evenfor a brief site visit, thorough preparation is stronglyrecommended. Preparation should include reviewing thepertinent geologic, oceanographic, and engineering litera-ture, compiling maps and photographs, and understandingthe scope of the problem or situation. The field inspec-tion should include observations by all members to beinvolved in the project, if at all possible.

(2) The duration of the field examination must besufficient to assess the major objectives of the study.Local residents, existing data records, and field monitor-ing equipment may need to be used. A site inspectionshould include observation of marine forces and pro-cesses, assessment of geomorphic indicators, visits toneighboring sites, and interviews with residents and otherlocal or knowledgeable individuals. Questions to beasked might include what, why, when, where, and howcome? Why does this section of the shore look as itdoes? How do humans influence the local environment?Is the problem geologic (natural) or man-made? Docatastrophic events, such as hurricanes, appear to havemuch impact on the region? A checklist of data to becollected at a coastal site visit is presented in Appen-dix H. A handy field notebook of geologic data sheets is

published by the American Geologic Institute (Dietrich,Durto, and Foose 1982).

c. Photographs and time sequences. Photography isoften an important tool for initial reconnaissance work aswell as for more detailed assessments of the study area.One special application of cameras involves the use oftime-lapse or time interval photography, which may behelpful in studies of geomorphic variability to observeshoreline conditions, sand transport (Cook and Gorsline1972), and wave characteristics. If the camera is set torecord short-term processes, relatively frequent photo-graphs are typically obtained. If historic ground photo-graphs are available, additional pictures can be acquiredfrom the same perspective. Changes in an area over time,applicable to both short- and long-term studies, can alsobe recorded with video photography. It is important thatpertinent photographic information be recorded in a fieldlog:

• Date.

• Time.

• Camera location.

• Direction of each photograph.

• Prominent landmarks, if any.

Date, location, and direction should be marked on slidemounts for each exposure.

d. Wave measurements and observations. It is oftenrelevant in studies of historic and process time scales toobtain data regarding wave conditions at the site. Instru-mented wave gauges typically provide the most accuratewave data. Unfortunately, wave gauges are expensive topurchase, deploy, maintain, and analyze. Often, they areoperated for a short term to validate data collected byvisual observation or hindcasting methods. Multiplegauges, set across the shore zone in shallow and deepwater, can be used to determine the accuracy of wavetransformation calculations for a specific locale.

(1) Types of wave gauges.

(a) Wave gauges can be separated into two generalgroups: directional and non-directional. In general, direc-tional gauges and gauge arrays are more expensive tobuild, deploy, and maintain than non-directional gauges.Nevertheless, for many applications, directional

5-11

EM 1110-2-181031 Jan 95

instruments are vital because the directional distribution ofwave energy is an important parameter in many applica-tions, such as sediment transport analysis and calculationof wave transformation. Wave gauges can be installed inbuoys, placed directly on the sea or lake bottom, ormounted on existing structures, such as piers, jetties, oroffshore platforms.

(b) Of the non-directional wave gauges, buoy-mounted systems such as the Datawell Waverider areaccurate and relatively easy to deploy and maintain. Dataare usually transmitted by radio between the buoy and anonshore receiver and recorder. Bottom-mounted pressuregauges measure water level changes by sensing pressurevariations with the passage of each wave. The gauges areeither self-recording or are connected to onshore recordingdevices with cables. Bottom-mounted gauges must bemaintained by divers unless the mount can be retrieved byhoisting from a workboat. Internal-recording gaugesusually need more frequent maintenance because the datatapes must be changed or the internal memory down-loaded. Advantages and disadvantages of self-containedand cable-telemetered gauges are listed in Table 5-1.Structure-mounted wave gauges are the most economicaland most accessible of the non-directional gauges,although their placement is confined to locations wherestructures exist. The recording devices and transmitterscan be safely mounted above water level in a protectedlocation.

(c) Directional wave gauges are also mounted inbuoys or on the seafloor (Figure 5-5). Arrays of non-directional gauges can be used for directional wave analy-ses. Directional buoy-type wave gauges are oftendesigned to collect other parameters such as meteorology.

(2) Placement of wave gauges. The siting of wavegauges along the coast depends on the goals of the moni-toring project, funds and time available, environmentalhazards, and availability of previously collected data.There are no firm guidelines for placing gauges at a site,and each project is unique. There are two approaches towave gauging: one is to deploy instruments near a proj-ect site in order to measure the wave and sea conditionsthat directly affect a structure or must be accounted for indesigning a project. The second approach is to deploy agauge further out to sea to measure regional, incidentwaves. In the past, when wave gauges were exceedinglyexpensive, researchers often opted to collect regional datawith a single instrument. Now, with lower costs for hard-ware and software, we recommend that several gauges bedeployed near the coast flanking the project area. Apriori knowledge of a site or practical considerations may

dictate gauge placement. The user must usuallycompromise between collecting large amounts of data fora short, intensive experiment, and maintaining the gaugesat sea for a longer period in order to try to observe sea-sonal changes. Table 5-2 summarizes some suggestedpractices based on budget and study goals. Suggestionson data sampling intervals are discussed in Section 5-5.

(3) Seismic wave gauge. Wave estimates based onmicroseismic measurements are an alternative means toobtain wave data in high-energy environments. Micro-seisms are very small ground motions which can bedetected by seismographs within a few kilometers of thecoast. It is generally accepted that microseisms arecaused by ocean waves and that the amplitudes and peri-ods of the motions correspond to the regional wave cli-mate. Comparisons of seismic wave gauges in Oregonwith in situ gauges have been favorable (Howell and Rhee1990; Thompson, Howell, and Smith 1985). The seismicsystem has inherent limitations, but deficiencies in waveperiod estimates can probably be solved with moresophisticated processing. Use of a seismometer for wavepurposes is a long-term commitment, requiring time tocalibrate and compare the data. The advantage of a seis-mograph is that it can be placed on land in a protectedbuilding.

e. Water level measurements and observations.

(1) To collect continuous water level data for site-specific, modern process studies, tide gauges must bedeployed near the project site. Three types of instrumentsare commonly used to measure water level:

(a) Pressure transducer gauges.These instrumentsare usually mounted on the seafloor or attached to struc-tures. They record hydrostatic pressure, which is con-verted to water level during data processing. A majoradvantage of these gauges is that they are underwater andsomewhat inaccessable to vandals. In addition, ones likethe Sea Data Temperature Depth Recorder are compactand easy to deploy.

(b) Stilling-well, float gauges. These instruments,which have been in use since the 1930’s, consist of afloat which is attached to a stylus assembly. A clockworkor electric motor advances chart paper past the stylus,producing a continuous water level record. The float iswithin a stilling well, which dampens waves and boatwakes. The main disadvantage of these gauges is thatthey must be protected from vandals. They are usuallyused in estuaries and inland waterways where piles or

5-12

EM 1110-2-181031 Jan 95

Table 5-1Self-Contained and Cable-Telemetry Wave Gauges; Advantages and Disadvantages

I. Self-contained gauges

A. Advantages

1. Deployment is often simple because compact instrument can be handled by a small dive team.2. Gauge can be easily attached to piles, structural members, or tripods.3. Field equipment can be carried by airplane to remote sites.4. Gauges will continue to function in severe storms as long as the mounts survive.5. Usually easy to obtain permits to deploy instruments (typically, notification to mariners must be posted).

B. Disadvantages

1. Gauge must be periodically recovered to retrieve data or replace storage media.2. Data collection time is limited by the capacity of the internal memory or data tapes. Researcher must compromise between sampling

density and length of time the gauge can be gathering data between scheduled maintenance visits.3. Battery capacity may be a limiting factor for long deployments.4. If bad weather forces delay of scheduled maintenance, gauge may reach the limit of its storage capacity. This will result in unsampled

intervals.5. While under water, gauge’s performance cannot be monitored. If it fails electronically or leaks, data are usually lost forever.6. Gauge may be struck by anchors or fishing vessels. The resulting damage or total loss may not be detected until the next mainte-

nance visit.

C. Notes

1. Data compression techniques, onboard data processing, and advances in low-energy memory have dramatically increased the storagecapacity of underwater instruments. Some can remain onsite as long as 12 months.

II. Data transmission by cable

A. Advantages

1. Data can be continuously monitored. If a failure is detected (by human analysts or error-checking computer programs), a repair teamcan be sent to the site immediately.

2. Because of the ability to monitor the gauge’s performance, infrequent inspection visits may be adequate to maintain systems.3. Frequency and density of sampling are only limited by the storage capacity of the shore-based computers.4. Gauge can be reprogrammed in situ to change sampling program.5. Electrical energy is supplied from shore.

B. Disadvantages

1. Permitting is difficult and often requires considerable effort.2. Lightning is a major cause of damage and loss of data.3. Cable to shore is vulnerable to damage from anchors or fishing vessels.4. Shore station may be damaged in severe storms, resulting in loss of valuable storm data.5. Shore station and data cable are vulnerable to vandalism.6. Backup power supply necessary in case of blackouts.7. Installation of cable can be difficult, especially in harbors and across rough surf zones.8. Installation often requires a major field effort, with vehicles on beach and one or two boats. Heavy cable must be carried to the site.9. Cable eventually deteriorates in the field and must be replaced.10. Cable may have to be removed after experiment has ended.

C. Notes

1. Some cable-based gauges have internal memory and batteries so that they can continue to collect data even if cable is severed.2. Ability to constantly monitor gauge’s performance is a major advantage in conducting field experiments.

bridges are available for mounting the well and recordingbox. Figure 5-6 is an example of tide data from Chocta-whatchee Bay, Florida.

(c) Staff gauges. Water levels are either recordedmanually by an observer or calculated from electric resis-tance measurements. The resistance staff gauges requirefrequent maintenance because of corrosion and biologicalfouling. The manual ones are difficult to use at night and

during storms, when it is hazardous for the observer to beat the site.

Typically, water level measurements recorded by gaugesare related to an established datum, such as mean sealevel. This requires that the gauge elevations be accu-rately measured using surveying methods. The maximumwater level elevations during extreme events can also be

5-13

EM 1110-2-181031 Jan 95

Figure 5-5. Bottom-mounted Sea Data TM 635-12 directional wave gauge mounted in tripod using railroad wheels ascorner weights

determined by examining water marks on structures orother elevated features.

(2) Water level information over paleoenvironmentaltime scales has been investigated by researchers usingstratigraphic coring, seismic techniques, and radiometricdating. Petroleum geologists have used seismic stratigra-phy to reconstruct ancient sea levels (Payton 1977, Sheriff1980).

f. Current measurements and observations.

(1) General techniques of current measurement.

(a) The observation of hydraulic phenomena can beaccomplished by two general approaches. One of these,Lagrangian, follows the motion of an element of matter in

its spatial and temporal evolution. The other, Eulerian,defines the motion of the water at a fixed point and deter-mines its temporal evolution. Lagrangian current measur-ing devices are often used in sediment transport studies,in pollution monitoring, and for tracking ice drift.Eulerian, or fixed, current measurements are important fordetermining the variations in flow over time at a fixedlocation. Recently developed instruments combine aspectsof both approaches.

(b) Four general classes of current measuring tech-nology are presently in use (Appell and Curtin 1990):

• Radar and Lagrangian methods.

• Spatially integrating methods.

5-14

EM 1110-2-181031 Jan 95

Table 5-2Suggested Wave Gauge Placement for Coastal Project Monitoring

I. High-budget project (major harbor; highly populated area)

A. Recommended placement:

1. One (or more) wave gauge(s) close to shore near the most critical features being monitored (example, near an inlet). Although near-shore, gauges should be in intermediate or deep water based on expected most common wave period. Depth can be calculated fromformulas in Shore Protection Manual (1984).

2. In addition, one wave gauge in deep water if needed for establishing boundary conditions of models.

B. Schedule:

1. Minimum: 1 year. Monitor winter/summer wave patterns (critical for Indian Ocean projects).

2. Optimum: 5 years or at least long enough to determine if there are noticeable changes in climatology over time. Try to include one ElNiño season during coverage for North American projects.

C. Notes:

1. Concurrent physical or numerical modeling: Placement of a gauge may need to take into account modellers’ requirements for input ormodel calibration.

2. Preexisting wave data may indicate that gauges should be placed in particular locations. As an alternative, gauges may be placed inlocations identical to the previous deployment in order to make the new data as compatible as possible with the older data. Long,continuous data sets are extremely valuable!

3. Hazardous conditions: If there is a danger of gauges being damaged by anchors or fishing boats, the gauges must be protected,mounted on structures (if available), or deployed in a location which appears to be the least hazardous.

II. Medium-budget project

A. Recommended placement:

1. One wave gauge close to shore near project site.

2. Obtain data from nearest NOAA National Data Buoy Center (NDBC) buoy for deepwater climatology.

B. Schedule: minimum 1 year deployment; longer if possible

C. Notes: same as IC above. Compatibility with existing data sets is very valuable.

III. Low budget, short-term project

A. Recommended placement: gauge close to project site.

B. Schedule: if 1-year deployment is not possible, try to monitor the season when the highest waves are expected (usually winter,although this may not be true in areas where ice pack occurs).

C. Notes: same as IC above. It is critical to use any and all data from the vicinity, anything to provide additional information on the waveclimatology of the region.

• Point source and related technology.

• Acoustic Doppler Current Profilers (ADCP) andrelated technology.

The large number of instruments and methods used tomeasure currents underscores that detection and analysisof fluid motion in the oceans is an exceedingly complexprocess. The difficulty arises from the large continuousscales of motion in the water. As stated by McCullough(1980), “There is no single velocity in the water, butmany, which are characterized by their temporal andspatial spectra. Implicit then in the concept ofa fluid’velocity’ is knowledge of the temporal and spatial aver-aging processes used in measuring it. Imprecise, orworse, inappropriate modes of averaging in time and/orspace now represent the most prominent source of error in

near-surface flow measurements.” McCullough’s com-ments were addressed to the measurement of currents inthe ocean. In shallow water, particularly in the surf zone,additional difficulties are created by turbulence and airentrainment caused by breaking waves, by suspension oflarge concentrations of sediment, and by the physicalviolence of the environment. Trustworthy current mea-surement under these conditions becomes a daunting task.

(2) Lagrangian.

(a) Dye, drogues, ship drift, bottles, temperaturestructures, oil slicks, radioactive materials, paper, woodchips, ice, trees, flora, and fauna have all been used tostudy the surface motion of the oceans (McCullough1980). Some of these techniques, along with the use of

5-15

EM 1110-2-181031 Jan 95

Figure 5-6. Tidal elevations from seven stations in Choctawhatchee Bay, FL, and the Gulf of Mexico. The overallenvelope of the seven curves is similar, but individual peaks are shifted in phase from station to station. Originaltide records courtesy of U.S. Army Engineer (USAE) District, Mobile

5-16

EM 1110-2-181031 Jan 95

mid-depth drogues and seabed drifters, have been widelyused in coastal studies. A disadvantage of all drifters isthat they are only quasi-Lagrangian sensors because,regardless of their design or mass, they cannot exactlyfollow the movement of the water (Vachon 1980). Nev-ertheless, they are particularly effective at revealing sur-face flow patterns if they are photographed or videorecorded on a time-lapse basis. Simple drifter exper-iments can also be helpful in developing a samplingstrategy for more sophisticated subsequent field investiga-tions. Floats, bottom drifters, drogues, and dye are usedespecially in the littoral zone where fixed current metersare adversely affected by turbulence. Resio and Hands(1994) analyze the use of seabed drifters and comment ontheir value in conjunction with other instruments.

(b) High frequency (HF) radar surface-current map-ping systems have been tested since the 1970’s. Theadvantage of using the upper high radar frequencies isthat these frequencies accurately assess horizontal currentsin a mean water depth of only 1 m (total layer thicknessabout 2 m). Hence, HF radar accurately senses horizontalcurrents in the uppermost layers of the oceans, whereother instruments such as moored current meters andADCP’s become inoperable (Barrick, Lipa, and Lilleboe1990). Nevertheless, HF radar has had limited success inthe oceanography community because of the difficulty inproving measurement accuracy and because of relativelyhigh system costs (Appell and Curtin 1990).

(c) Large-scale coastal circulation can be observed insatellite images, as seen in Figure 5-3.

(3) Spatially integrating methods. To date, experi-ments in spatially averaging velocity by observinginduced electrical fields have been conducted by towingelectrodes from ships or by sending voltages in abandonedunderwater telephone cables. Some of these experimentshave been for the purpose of measuring barotropic flow inthe North Pacific (Chave, Luther, and Filloux 1990; Spain1990 - these two papers provide a substantial summary ofthe mathematics and methods). This author is unaware ofwhether these techniques have been tested in shallowwater or in restricted waterways such as channels. At thistime, therefore, spatially integrating methods appear tohave no immediate application to coastal engineeringstudies.

(4) Point source (Eulerian) and related technology.

(a) In channels, bays, and offshore, direct measure-ments of the velocity and direction of current flow can bemade by instruments deployed on the bottom or at various

levels in the water column. Two general classes of cur-rent meters are available: mechanical (impeller-type) andelectronic. Several types of electronic current meters arein common use, including electromagnetic, inclinometer,and acoustic travel-time (Fredette et al. 1990, McCullough1980; Pinkel 1980).

(b) Impeller current meters measure currents bymeans of a propeller device which is rotated by the cur-rent flow. They serve as approximate velocity componentsensors because they are primarily sensitive to the flowcomponent in a direction parallel to their axle. Varioustypes of propeller design have been used to measure cur-rents, but experience and theoretical studies have shownthat the ducted propellers are more satisfactory in measur-ing upper ocean currents than rotor/vane meters (Davisand Weller 1980). Impeller/propeller meters are consid-ered to be the most reliable in the surf zone (Teleki,Musialowski, and Prins 1976), as well as the leastexpensive. One model, the Endeco 174, has been widelyused by CERC for many years throughout the country.Impeller gauges are subject to snarling, biofouling, andbearing failures, but are more easily repaired in the fieldand are more easily calibrated than other types (Fredetteet al. 1990).

(c) Electronic current meters have many features incommon, although they operate on different principles.Their greatest common advantages are rapid response andself-contained design with no external moving parts.They can be used in real-time systems and can be used tomeasure at least two velocity components. The degree ofexperience of the persons working with the instrumentsprobably has more influence on the quality of dataacquired than does the type of meter used (Fredette et al.1990). The InterOcean Systems S4 electromagnetic meterhas been successfully used by CERC at field experiments.

(5) ADCPS. These profilers operate on the principleof Doppler shift in the backscattered acoustic energycaused by moving particles suspended in the water.Assuming that the particles have the same velocity as theambient water, the Doppler shift is proportional to thevelocity components of the water within the path of theinstrument’s acoustic pulse (Bos 1990). The back-scattered acoustic signal is divided into parts correspond-ing to specific depth cells, often termed “bins.” The binscan be various sizes, depending upon the depth of waterin which the instrument has been deployed, the frequencyof the signal pulse, the time that each bin is sampled, andthe acceptable accuracy of the estimated current velocity.Much excitement has been generated by ADCP’s, bothamong scientists working in shallow water and in the

5-17

EM 1110-2-181031 Jan 95

deep ocean (a comprehensive bibliography is listed inGordon et al. (1990)). A great advantage of usingADCP’s in shallow water is that they provide profiles ofthe velocities in the entire water column, providing morecomprehensive views of water motions than do strings ofmultiple point source meters. ADCP data are inherentlynoisy, and signal processing and averaging are critical tothe successful performance of the gauges (Trump 1990).

(6) Indirect estimates of currents. Indirect estimatesof current speed and direction can be made from theorientation, size, and shape of bed forms, particularly inshallow water. Widespread use of side-scan sonar hasmade this type of research possible in bays, inlets, andoffshore. Sedimentary structures on the seafloor arecaused by the hydrodynamic drag of moving water actingon sediment particles. The form and shape of bottomstructures reflect the effects and interaction among tidalcurrents, waves, riverine flow, and longshore currents.These complex interactions especially affect bedforms intidal channels and other restricted waterways. Bedformsreflect flow velocity, but are generally independent ofdepth (Clifton and Dingler 1984; Boothroyd 1985). Theirshape varies in response to increasing flow strength(Hayes and Kana 1976). Bedform orientation and associ-ated slipfaces also provide clues to flow direction(Morang and McMaster 1980; Wright, Sonu, and Kielhorn1972).

g. Grab sampling and samplers.

(1) Seafloor sediments in coastal areas can showgreat spatial and temporal variation. The surface sedi-ments may provide information about the energy of theenvironment as well as the long-term processes and move-ment of materials, such as sediment transport pathways,sources and sinks. Bed surface sediments are typicallycollected with grab samplers and then analyzed usingstandard laboratory procedures. These tests are describedin detail in other sources (Fredette et al. 1990; Buller andMcManus 1979).

(2) There are a variety of grab type samplers of dif-ferent sizes and design that are used for collecting surfacesediment samples (described in detail in Bouma (1969)).Most consist of a set of opposing, articulatedscoop-shaped jaws that are lowered to the bottom in anopen position and are then closed by various trip mecha-nisms to retrieve a sample. Many grab samplers are smallenough to be deployed and retrieved by hand; othersrequire some type of lifting gear. If there is gravel in thesample, at least 2 to 3 litres of sample are needed forreliable grain size distribution testing.

(3) A simple and inexpensive dredge sampler can bemade of a section of pipe that is closed at one end. It isdragged a short distance across the bottom to collect asample. Unlike grab samples, the dredged samples arenot representative of a single point and may have lostfiner material during recovery. However, dredge samplersare useful in areas where shells or gravel which preventcomplete closure of the jaws are present.

(4) Although obtaining surficial samples is helpfulfor assessing recent processes, it is typically of limitedvalue in stratigraphic study because grab samplers usuallyrecover less than 15 cm of the sediment. Generally, theexpense of running tracklines in coastal waters for thesole purpose of sampling surficial sediments is not eco-nomically justified unless particularly inexpensive boatscan be used. Occasionally, grab and dredge samples aretaken during geophysical surveys, but the sampling opera-tions require the vessel to stop at each station, thus losingsurvey time and creating interrupted data coverage. Pre-cise offshore positioning now allows grab samples to becollected at specific locations along the boat’s track afterthe survey has been run and the data examined.

h. Stratigraphic sampling.

(1) Sediments and sedimentary rock sequences are arecord of the history of the earth and its changing envi-ronments, including sea-level changes, paleoclimates,ocean circulation, atmospheric and ocean geochemicalchanges, and the history of the earth’s magnetic field. Byanalyzing stratigraphic data, age relations of the rockstrata, rock form and distribution, lithologies, fossilrecord, biopaleogeography, and episodes of erosion anddeposition at a coastal site can be determined. Erosionremoves part of the physical record, resulting in unconfor-mities. Often, evidence of erosion can be interpretedusing physical evidence or dating techniques.

(2) Sediment deposits located across a zone thatranges from the maximum water level elevation to thedepth of the wave base are largely indicative of recentprocesses. Within this zone in unconsolidated sediments,simple reconnaissance field techniques are available forcollecting data. The techniques often use ordinary con-struction equipment or hand tools. Smaller efforts requireshovels, hand augers, posthole diggers, or similar hand-operated devices. Larger-scale efforts may includetrenches, pits or other large openings created for visualinspection, sample collection, and photography (Fig-ure 5-7). A sedimentary peel can be taken from theexposed surface. The peel retains the original

5-18

EM 1110-2-181031 Jan 95

Figure 5-7. Trench excavated in the edge of a sand dune, eastern Alabama near Alabama/Florida state line

arrangement of sedimentary properties (Bouma 1969).Often, undisturbed chunk or block samples and disturbedjar or bag samples are carved from these excavations andtaken back to the laboratory.

(3) Rates and patterns of sedimentation can be deter-mined using marker horizons. Marker horizons may

occur in relation to natural events and unintentionalhuman activities or they may be directly emplaced for theexpress purpose of determining rates and patterns of sedi-mentation. Recently, several studies have estimated ratesof sedimentation in marshes by spreading feldsparmarkers and later measuring the thicknesses of materialsdeposited on the feldspar with cryogenic coring devices.

5-19

EM 1110-2-181031 Jan 95

(4) The petrology and mineralogy of rock samplescan be used to identify the source of the sediment. Thiscan indicate if river flow has changed or if coastal cur-rents have changed directions. Mineralogy as it pertainsto sediment budgets is discussed in Meisburger (1993)and Wilde and Case (1977).

(5) Direct sampling of subbottom materials is oftenessential for stratigraphic studies that extend beyondhistoric time scales. Table 5-3 lists details on a numberof subaqueous sediment sampling systems that do notrequire drill rigs. One system listed in Table 5-3, thevibracorer, is commonly used by geologists to obtain sam-ples in the marine and coastal environment. Vibratorycorers consist of three main components: a frame, coringtube or barrel, and a drive head with a vibrator (Fig-ure 5-8). The frame consists of a quadrapod or tripodarrangement, with legs connected to a vertical beam. Thebeam supports and guides the core barrel and vibrator andallows the corer to be free-standing on the land surface orseafloor. The core may be up to 3 or 4 m long, which isadequate for borrow site investigations and many othercoastal studies.

(6) While common vibratory corers are capable ofpenetrating up to 5 m or more of unconsolidated sedi-ment, actual performance depends on the nature of thesubbottom material. Under unfavorable conditions, verylittle sediment may be recovered. Limited recoveryoccurs for several reasons, chief among these being lackof penetration of the core barrel. In general, stiff clays,gravel and hard-packed fine to very fine sands are usuallymost difficult to penetrate. Compaction and loss of mate-rial during recovery can also cause a discrepancy betweenpenetration and recovery. In comparison with rotary soilboring operations, vibratory coring setup, deployment,operation, and recovery are rapid. Usually a 3-m core canbe obtained in a manner of minutes. Longer cores requirea crane or some other means of hoisting the equipment, aprocedure that consumes more time, but is still compara-tively rapid. Success with vibracoring depends on someprior knowledge of sediment type in the region.

(7) Cores can be invaluable because they allow adirect, detailed examination of the layering and sequencesof the subsurface sediment in the study area. Thesequences provide information regarding the history of thedepositional environment and the physical processes dur-ing the time of sedimentation. Depending upon the infor-mation required, the types of analysis that can beperformed on the core include grain size, sedimentarystructures, identification of shells and minerals, organiccontent, microfaunal identification, (pollen counts) x-ray

radiographs, radiometric dating, and engineering tests. Ifonly information regarding recent processes is necessary,then a box corer, which samples up to 0.6-m depths, canprovide sufficient sediment. Because of its greater width,a box corer can recover undisturbed sediment from imme-diately below the seafloor, allowing the examination ofmicrostructure and lamination. These structures are usu-ally destroyed by traditional vibratory or rotary coring.

(8) If it is necessary to obtain deep cores, or if thereare cemented or very hard sediments in the subsurface,rotary coring is necessary. Truck- or skid-mounted drill-ing rigs can be conveniently used on beaches or on bargesin lagoons and shallow water. Offshore, rotary drillingbecomes more complex and expensive, usually requiringjack-up drilling barges or four-point anchored drill ships(Figure 5-9). An experienced drilling crew can sample100 m of the subsurface in about 24 hr. Information ondrilling and sampling practice is presented in EM 1110-1-1906 and Hunt (1984).

i. Sediment movement and surface forms. Of greatimportance in investigations of geologic history is tracingsediment movement. This includes identifying the loca-tions of sediment sources and sinks, quantifying sedimenttransport rates, and discovering the pathways. Sedimenttransportation is influenced by grain properties such assize, shape, and density, with grain size being mostimportant. Differential transport of coarse and fine, angu-lar and rounded, and light and heavy grains leads to grad-ing. Field visits to a locality are often repeated to assesstemporal variability of these phenomena. Simultaneousmeasurements of energy processes, such as current andwaves, are often required to understand the rates andmechanisms of movement.

(1) Measurement of sediment movement.

(a) The measurement of suspended and bed loadsediment movement in the surf zone is an exceedinglydifficult process. There are a variety of sampling devicesavailable for measuring suspended and bed load transportin the field (Dugdale 1981; Seymour 1989), but thesedevices have not performed properly under some condi-tions or have been expensive and difficult to use. Forthese reasons, new sampling procedures are being devel-oped and tested at CERC and other laboratories. Pointmeasurements of sediment movement can be performedby two general procedures:

• Direct sampling and weighing of a quantity ofmaterial.

5-20

EM 1110-2-181031 Jan 95

Table 5-3Subaqueous Soil Sampling Without Drill Rigs and Casing

Device Application Description Penetration depth Comments

Petersen dredge Large, relatively intact“grab” samples of sea-floor.

Clam-shell type grab weighingabout 1,000 lb with capacityabout 0.4 ft3

To about 4 in. Effective in water depths to200 ft. More with additionalweight.

Harpoon-type gravitycorer

Cores 1.5- to 6-in.-dia.in soft to firm soils.

Vaned weight connected to cor-ing tube dropped directly fromboat. Tube contains liners andcore retainer.

To about 30 ft. Maximum water depthdepends only on weight.Undisturbed (UD) samplingpossible with short, large-diameter barrels.

Free-fall gravitycorer

Cores 1.5- to 6-in. dia.in soft to firm soils.

Device suspended on wire ropeover vessel side at height aboveseafloor about 15 ft and thenreleased.

Soft soils to about17 ft.Firm soils to about10 ft.

As above for harpoon type.

Piston gravity corer(Ewing gravity corer)

2.5-in. sample in soft tofirm soils.

Similar to free-fall corer exceptthat coring tube contains a pistonthat remains stationary on theseafloor during sampling.

Standard core barrel10 ft; additional 10-ftsections can beadded.

Can obtain high-quality UDsamples.

Piggott explosivecoring tube

Cores of soft to hardbottom sediments.

Similar to gravity corer. Driveweight serves as gun barrel andcoring tube as projectile. Whentube meets resistance of sea-floor, weighted gun barrel slidesover trigger mechanism to fire acartridge. The exploding gasdrives tube into bottom sedi-ments.

Cores to 1-7/8 in. andto 10-ft lengths havebeen recovered in stiffto hard materials.

Has been used successfullyin 20,000 ft of water.

NorwegianGeotechnical Insti-tute gas-operatedpiston

Good-quality samplesin soft clays.

Similar to the Osterberg pistonsampler except that the piston onthe sampling tube is activated bygas pressure.

About 35 ft.

Vibracorer High-quality samples insoft to firm sediments.Dia. 3-1/2 in.

Apparatus is set on seafloor. Airpressure from the vessel acti-vates an air-powered mechanicalvibrator to cause penetration ofthe tube, which contains a plasticliner to retain the core.

Length of 20 and40 ft. Rate of penetra-tion varies with mate-rial strength. Samplesa 20-ft core in softsoils in 2 min.

Maximum water depth about200 ft.

Box corer Large, intact sliceof seafloor.

Weighted box with closure ofbottom for benthic biologicalsampling.

To about 1 ft. Central part of sample isundisturbed.

(Adapted from Hunt (1984))

• Detection of the fluid flow by electro-optical oracoustic instruments deployed in the water.

(b) Two general methods are available to directlysample the sediment in suspension and in bed load. First,water can be collected in hand-held bottles or can beremotely sucked into containers with siphons or pumpapparatus. The samples are then dried and weighed. Thesecond method is to trap a representative quantity of thesediment with a mesh or screen trap through which thewater is allowed to flow for a fixed time. A fundamentalproblem shared by both methods is the question of

whether the samples are truly representative of the sedi-ment in transport. For example, how close to the seabedmust the orifice be to sample bed load? If it is highenough to avoid moving bed forms, will it miss some ofthe bed load? Streamer traps made from mesh are inex-pensive to build but difficult to use. The mesh must besmall enough to trap most of the sediment but must allowwater to flow freely. Kraus (1987) deployed streamers atDuck, NC, from stainless steel wire frames (Figure 5-10).Kraus and Dean (1987) obtained the distribution of long-shore sand transport using sediment traps. At this time,sediment traps are still research tools and are not

5-21

EM 1110-2-181031 Jan 95

Figure 5-8. Front view of lightweight vibracorermounted on a barge

Figure 5-9. Rotary drilling operations underway from a4-point anchored drill ship. Drilling is conducted 24 hrusing two crews

commonly used. A fundamental limitation of traps is thatthey can usually only be used in mild conditions. Inwinter and during storms it is too hazardous for the fieldtechnicians to maintain the equipment. Perversely, it isunder these harsher conditions when the greatest sedimentmovement occurs. Another fundamental problem is relat-ing the instantaneous measured suspended and bedloadtransport to long-term sediment movement. Because ofthe extreme difficulty of conducting research in the surfzone, answers to these questions remain elusive.

(c) Electronic instruments are being developed todetect or estimate sediment transport. They have someadvantages over direct sampling procedures. Theseinclude the ability to measure the temporal variations ofsuspended or bed load sediment and the ability to be usedin cold water or in harsh conditions. (Note, however, thatin severe storms, essentially no man-made devices havesurvived in the surf zone.) Their disadvantages includethe difficulty of calibrating the sensors and testing theiruse with different types of sand and under differenttemperatures. In addition, many of these instruments areexpensive and not yet commonly available. Sternberg(1989) and Seymour (1989) discuss ongoing research todevelop and test new instruments for use in sedimenttransport studies in estuarine and coastal areas.

(d) Sediment movement, both bed load or total load,can also be measured with the use of natural and artificialtracers (Dugdale 1981). Heavy minerals are naturaltracers which have been used in studies of sedimentmovement (McMaster 1960; Wilde and Case 1977).Natural sand can also be labelled using radioactive iso-topes and fluorescent coatings (Arlman, Santema, andSvasek 1958; Duane 1970; Inman and Chamberlain 1959;Teleki 1966). Radioactive tracers are no longer usedbecause of health and safety concerns. When fluorescentdyes are used, different colors can be used simultaneouslyon different size fractions to differentiate between succes-sive experiments at one locality (Ingle 1966). Artificialgrains, which have the same density and hydraulicresponse of natural grains, can also be used in tracerstudies. Aluminum cobble has been used by Nicholls andWebber (1987) on rocky beaches in England. The alumi-num rocks were located on the beaches using metal detec-tors. Nelson and Coakley (1974) review artificial tracermethods and concepts.

(e) As with other phenomena, the experimentaldesign for tracer studies may be Eulerian or Lagrangian.For the time integration or Eulerian method, the tracergrains are injected at a constant rate over a given interval

5-22

EM 1110-2-181031 Jan 95

Figure 5-10. Side view of steel frame and polyester mesh sediment trap used at Duck, NC, by Kraus (1987) duringCERC’s DUCK-85 field experiments

of time. For the space integration or Langrangianmethod, the tracers are released over an area at the sametime. The choice of the method depends upon the natureof the problem. Field experiments must be designedcarefully to isolate the parameter of interest that is to bemeasured or traced. For example, if the purpose of thestudy is to assess bed-load transport, then care must betaken not to introduce tracers into the suspended load inthe water column.

(2) Use of subsurface structure to estimate flowregime. An introduction to bed form shape and nomen-clature has been presented in Chapter 4.

(a) Several useful indices of foreset laminae, whichmay assist in making qualitative estimates of the strengthof currents in modern and ancient sediments, are given byJopling (1966). These include: (1) maximum angle ofdip of foreset laminae (at low velocities the angle may

5-23

EM 1110-2-181031 Jan 95

exceed the static angle of repose whereas at high veloci-ties the angle is less than the static angle); (2) character ofcontact between foreset and bottomset (the contactchanges from angular to tangential to sigmoidal withincreasing velocity); (3) laminae frequency measured atright angles to bedding (there are more laminae per unitarea with increasing velocity); (4) sharpness or texturalcontrast between adjacent laminae (at higher velocitieslaminae become less distinct); and (5) occurrence ofregressive ripples (regressive ripples indicate relativelyhigher velocities).

(b) Measurements of bed forms can be accomplishedon exposed sand banks at low water using surveyingtechniques or large-scale aerial photographs. Dimension-less parameters of ripples and other bedforms can indicatedepositional environment (Tanner 1967). The flow direc-tions can be assessed in terms of the trace of the crestline(Allen 1968). Wave-formed structures reflect the velocityand direction of the oscillatory currents as well as thelength of the horizontal component of orbital motion andthe presence of velocity asymmetry within the flow(Clifton and Dingler 1984). The flow strength for inter-tidal estuarine bed forms can also be estimated for a givenflow depth by the velocity-depth sequence of bed forms(Boothroyd 1985).

j. Navigation and positioning equipment.

(1) Accurate positioning is essential for most geologi-cal monitoring studies. Several types of positioning andnavigation systems are available for coastal studies, withthe most common being Loran-C and Global PositioningSystems (GPS). Other technologies, such as short-rangemicrowave and optical systems, are also in common use(Fredette et al. 1990).

(2) Loran-C computes microsecond time differencesusing pulsed low-frequency radio waves between net-works and receivers. The differences are then computedas lines of position. The receivers can be used up toabout 2,000 km from the networks with reasonable accu-racy. The absolute accuracy of Loran-C varies from 180to 450 m, while the repeatable accuracy varies from 15 to90 m.

(3) Global Positioning System(GPS) is a revolutionin electronic navigation for the military because of itsunmatched ability to provide rapid and extremely accurateposition fixes around the world under all weather condi-tions. The system is not yet fully operational and someof the satellites have not yet been launched. Unfortu-nately for civilian users, the Department of Defense has

implemented a national security program calledSelectiveAvailability (SA), which deliberately degrades GPS accu-racy by distorting the satellite signals. There are twotypes of GPS receivers:

• Precise Positioning Systems(PPS), which areavailable only to the military and “approved civil-ians,” contain electronic chips which recognizeand correct the SA distortion.

• Standard Positioning Systems(SPS) are availablecommercially for boaters and civilians. The SAdistortion makes these units accurate to 100 m95 percent of the time and 300 m for the remain-ing 5 percent. The problem with SPS is thatcivilian users do not know the extent of the SAdistortion or when it is in effect, therefore makingit impossible to determine the level of accuracy ofthe GPS readings at any given time.

In an effort to provide accuracy of 12-20 m in harborsand harbor approaches, the U.S. Coast Guard has beendevelopingDifferential GPS (DGPS) for coastal waters.This procedure attempts to cancel the error which SAimposes. Using land-based receivers at specific, knownlocations (Coast Guard stations, lighthouses), the CoastGuard receives simultaneous signals from 12 satellites,determines the difference between the exact and GPS-reported locations, calculates a correction, and transmitsthe correction over local radio frequencies to nearby ves-sels. Boats must be equipped with special receivers todemodulate the signal and apply it to the GPS signal thatthe boat is receiving.

As of 1994, only 10 of the planned 47 U.S. DGPS sta-tions have been installed, and the operating stations arestill in the prototype stage. Several more years of adjust-ing and tuning are anticipated. Users can contact theCoast Guard’s GPS Information Center (Alexandria, VA;tel 703/313 5900) for up-to-date information on the statusof the system. In light of the developmental stage ofDGPS technology,users are cautioned against relyingon manufacturers’ claims of pinpoint accuracy. Thetechnology is simply not yet developed to this level. Useof GPS for USACE surveys is discussed inEM 1110-1-1003.

(4) Navigation (positioning) error standards havebeen established for USACE hydrographic surveys. Threegeneral classes of surveys have been defined(EM 1110-2-1003):

5-24

EM 1110-2-181031 Jan 95

• Class 1 - Contract payment surveys.

• Class 2 - Project condition surveys.

• Class 3 - Reconnaissance surveys.

Although the requirements of geologic site surveys maynot be the same as those of USACE hydrographic sur-veys, the accuracy standards are useful criteria whenspecifying quality control requirements in contractualdocuments. The frequency of calibration is the majordistinguishing factor between the classes of survey, anddirectly affects the accuracy and adequacy of the finalresults. With the increasing use of Geographic Informa-tion Systems (GIS) for analysis and manipulation of data,high standards of accuracy are imperative. Calibrationsare time-consuming and reduce actual data collectiontime. Nevertheless, this must be countered with the eco-nomic impact that low quality data may be useless or mayeven lead to erroneous conclusions (leading, in turn, toincorrectly designed projects and possible litigation).

(5) The maximum allowable tolerances for each classof survey are shown in Table 5-4.

(6) Table 5-5 depicts positioning systems which areconsidered suitable for each class of survey. The tablepresumes that the typical project is located within 40 km(25 miles) of a coastline or shoreline reference point.Surveys further offshore should conform to the standardsin the NOAA Hydrographic Manual (NOAA 1976). Plan-ning and successful implementation of offshore surveysare sophisticated activities and should be carried out bypersonnel or contractors with considerable experience anda successful record in achieving the accuracies specifiedfor the particular surveys.

k. Geophysical techniques.

(1) Geophysical survey techniques, involving the useof sound waves and high quality positioning systems on

ocean vessels, are widely used for gathering subsurfacegeological and geotechnical data in coastal environments.Geophysical procedures provide indirect subsurface dataas opposed to the direct methods such as coring and tren-ching. The use of geophysical methods can assist inlocating and correlating geologic materials and features bydetermining acoustic transparency, diffraction patterns,configuration and continuity of reflectors, and apparentbedding patterns. Inferences can often be made usingthese measures of stratigraphic and lithologic characteris-tics and important discontinuities. Table 5-6 lists frequen-cies of common geophysical tools.

(2) Fathometers or depth-sounders, side-scan sonar,and subbottom profilers are three major types of equip-ment used to collect geophysical data in marine explora-tion programs. All three systems are acoustic devices thatfunction by propagating acoustic pulses in the water andmeasuring the lapsed time between pulse initiation and thearrival of return signals reflected from various features onor beneath the bottom. These systems are used to obtaininformation on seafloor geomorphology, bottom featuressuch as ripple marks and rock outcrops, and the underly-ing rock and sediment units. Acoustic depth-sounders areused for conducting bathymetric surveys. Side-scan sonarprovides an image of the aerial distribution of sedimentand surface bed forms and larger features such as shoalsand channels. It can thus be helpful in mapping direc-tions of sediment motion. Subbottom profilers are used toexamine the near-surface stratigraphy of features belowthe seafloor.

(3) A single geophysical method rarely providesenough information about subsurface conditions to beused without actual sediment samples or additional datafrom other geophysical methods. Each geophysical tech-nique typically responds to several different physical char-acteristics of earth materials, and correlation of data fromseveral methods provides the most meaningful results.All geophysical methods rely heavily on experiencedoperators and analysts.

Table 5-4Maximum Allowable Errors for Hydrographic Surveys

Survey Classification

Type of Error1

Contract Payment2

Project Condition3

Reconnaissance

Resultant two-dimensional one-sigma RMS positional errornot to exceed 3 m 6 m 100 m

Resultant vertical depth measurement one-sigma standarderror not to exceed

± .152 m(± 0.5 ft)

± .305 m(± 1.0 ft)

± .457 m(± 1.5 ft)

(From EM 1110-2-1003)

5-25

EM 1110-2-181031 Jan 95

Table 5-5Allowable Horizontal Positioning System Criteria

Positioning System

EstimatedPositionalAccuracy

(meters, RMS)

Allowable forSurvey Class

1 2 3

Visual Range Intersection 3 to 20 No No Yes

Sextant Angle Resection 2 to 10 No Yes Yes

Transit/Theodolite Angle Intersection 1 to 5 Yes Yes Yes

Range Azimuth Intersection 0.5 to 3 Yes Yes Yes

Tag Line (Static Measurementsfrom Bank)< 457 m (1,500 ft) from baseline> 457 m (1,500 ft) but < 914 m (3,000 ft)> 914 m (3,000 ft) from baseline

0.3 to 11 to 5

5 to 50+

YesNoNo

YesYesNo

YesYesYes

Tag Line (Dynamic)< 305 m (1,000 ft) from baseline> 305 m (1,000 ft) but < 610 m (2,000 ft)> 610 m (2,000 ft) from baseline

1 to 33 to 6

6 to 50+

YesNoNo

YesYesNo

YesYesYes

Tag Line (Baseline Boat) 5 to 50+ No No Yes

High-Frequency EPS*(Microwave or UHF) 1 to 4 Yes Yes Yes

Medium-Frequency EPS 3 to 10 No Yes Yes

Low-Frequency EPS (Loran) 50 to 2000 No No Yes

Satellite Positioning:DopplerSTARFIX

100 to 3005

NoNo

NoYes

NoYes

NAVSTAR GPS:**Absolute Point Positioning (No SA)Absolute Point Positioning (w/SA)Differential Pseudo RangingDifferential Kinematic (future)

1550 to 100

2 to 50.1 to 1.0

NoNoYesYes

NoNoYesYes

YesYesYesYes

* Electronic Positioning System** Global Positioning System

(From EM 1110-2-1003)

(4) Bathymetric surveys are required for many studiesof geology and geomorphology in coastal waters. Echosounders are most often used to measure water depthsoffshore. Errors in acoustic depth determination arecaused by several factors:

(a) Velocity of sound in water. The velocity in near-surface water is about 1,500 m/sec but varies with waterdensity, which is a function of temperature, depth, andsalinity. For high- precision surveys, the acoustic velocityshould be measured onsite.

(b) Boat-specific corrections. As the survey pro-gresses, the vessel’s draft changes as fuel and water areused. Depth checks should be performed several timesper day to calibrate the echo sounders.

(c) Survey vessel location with respect to knowndatums. An echo sounder on a boat simply measures thedepth of the water as the boat moves over the seafloor.However, the boat is a platform that moves verticallydepending on oceanographic conditions such as tides andsurges. To obtain water depths that are referenced to a

5-26

EM 1110-2-181031 Jan 95

Table 5-6Summary of Acoustic Survey Systems

Acoustic System Frequency (kHz) Purpose

Sea floor and water column

Echosounder 12 - 80 Measure water depth for bathymetric mapping

Water column bubble detector(tuned transducer)

3 - 12 Detect bubble clusters, fish, flora, debris in water column

Side-scan sonar 38 - 250 Map sea floor topography, texture, outcrops, man-made debris,structures

Sub-bottom profilers

Tuned transducers 3.5 - 7.0 High resolution sub-bottom penetration

Electromechanical:

Acoustipulse® 0.8 - 5.0 Bottom penetration to ∼30 m

Uniboom® 0.4 - 14 15 - 30 cm resolution with 30 - 60 m penetration

Bubble Pulser ∼ 0.4 Similar to Uniboom®

Sparker:

Standard 50 - 5,000 Hz Use in salt water (minimum 20 ‰), penetration to 1,000 m

Optically stacked (same) Improved horizontal resolution

Fast-firing4 KJ & 10 KJ

(same) Improved horizontal and vertical resolution

De-bubbled,de-reverberated

(same) Superior resolution, gas-charged sediment detection

Multichanneldigital

(same) Computer processing to improve resolution, reduce noise

(From Sieck and Self (1977), EG&G®, Datasonics®, and other literature)

known datum, echo sounder data must be adjusted in oneof two ways. First, tides can be measured at a nearbystation and the echo sounder data adjusted accordingly.Second, the vertical position of the boat can be constantlysurveyed with respect to a known land datum and theseresults added to the water depths. For a class 1 survey,either method of data correction requires meticulous attention to quality control.

(d) Waves. As the survey boat pitches up and down,the seafloor is recorded as a wavey surface. To obtainthe true seafloor for the highest quality surveys, trans-ducers and receivers are now installed on heave-compen-sating mounts. These allow the boat to move verticallywhile the instruments remain fixed. The most commonmeans of removing the wave signal is by processing thedata after the survey. Both methods are effective,although some contractors claim one method is superior tothe other.

Even with the best efforts at equipment calibration anddata processing, the maximum practicable achievableaccuracy for nearshore depth surveys using echosoundersis about ± 0.15 m (EM 1110-2-1003). The evaluation ofthese errors in volumetric calculations is discussed in Sec-tion 5-5. Survey lines are typically run parallel to oneanother, with spacing depending on the survey’s purposeand the scale of the features to be examined.

(5) In geophysical surveys, the distance between thesound source and reflector is computed as velocity ofsound in that medium (rock, sediment, or water) dividedby one half of the two-way travel time. Thismeasurement is converted to an equivalent depth andrecorded digitally or on a strip chart.

(6) The principles of subbottom seismic profiling arefundamentally the same as those of acoustic depth sound-ing. Subbottom seismic devices employ a lower

5-27

EM 1110-2-181031 Jan 95

frequency, higher power signal to penetrate the seafloor(Figure 5-11). Transmission of the waves through earthmaterials depends upon the earth material properties, suchas density and composition. The signal is reflected frominterfaces between sediment layers of different acousticalimpedance (Sheriff 1980). Coarse sand and gravel, gla-cial till and highly organic sediments are often difficult topenetrate with conventional subbottom profilers, resultingin poor records with data gaps. Digital signal processingof multi-channel data can sometimes provide useful datadespite poor signal penetration. Spacing and griddimensions again depend upon the nature of the investiga-tion and the desired resolution.

(7) Acoustic characteristics are usually related tolithology so that seismic reflection profiles can be consid-ered roughly analogous to a geological cross section ofthe subbottom material. However, because of subtlechanges in acoustic impedence, reflections can appear onthe record where there are minor differences in the lithol-ogy of underlying and overlying material. Also,

significant lithologic differences may go unrecorded dueto similarity of acoustic impedance between boundingunits, minimal thickness of the units, or masking by gas(Sheriff 1980). Because of this, seismic stratigraphyshould always be considered tentative until supported bydirect lithologic evidence from core samples. Signalprocessing procedures are being developed by the USACEto analyze waveform characteristics of outgoing andreflected pulses. With appropriate field checks, the sea-floor sediment type and hardness can be modeled, reduc-ing the need for extensive coring at a project site.

In shallow coastal areas, it is common practice to use jetprobing to accompany subbottom seismic surveys. This isespecially important when there is a thin veneer of sandover more resistant substrate.

(8) The two most important parameters of a sub-bottom seismic reflection system are its vertical resolu-tion, or the ability to differentiate closely spacedreflectors, and penetration. As the dominant frequency of

Figure 5-11. Principles of obtaining subbottom seismic data

5-28

EM 1110-2-181031 Jan 95

the output signal increases, the resolution becomes finer.Unfortunately, raising the frequency of the acoustic pulsesincreases attenuation of the signal and consequentlydecreases the effective penetration. Thus, it is a commonpractice to use two seismic reflection systems simulta-neously during a survey; one having high resolution capa-bilities and the other capable of greater penetration.

(9) Side-scan sonar is used to distinguish topographyof the seafloor. Acoustic signals from a source towedbelow the water surface are directed at a low angle toeither or both sides of a trackline, in contrast with thedownward-directed Fathometer and seismic reflectionsignals (Figure 5-12). The resulting image of the bottomis similar to a continuous aerial photograph. Detailedinformation such as spacing and orientation of bed formsand broad differences of seafloor sediments, as well asfeatures such as rock outcrops, boulders, bed forms, andman-made objects, can be distinguished on side-scan. Itis generally recommended that bathymetry be run in con-junction with side-scan to aid in identifying objects with

subtle vertical relief. The side-scan system is sensitive tovessel motion and is most suitable for use during calmconditions.

(10) Commonly available side-scan sonar equipment,at a frequency of 100 khz, is capable of surveying theseafloor to over 500 m to either side of the vessel track-line; thus, a total swath of 1 km or more can be coveredat each pass. To provide higher resolution output at closerange, some systems are capable of dual operation usingboth 500-khz and 100-khz frequency signals. The dataare simultaneously recorded on separate channels of afour-channel recorder. Digital side-scan sonar systems areavailable that perform signal processing to correct forslant range to seafloor targets and correct for surveyvessel speed. The resulting records show the true x-ylocation of seafloor objects, analagous to maps or aerialphotographs. The digital data can be recorded onmagnetic media, allowing additional signal processing orreproduction at a later date.

Figure 5-12. Side-scan sonar in operation

5-29

EM 1110-2-181031 Jan 95

(11) The identification of potential sand borrow sitesfor beach renourishment has become an increasinglyimportant economic and environmental issue in recentyears. Reconnaissance surveys to identify potential sitesare made using high-resolution seismic profilers, side-scansonar, and echo sounders. Suggested survey proceduresare discussed in Appendix I.

(12) Ground-penetrating radar (GPR) is a relativelynew technique for subsurface exploration. In contrast tothe acoustic systems described above, GPR is used sub-aerially. The radio portion of the electromagnetic spec-trum is emitted from the source and reflected back to thesensors. The transparency of geologic materials varies.Sands and limestones are typically reasonably transparent.The use of GPR in marine environments is limitedbecause salt water is non-transparent to electromagneticradiation in the radio frequencies. Fitzgerald et al. (1992)used GPR as a tool to study beach ridge barriers inBuzzards Bay, Massachusetts. GPR has been very usefulin the Great Lakes to detect buried channels and tilloutcrops.

l. Morphologic and bathymetric profiles.

(1) Periodic topographic and nearshore bathymetricsurveys constitute the most direct and accurate means ofassessing geologic and geomorphic changes over moderntime scales. Time series data, such as repeated beachprofiles, allow the assessment of erosion and accretion inthe coastal zone. The preferred surveying techniqueinvolves collecting a series of shore-normal profile lines.These must extend landward of the zone that can be inun-dated by storms, usually behind the frontal dunes. Thelines should extend seaward deep enough to include theportion of the shoreface where most sediment moves (i.e.,to beyond closure, as defined in Chapter 4).

(2) Permanent or semi-permanent benchmarks arerequired for reoccupying profile sites over successivemonths or years. On rapidly transgressing coasts, thesebenchmarks should be located at the landward end of theprofile line in order to minimize their likelihood of beingdamaged in storms. The locations of survey monumentsmust be carefully documented and referenced to othersurvey markers or control points. The ability to accu-rately reestablish a survey monument is very importantbecause it ensures that profile data collected over manyyears will be comparable (Hemsley 1981). Locationswhich might experience dune burial should be avoided,and care should also be taken to reduce the visibility ofbenchmarks to minimize damage by vandals.

(3) Both the frequency of the sampling and the over-all duration of the project must be considered when plan-ning a beach profiling study. Morphologic changes ofbeaches can occur over varying time scales, and if long-term studies are to be conducted, the dynamic nature ofthe beach should be taken into account. Often, it is finan-cially or logistically impractical to conduct frequent,repeated surveys for a sufficient length of time to obtainreliable and comprehensive information on long-termprocesses at the study area. Nonetheless, resurveying ofprofile lines over a period of more than one year can beof substantial help in understanding the prevailing sea-sonal changes. Resurveying of control profile lines atselected time intervals can reveal seasonal patterns. Inaddition, special surveys can be made after significantstorms to determine their effects and measure the rate ofrecovery of the local beach system. At a minimum, sum-mer and winter profiles are recommended. Unfortunately,there are no definitive guidelines for the timing andspacing of profile lines. Table 5-7 outlines a suggestedsurvey schedule for monitoring beach fill projects. Insummary, observation over a period of time is recom-mended in order to document the range of variability ofmorphology and bathymetry.

(4) Some issues concerning the spatial aspects ofstudy include the spacing of profiles, longshore dimen-sions, and cross-shore dimensions. Profile lines should bespaced at close enough intervals to show any significantchanges in lateral continuity. In a cross-shore direction,the uppermost and lowermost limits of the profiles shouldbe located where change is unlikely to occur, and shouldadequately cover the most active zones such as the shoreand upper shoreface. The preferred closure depth is at thetoe of the shoreface, although a selected depth contourwhere variability becomes minimal is acceptable. Histori-cal shorelines are an important component of where theseuppermost and lowermost limits are located, particularlyalong rapidly changing coastlines. For example, shoreand dune deposits that are now inland from the modernshoreline are likely to be affected by marine or lacustrineprocesses only during large storms. Large-scale aerialphotographs or maps of these interior areas are usuallyadequate for examining these more stable features.Appropriate longshore dimensions of the survey griddepend upon the nature of the problem. Profile linesshould be connected with a shore-parallel survey to deter-mine positions and elevations of each profile relative toone another.

5-30

EM 1110-2-181031 Jan 95

Table 5-7Example of Beach Fill Area Profile Survey Scheme

Year Times/Year Number of Profiles

pre-fill 2 Collect within fill area and at control locations in summer and winter months to characterizeseasonal profile envelope (beach & offshore).

post-fill 1 Collect all profiles immediately after fill placement at each site (beach & offshore) to documentfill volume. Collect control profiles immediately after project is completed.

1 4 Four quarterly survey trips collecting all beach and offshore profiles out to depth of closure.Begin series during the quarter following the post-fill survey.

Continue year 1 schedule to time of renourishment (usually 4-6 years). If project is a single nourishment, taper surveys in subsequentyears:

2 2 6- and 12-month survey of all beach and offshore profiles

3 2 6- and 12-month survey of all beach and offshore profiles.

4 1 12-month survey of beach and offshore profiles.

Note:If project is renourished, repeat survey schedule from post-fill immediately after each renourishment to document new fillquantity and behavior.Project-specific morphology and process requirements may modify this scheme.Monitoring fill after major storms is highly desirable to assess fill behavior and storm protection ability. Include both pro-file and sediment sampling. Conduct less than one week after storm conditions abate to document the beach and off-shore response.

(5) Onshore (beach) profiles.

(a) Onshore portions of profiles are surveyed usingstandard land survey techniques and equipment. Equip-ment commonly used in surveys includes transits, levels,or theodolites, which are used for siting survey rods.Detailed information concerning techniques and equip-ment can be found in textbooks (i.e. Brinker and Wolf(1984)).

(b) Surveys are preferably conducted during low tide,when the profile line can be extended as far seaward aspossible. A typical cross-shore profile survey can consistof around 25 to 50 points over a total length of 600 to1,000 m. Data point spacing is variable, with more pointstaken over areas with complex elevation change such as aberm scarp or the nearshore bar trough and crest. Meas-urements are usually made every 5 to 10 m along thesubaerial profile, or at a shorter interval to define majormorphologic features. Standard procedure places thesurvey instrument at the baseline and proceeds seaward.

(6) Extending profile lines offshore beyond wadingdepths requires boats or amphibious vehicles. Amphibi-ous vehicles are better-suited to this task because they cantraverse the sea-land boundary and maintain the continuityof profile lines. Acoustic echo sounders can be used forcontinuous profiling seaward of the breaker zone, but thesignals are usually disrupted by breaking waves, and boats

suitable for offshore use cannot approach the shore closeenough to connect directly with a land profile. High-precision electronic navigation is recommended if thesurveys extend offshore more than a few hundred meters.

(7) Sea sleds.

(a) During calm weather conditions, sea sleds havebeen successfully used to obtain shoreface profiles closeto shore. A sea sled consists of a long, upright stadia rodmounted vertically on a base frame with sledlike runners(Clausner, Birkemeier, and Clark 1986) or a sled-mountedmast with a prism for use by total station survey system(Fredette et al. 1990). The sled is towed, winched, orotherwise propelled along the profile lines while frequentdepth and position data are determined using onshore in-struments. Because the sea sled does not float, elevationsare not subject to wave or tide variations, thus providing amore accurate comparison between repeated surveys. Atpresent, it is not possible to obtain bottom samples with asea sled; these must be obtained from a boat or amphibi-ous vehicle working in conjunction with the sled. Sledsare currently limited to use within 4 km of the coast andwater depths of 12 m, less than the height of the sledmasts. A limitation of sleds is that they normally must beused at sites with road access to the beach. It is verydifficult to use them if the shore is revetted or armored.Also, sleds cannot be used if the offshore topography isrough (i.e., till or coral outcrops, glacial boulders).

5-31

EM 1110-2-181031 Jan 95

(b) When conducting a sled survey, the tow boat isnavigated based upon a continuous report of the sled’scoordinates transmitted from the shore station. The sledshould be kept to within 2 to 3 m of the shore-normalprofile line 95 percent of the time. Measurements of thesled position are usually read at approximately 10-mintervals along the profile line close to shore to resolvebar/trough features, and increased to 15- to 20-m intervalsfurther offshore (Birkemeier et al. 1985, Stauble et al.1993). The positioning measurements are automaticallyrecorded by a data logger and copied to a computer forprocessing or editing at the end of each survey day.

(8) A helicopter bathymetric surveying system hasbeen in use at USAE District, Portland, since the 1960’s.The big advantage of this procedure is that land-accuracysurveys can be conducted offshore in high waves and nearstructures, conditions under which a boat could not per-form (Pollock 1995). A helicopter is fitted with aweighted, calibrated cable and prisms. A total stationsurvey system is set up onshore to measure the location ofthe cable. Soundings are commonly taken at 8-m inter-vals along profile lines up to 2,500 m offshore.Operations are limited by poor visibility or winds over15-20 m/sec (30-40 knots).

(9) The Coastal Research Amphibious Buggy(CRAB), a self-propelled vehicle, was developed to makecontinuous onshore-offshore profiles and obtain bottomsamples. The CRAB is a tripod mounted on wheels andis propelled by hydraulic motors. It can move under itsown power across the beach and shoreface to a depth ofabout 8 m. It has been widely used at the CERC FieldResearch Facility at Duck, North Carolina. Both theCRAB and sea sled are important tools for characterizingsubmarine bars and the overall morphology of offshoreprofiles (Stauble 1992).

m. Prototype monitoring.

Prototype testing and monitoring involve bringing togethermultiple means of investigating and measuring the pro-cesses and responses of a coastal site. Prototype studiesoften involve physical experiments, conducted under idealor well-monitored conditions in the field. The purpose ofmany prototype studies is to test and evaluate theoreticalformulae or conceptual asssumptions. Prototype studies,in other instances, are conducted to assess the status andvariations of environmental conditions at a site and todevelop information for guidance in construction ofstructures.

5-4. Laboratory Techniques and Approaches

a. Laboratory observation and experiment. Thecharacteristics of samples obtained in the field can befurther analyzed in the laboratory. Some properties thatare commonly examined include: (1) sediment properties,such as grain size, shape, and density, mineralogy, andheavy mineral type and content; (2) stratigraphicproperties, which can be characterized using core descrip-tion, preservation, and analysis techniques; and (3) geo-chronological history, obtained from radiometric datingand a variety of relative dating approaches. In order toachieve maximum benefit from laboratory analyses, thecoastal scientist must be cognizant of the limitations andvariance of precision and accuracy of each test andprocedure.

(1) Laboratory analysis of sediment.

(a) Sediments can be classifed into size rangeclasses. Ranked from largest to smallest, these includeboulders, cobbles, gravel, sand, silt, and clay (Table 5-8).Particle size is often expressed as D, or the diameter inmillimeters, and sometimes includes a subscript, such asD84, to indicate the diameter corresponding to the listedpercentile. As an alternative, grain size is often expressedin phi (φ) units, whereφ = -log2 D (Hobson 1979). Thisprocedure normalizes the grain size distribution andallows computation of other size statistics based on thenormal distribution.

(b) Grain-size analysis involves a series of proce-dures to determine the distribution of sediment sizes in agiven sample. An important aspect of the laboratoryanalysis program, which must be designed into the fieldsampling scheme, is to obtain sufficient sediment to ade-quately determine the sediment population characteristics(Table 5-9). Large samples should be divided using asample splitter to prevent clogging of sieves. Particleaggregates, especially those in the silt-clay range whichshow cohesive properties, should be separated and dis-persed by gentle grinding and use of a chemical disper-sant (sodium hexametaphospate) before analysis. Notethat depending on the purpose of the study, it may beimportant to preserve the hydraulic characteristics of sedi-ment aggregates (i.e., clay balls, cemented sand, or shellfragments). In these circumstances, it is best to not splitor mechanically grind the samples.

5-32

EM 1110-2-181031 Jan 95

5-33

EM 1110-2-181031 Jan 95

Table 5-9Minimum Weight of Sample Required for Sieving

Maximum Particle Size Present in WeightSubstantial Proportion (> 10%) of Sample

in. mm kg2.5 64 502.0 50 351.5 40 151.0 25 50.75 20 20.50 12.5 10.38 10 0.50.25 6.3 0.2

2.4 0.1

British Standards Institution (1975). Note: quantities specified inASTM Standard D2487-92 are similar.

(c) Laboratory techniques used to estimate sedimentdiameter depend in part on the grain size. Pebbles andcoarser sediments can be directly measured with calipersor by coarse sieves. The grain-size distribution of sand isdetermined directly by sieve analysis, sedimentation tubes,or Coulter counter. Silt and clay-sized material is deter-mined indirectly by hydrometer or pipette analysis, or theuse of a Coulter counter. The size distribution of mixedsediments is determined by using a combination of sieveand hydrometer or pipette analyses. Practical proceduresfor conducting laboratory grain size and mineralogicaltests on sediment samples are covered by Folk (1980) andLewis (1984). Laboratory manuals more oriented towardsengineering applications include EM 1110-2-1906 andthose produced by the American Society for Testing andMaterials (1964) and Bowles (1986).

(d) Coastal sediments reflect the relative importanceof various source areas, and transport processes. Somesources of coastal sediments include river basins thatempty into the coastal zone, nearshore cliffs and uplandsthat are denuded by waves, wind, transported materialmass wasting and slope wash, and sediments transportedby longshore currents. Because gravel and larger particlesrequire more energy to be transported, they are typicallyfound close to their source. In contrast, silt and clay maybe transported long distances. The size fraction distribu-tion is determined by the composition of the source rocksand weathering conditions. The mineralogy of sediments,especially clays, shows that variations are controlled bysource rocks and weathering conditions. Resistant miner-als, such as quartz and feldspars, comprise most coastaldeposits (Table 3-2). However, as tracers, the least com-mon minerals are generally the best indicators of source.

(e) Heavy minerals can provide information regard-ing source and process and other aspects of geomorphicvariability in the coastal zone (Brenninkmeyer 1978;Judge 1970; McMaster 1960; Neiheisel 1962). Pro-nounced seasonal variations in heavy minerals may occurin beach and nearshore samples. Lag deposits of heavyminerals are often seen on the beach after storms.

(f) Analysis of size and texture can also be used todistinguish among sediments that may have come fromthe same original source area. As an example, Mason andFolk (1958) used size analysis to differentiate dune andbeach sediments on Mustang Island, Texas.

(g) A variety of techniques are used to identify themineralogy of coastal sediments. Mineralogy of coarsesediments and rocks is typically assessed using laboratorymicroscopes. Clay mineralogy is usually assessed withX-ray diffraction methods or electron microscopy. Heavyminerals are separated from light minerals using bromo-form (specific gravity of 2.87) after washing and sieving.In unconsolidated sediments, heavy mineral samples areexamined under a microscope to determine approxima-tions or percentages of mineral types.

(2) Core description and analysis.

(a) Core description is widely used to characterizethe features and depositional environments of sediments.After being collected in the field, core barrels are sealedto retain moisture. In the laboratory, they are cut in halflengthwise. One side of the core is used for descriptionand the other for radiography, peels, and subsampling forgrain size analysis, palynology, and organic materials.Cores are often photographed soon after splitting, whilethe exposed surfaces are still fresh.

(b) A hypothetical USACE core drilling log isshown in Figure 5-13 completed in the level of detailnecessary for coastal geologic studies. An alternate sheetused at some universities is shown in Figure 5-14.Important characteristics of the sedimentary sequence thatneed to be described include grain size variations, sedi-mentary structures and directions, and occurrences ofcyclic bedding, such as varves. Evidence of plant rootsand features such as color changes, mottling, discontinu-ities, and other variations in physical characteristics maybe indicators of key changes. Roots, for example, oftencorrespond to marshes in coastal sequences. Fossils andpollen in stratigraphic sequences are indicative of paleoen-vironmental characteristics and changes. Techniques for

5-34

EM 1110-2-181031 Jan 95

Figure 5-13. Example of hypothetical USACE drilling log. Descriptions and notations by field geologist should besufficiently complete to allow an interpretation of depositional environment and other factors indicating paleoenvi-ronmental characteristics

5-35

EM 1110-2-181031 Jan 95

Figure 5-14. Example core description form used for sedimentary environments (courtesy of Dr. Harry Roberts,Louisiana State University)

5-36

EM 1110-2-181031 Jan 95

analysis and interpretation of such evidence can be foundin Faegri and Iverson (1975) and Kapp (1969).

(c) Grain size variation in cores can yield much infor-mation about the sedimentary environments and thus thegeologic history of the region. Coarser fractions settlefirst, followed by silts and clays. This separation is afunction of particle settling velocities, which vary depend-ing upon particle size, density, shape, and the nature ofthe transport media. Changes in the environment of depo-sition can result in the clay fraction being separated fromgranular material both spatially and temporally. Forexample, silt and clay are usually deposited further fromshore than granular material.

(d) X-ray radiography is an imaging method thatamplifies contrasts in grain size, mineralogicalcomposition, packing density, water content, diageneticproducts, sedimentary structures and geochemical inclu-sions in cores that otherwise appear homogeneous(Roberts 1981). Being able to distinguish these featuresmay assist in understanding the sequence of geomorphicchanges that occurred at that site. For example, the scaleand direction of bed forms can be used to estimate paleo-currents. Marker horizons are related to a date or a sig-nificant event. Peat indicates stability and growth at ornear sea level. Radiography is based on the differentialtransmission of X-ray radiation through a sample ontosensitized X-ray photographic film. Variations in textureas well as chemical composition throughout the sedimentresult in differential attenuation of the incident X-rayradiation before it reaches the underlying film. Samplesof uniform thickness (about 1 cm) that are cut lengthwisewith a wire knife provide the best results in radiography(Roberts 1981).

(e) The occurrence of paleosols in cores may alsoprovide important information toward assessing the geo-logic history of coasts. In terrestrial coastal environments,there may be prolonged periods of minimal sedimentationduring which soil development may occur, followed byperiods of relatively rapid sedimentation without soildevelopment.1 This scenario is characteristic of recentsea level changes during the Quaternary. As alternativescenarios, such cycles could occur in a semi-protected saltmarsh subject to sedimentation during a severe storm or

_____________________________1 The term “soil” in this context refers to unconsolidatedsurficial sediment which supports plant life. This is amore restrictive definition than the one typically used inengineering texts, which refers to soil as any unconsoli-dated material, even if barren of plant life.

in a soil which subsided as a result of rapid burial byother sediments. As with modern soils, horizon color andhorizon assemblages based on color permit an initialidentification. Important paleosols, which may reflectonly limited pedogenesis, are represented only by thin,dark, organic horizons. Less apparent chemical and phys-ical changes in sediments which were exposed to atmo-spheric and meteorological processes may also occur.Soils that are uniform over a wide area can sometimes beused as approximate marker horizons and thus are valu-able for relative dating purposes. In some circumstances,soils may also contain enough organic material to besuitable for radiocarbon dating.

(3) Geochronology. Geochronology is the study oftime in relationship to the history of the earth. Geochron-ology encompasses a variety of radiometric and non-radiometric techniques, which collectively can datematerials whose ages extend from near-present throughthe Pleistocene and earlier. Radiometric techniques varyin precision, in time range, in the types of materials thatcan be analyzed, and the type of information that resultsare capable of providing. Non-radiometric techniques thatmay be useful in coastal areas include archives, archeol-ogy, dendrochronology, thermoluminescence dating, mag-netostratigraphy or paleomagnetic dating, paleoecology,the use of weathering and coating indices. Use of mul-tiple techniques typically provides the best results forassessing the geologic history of coasts.

(4) Radiometric dating and isotopes.

(a) Radiometric dating techniques have been usedsince the 1950’s. Many natural elements are a mixture ofseveral isotopes, which have the same chemical propertiesand atomic numbers but different numbers of neutronsand hence atomic masses. Radiometric methods of datingare based on radioactive decay of unstable isotopes. Theduration of time leading to the state where half the origi-nal concentration remains is known as the half-life. Ingeneral, the useful dating range of individual isotopicmethods is about ten times their half-life. The radio-metric isotopes Carbon-14, Potassium-Argon 40, Caesium-137, Lead-210, and Thorium-230 are the most commonlyused in standard geologic investigations (Faure 1977;Friedlander, Kennedy, and Miller 1955).

(b) Radiocarbon (Carbon-14 or14C) dating is per-haps the most widely used technique for assessing the ageof Holocene and late Pleistocene organic materials. Oncean organism or plant dies, its radiocarbon (14C) content isno longer replenished and begins to decrease exponenti-ally, achieving a half-life after some 5,730 years.

5-37

EM 1110-2-181031 Jan 95

Substances that are often examined with14C datinginclude wood, charcoal, peat, shells, bones, aqueous car-bonates, rope, and soil organics. Recent developmentsusing mass spectrometers allow detection of absoluteamounts of14C content in samples as small as 5 mg. Tobe comparable, radiocarbon dates are adjusted to a zeroage at AD 1950. Analytical error factors are given as oneor two standard deviations about the mean. Other errors,associated with sample contamination, changes in atmo-spheric or oceanic14C content, and fractionation, are moredifficult to estimate. Absolute dates of samples less than150 years old or greater than 50,000 years old are current-ly considered to be ambiguous.

(c) Potassium:argon dating (Potassium-Argon-40or K:Ar) can be applied to a wide range of intrusive andextrusive igneous rocks that contain suitable minerals. Inaddition to constraints on rock type, it is necessary for thesample to be unaltered by weathering or other geologicalprocesses that may allow diffusion of radiogenic argonfrom the sample. The occurrence of such rocks alongcoasts is generally restricted to regions adjacent to plateboundaries and regions of active tectonics. Potassium-argon dating of Holocene deposits is generally imprecise,with errors of ± 15 to 30 percent. Only certain minerals,particularly those with a high K and low atmospheric Arcontent, are suitable for extending the K:Ar dates into thelate Pleistocene. For these reasons, it has limited applica-tions in studies of the geologic history of coasts.

(d) Fission-track dating was developed as a comple-mentary technique to potassium:argon (K:Ar) dating.Most applications to Quaternary deposits have involveddating airfall volcanic ash or glass deposits, a field knownas tephrochronology. This material usually has widedistribution and geologically speaking has infinitely nar-row depositional time duration. However, it is oftenabsent or quickly removed in many coastal settings. Ifpresent, the rapid deposition and large aerial extent of ashmake it an excellent tool for correlation of rock strata,which can provide radiometric age dates. A listing ofsome of the important volcanic ash layers in North Amer-ica, which include very recent to Pleistocene dates, can befound in Sarna-Wojcicki, Champion, and Davis (1983).

(e) Cesium-137 (137Cs) is an artificial isotope, primar-ily produced during the atmospheric testing of nuclearweapons. These tests began in the 1940’s, peaked in theearly 1960’s, and have declined since the advent ofnuclear test ban treaties (Wise 1980).137Cs is stronglyabsorbed onto sediment or soil and has been used in stud-ies of soil erosion and sediment accumulation in wetlands,

lakes, and floodplains. The timing of very recent events(post-1954) and human impacts on coastal ecosystems canbe improved using such techniques.

(f) Lead-210 (210Pb) is an unstable, naturally occur-ring isotope with a half-life of just over 22 years and adating range of 100 to 200 years (Oldfield and Appleby1984, Wise 1980). It forms as part of a decay chain fromRadium-226, which escapes into the atmosphere as theinert gas Radon-222. The excess or unsupported210Pbreturns to the earth as rainfall or dry fallout, and can beseparated from that produced by in situ decay. Applica-tions in coastal environments are limited but show goodpotential. This technique would be of greatest value inlow-energy environments and would allow documentationof the timing of recent events and human impacts oncoastal ecosystems.

(g) Thorium-230/Uranium-234 (230Th/234U), a usefuldating technique that complements other methods, isapplicable for dating coral sediments. The techniqueinvolves comparing the relative amounts of the radioactiveisotope of thorium,230Th, with that of uranium,234U.Thorium-230 increases in coral carbonate from zero at thedeath of the organism to an equilibrium with Uranium-234 at 0.5 million years, allowing samples as old asmiddle Pleistocene to be dated.

(5) Non-radiometric methods of dating and relativedating.

(a) Archival and archeological documentation canassist in understanding the geologic history of coasts.Historical and social documents may contain detaileddescriptions of major storms, of ice movements, of shore-line changes, and of other catastrophic events. Historicalrecords are most useful if they correspond to a particulardate or specified range of time, as do newspaper reports.Archeological evidence can provide important clues forassessing Holocene environmental changes. Pottery, stonetools, coins, and other artifacts can be assigned ages andthus may be of assistance in dating surface and subsurfacedeposits. If discovered in a stratigraphic sequence, cul-tural artifacts provide a minimum age for deposits beneathand maximum age for deposits above. Archeologicalevidence, such as buried middens, inland ports, or sub-merged buildings, may also indicate shoreline changes andsometimes can be used to estimate rates of deposition incoastal areas. For example, the Holocene MississippiRiver deltaic chronology was revised using artifacts asindicators of the age of the deltaic surfaces (McIntire1958).

5-38

EM 1110-2-181031 Jan 95

(b) Thermoluminescence (TL), a technique that iscommonly practiced in archeology for dating pottery, hasbeen extended for use in geological studies. It has beenused for dating a variety of Pleistocene sediments, includ-ing loess. For geological purposes, TL needs furtherrefinement because most results to date are considered inerror, generally being too young. It does, however, gen-erally provide a good estimate of stratigraphic order.Thermoluminescence dating has the best potential whereclay-fired artifacts are present and has promise for datinga variety of deposits of Quaternary age.

(c) Magnetostratigraphy or paleomagnetic dating is ageochronologic technique that is used in conjunction withcorrelations of regional radiometric dates andpaleomagnetic characteristics. Because the earth’s mag-netic field changes constantly, the magnetic characteristicsof rock and sediments can be used to determine an agefor materials. The most dramatic changes are reversals,in which the earth’s polarity switches from the north tothe south pole. The reversals are relatively infrequentoccurrences, with the most recent one occurring700,000 years ago. Less dramatic secular variations ofthe geomagnetic field, however, can also be important inhelping to provide a time scale useful for dating overhundreds or thousands of years by linking magnetic prop-erties with time scales established by radiometric tech-niques. The combination of declination (the anglebetween true and magnetic north), inclination (the dip ofthe earth’s magnetic field), and magnetic intensity pro-duces a characteristic paleomagnetic signature for a partic-ular location and time. The magnetic alignments can beincorporated and preserved in baked materials, in sedi-ment particles which settle out in standing water, and incooled magma. The technique is most-suited to lakesediments containing homogeneous particle sizes andorganics. This technique can be used in places where themagnetostratigraphy has been linked with radiometricdates and can be extended to over 200 million yearsbefore present.

(d) Dendrochronology or tree ring dating can provideprecise data regarding minimum age of a geomorphicsurface. It can also provide proxy data concerning envi-ronmental stresses, including climatic conditions such ascold temperatures and droughts. In some parts of theworld, overlapping sets of rings on trees have been usedto construct a comprehensive environmental history of theregion.

(e) Lichenometry is the study of the establishmentand development of lichen to determine a relative chro-nology (Worsley 1981). Although used most extensively

for studies of glacier fluctuations, this technique also hasapplication in shoreline dating. The method involves themeasurement of thallus size, with increasing diameterrepresenting increasing age. It is valid from about tenyears to a few centuries before present. This measure-ment is often conducted in the field with a ruler or withcalipers. Field techniques differ, although normally thelargest diameters are measured. Although there has been alack of critical assessment of the technique, the majorityof research shows that the technique gives reasonabledates when applied to a variety of environments.

(f) Paleoecology is the study of fossil organisms inorder to reconstruct past environments. Pollen analysis,or palynology, is the single most important branch ofpaleoecology for the late Pleistocene and Holocene. Usesof paleoecological tools include: (a) the establishment ofrelative chronologies and indirect dating by means ofcorrelation with other dated sequences; (b) characteriza-tion of depositional environments at or near the samplingsite, since certain species and combinations of species areadapted to certain conditions; (c) reconstruction of thepaleoenvironmental and paleoclimatic conditions;(d) establishment of human-induced transformations of thevegetation and land use regime (Oldfield 1981).

(g) The use of weathering and coating indices forrelative age dating in geomorphology is rapidly increas-ing. Using laboratory microscopes, samples are calibratedwith those of known age and similar chemistry for eachgeographic area. One such method, obsidian hydrationdating, is based on the reaction of the surface of obsidianwith water from the air or soil, which produces a rindwhose thickness increases with time (Pierce, Obradovich,and Friedman 1976). Rock varnish-cation ratio dating isused primarily in deserts, where rocks develop a coating(Dorn 1983). Emery (1941) used dated graffiti to deter-mine the rates of erosion and weathering in sandstonecliffs.

(h) Varve chronology may be useful in quiescent orlow-energy basins where thin laminae of clay and silt aredeposited. In glaciated coastal areas, the thin layers orvarves are usually annual deposits. The sequences ofsuccessive graded layers can be discerned visually. Colorvariations occur because usually the winter season depos-its have a higher organic material content. The result isalternating light-colored, gray-brown sediment layers anddark-colored organic layers.1 Varve chronology rarelyextends beyond about 7,000 years._____________________________1 In freshwater lakes, varves are caused by clay-siltdeposition cycles. The silt settles out in spring and sum-mer, and the clay in fall and winter.

5-39

EM 1110-2-181031 Jan 95

(i) A major limitation of varve chronology is the factthat in the marine environment, annual varves are usuallyonly preserved in anoxic basins, where a lack of oxygencauses a dearth of bottom-dwelling animals. Otherwise,mollusks, worms, fish, and crustaceans thoroughly reworkthe seafloor. This reworking, known as bioturbation,thoroughly destroys near-surface microstructure in most ofthe shallow-water portions of the world’s oceans. Exam-ples of anoxic basins include portions of the Black Sea(Anderson, Lyons, and Cowie 1994) and Saanich Inlet inBritish Columbia. The latter receives an annual input ofclays from the Fraser River. Yearly variations in thedischarge of the Fraser River’s spring freshet cause chan-ges in varve thicknesses.

b. Physical models.

(1) The use of physical models can be invaluable inunderstanding how geomorphic variability occurs incoastal areas. Physical modelling provides an opportunityfor reducing the complexity of natural systems, for scalingdown dimensions, and for accelerating change over timeso that detailed interactions can be identified. Physicalmodels can be applied in studies of hydrodynamics, sedi-ments, and structures. In studies of coastal processes andresponses, the wave tank is both the simplest and themost utilized physical model.

(2) Physical models are typically either two- or three-dimensional. A wave tank is considered to be a two-dimensional model because changes over length and overdepth can be examined. Where variations over width arealso investigated, the model is considered to be three-dimensional. A three-dimensional model or basin mayhave a variety of types of bottoms, including beds that arefixed, fixed with tracers, or moveable. Physical modelsrequire precise scaling and calibration, and much designand construction expertise must be devoted to their initialconstruction. Once set up, however, they allow for directmeasurement of process elements, repeated experimentsover a variety of conditions, and the study and isolation ofvariables that are difficult to assess in the field.

(3) Some examples of physical model experiments(conducted principally in wave tanks) that helped eluci-date geomorphologic variability of coasts include studiesof littoral drift blockage by jetties (Seabergh and McCoy1982), breaker type classification (Galvin 1968), experi-ments of cliff erosion (Sunamura 1983), relationships ofstorm surge or short-term water level changes to beachand dune erosion, and studies of suspended sedimentconcentration under waves (Hughes 1988).

(4) Large-scale physical models of harbors, rivers,and estuaries have been built and tested at the WaterwaysExperiment Station in order to examine the effects ofjetties, weirs, channel relocations, and harbor constructionon hydrodynamics and shoreline changes in these complexsystems. Measurements made by gauges at prototype (i.e.field) sites have sometimes been used to help calibrate thephysical models. In turn, the results of tests run in thephysical models have identified locations where gaugesneeded to be placed in the field to measure unusual condi-tions. An example is provided by the Los Angeles/LongBeach Harbor model (Figure 5-15). In operation since theearly 1970’s, it has been used to predict the effects ofharbor construction on hydrodynamics and water quality.As part of this project, wave gauges were deployed in thetwo harbors at selected sites. Figure 5-16 is an exampleof wave data from Long Beach Harbor. Although the twogauge stations were only a few hundred meters apart, theinstrument at sta 2 occasionally measured unusually highenergy compared to sta 1. The cause of these energyevents is unknown but is hypothesized to be related tolong-period harbor oscillations. The lesson is that a usermust discard data without considering the siting of theinstruments.

5-5. Analysis and Interpretation of Coastal Data

a. Background.

(1) All geologic and engineering project data,whether obtained from existing sources, field prototypecollection, laboratory analyses, or physical models, mustbe analyzed and interpreted to ultimately be useful ingeologic and geomorphic studies. The analysis proce-dures depend upon the type of data collected. Someanalyses require subjectivity or interpolation, such asconstructing geologic cross sections or making seismicinterpretations. Others are highly objective involvingcomputer probabilistic models. A coastal scientist orengineer should be aware of the assumptions, limitations,and errors involved, and should attempt to provide suffi-cient information so that his data can be replicated, analy-ses tested, and the interpretation supported.

(2) Computers play an important role in analysis andinterpretation of data from various sources. Statisticaltechniques are applied to a variety of data, including:(a) spectral analysis of wave characteristics; (b) waverefraction analysis; (c) time series analysis of water leveldata; (d) Fourier analysis of current data; (e) momentmeasures of grain size; (f) eigenvectors of shorelinechange; and (g) the use of fractals in shoreline geometry.

5-40

EM 1110-2-181031 Jan 95

Figure 5-15. Physical model of Los Angeles/Long Beach Harbor. Instruments on tripods are water level gauges

Computers are also used for numerous types of calcula-tions, such as volumetric changes in beach profiles, aswell as two-and three-dimensional plotting of thesechanges. If numerous types of spatial data exist for alocation, they may be entered into a Geographic Informa-tion System (GIS) so that important questions can beaddressed involving spatial changes. Computer softwareand hardware are also used for analysis, classification, andinterpretation of digital remotely sensed data from satel-lites and aircraft.

(3) The following sections briefly outline some con-cepts and procedures pertinent to analyses of coastal data.The reader is referred to specialized texts for detaileddescriptions of the underlying mathematics and data pro-cessing methods.

b. Wave records.

(1) General procedures.

(a) To an observer on the shore or on a boat, the seasurface usually appears as a chaotic jumble of waves ofvarious heights and periods, moving in many differentdirections. Wave gauges measure and record the chang-ing elevation of the water surface. Unfortunately, thesedata, when simply plotted against time, reflect the com-plexities of the sea’s surface and provide little initialinformation about the characteristics of the individualwaves which were present at the time the record wasbeing made (Figure 5-17). Once the water elevation dataare acquired, further processing is necessary in order toobtain wave statistics that can be used by coastal scien-tists or engineers to infer what wave forces have influ-enced their study area.

(b) Wave data analysis typically consists of a seriesof steps:

• Data transfer from gauge to computer.

5-41

EM 1110-2-181031 Jan 95

Figure 5-16. Comparison of wave gauge pressure measurements recorded at Long Beach Harbor sta 1 and 2.Although the two stations were only a few hundred meters apart, unusual energy events were recorded at sta 2which did not appear at sta 1. The abrupt shifts in the curve at each 2-hr interval represent changes in tide height.Each 2048-point record is 34.13 min long and each new wave burst is recorded at a 2-hr interval

Figure 5-17. Example of continuous wave pressure record and wave burst sampling of pressure data

5-42

EM 1110-2-181031 Jan 95

• Conversion of data from voltage readings to engi-neering units.

• Initial quality control inspection.

• Spectral analysis.

• Additional quality control (if necessary).

• Summary statistics in table and plot form.

• Plots of individual wave bursts or specialprocessing.

It is beyond the scope of this manual to discuss details ofthe above procedures. This section will summarize someaspects of data collection, quality control, analysis, andterminology. Because of the complexity of the subject,the reader is referred to Bendat and Piersol (1986),Horikawa (1988), and Weaver (1983) for additionalreferences.

(2) Data collection planning.

A continuous time series of raw pressure values plottedwith time along the x-axis is shown in Figure 5-17.Because it is impractical and too expensive to collect datacontinuously throughout the day, discrete time series or“bursts” are collected at predetermined intervals (oftenevery 2, 4, or 6 hr; Figure 5-17). Wave bursts typicallyconsist of 1,024 or 2,048 consecutive pressure, U-velocity,and V-velocity1 samples. At a sampling frequency of1 Hz, these produce time series of 17.07 min and34.13 min, respectively. Clearly, it would be desirable toacquire wave bursts frequently, but the sheer amount ofdata would soon overwhelm an analyst’s ability to orga-nize, interpret, and store the records. A researcher whoplans a data acquisition program must balance the need tocollect data frequently versus the need to maintain gaugesin the field for an extended period. There is a temptationto assume that as long as the gauges are at sea, theyshould be programmed to collect absolutely as much dataas possible. However, data management, analysis, andarchiving can cost at least as much as the deployment andmaintenance of the gauges. It is essential that these anal-ysis costs be factored into the project budget. Typicalsampling schemes used at CERC projects are listed inTable 5-10.

(3) Quality control of wave data.

_____________________________1 Orthogonal horizontal water velocity measurements.

Table 5-10Wave Data Sampling Intervals, Typical CERC Projects

Instrument LocationSample intervalhr

Sea Data self-contained wavegauge

Ocean coast-lines

4 or 6

Sea Data self-contained wavegauge

Great lakes 2 or 3

CERC Directional Wave Gauge Ocean coast-lines

1

NOAA wave and meteorologybuoys

Oceans andlakes

1

(a) One aspect of wave analysis, which is absolutelycritical to the validity of the overall results, is the qualitycontrol procedures used to ensure that the raw data col-lected by the gauges are truly representative of the waveclimate at the site. Wave gauges are subject to mechani-cal and electrical failures. The pressure sensors may beplugged or may be covered with growths while under-water. Nevertheless, even while malfunctioning, gaugesmay continue to collect data which, on cursory examina-tion, may appear to be reasonable. As an example, Fig-ure 5-18 shows pressure records from two instrumentsmounted on the same tripod off the mouth of Mobile Bay,Alabama. The upper record in the figure is from a gaugewith a plugged pressure orifice. The curve reflects theoverall change in water level caused by the tide, but highfrequency fluctuations caused by the passing of waveshave been severely damped. The damping is more obvi-ous when a single wave burst of 1,024 points is plotted(Figure 5-19). Without the record from the second gauge,would an analyst have been able to conclude that the firstinstrument was not performing properly? This type ofdetermination can be especially problematic in a low-energy environment like the Gulf of Mexico, where calmweather can occur for long periods.

(b) Another difficult condition to diagnose occurswhen the wave energy fluctuates rapidly. Many comput-erized analysis procedures contain user-specifiedthresholds to reject records that contain too many noisespikes. Occasionally, however, violent increases inenergy do occur over a short time, and it is important thatthe analysis procedures do not reject these records withoutverification. As an example, one of two gauges in LongBeach harbor (the lower curve in Figure 5-16) may havemalfunctioned and written many noise spikes on the tape.In reality, the gauge recorded unusual energy eventswithin the harbor. Another example, from Burns Harbor,Indiana, is shown in Figure 5-20. When wave height wasplotted against time, numerous spikes appeared. In this

5-43

EM 1110-2-181031 Jan 95

Figure 5-18. Pressure data collected by two gauges mounted on a tripod off Mobile Bay, Alabama. The upperrecord is from a gauge with a plugged pressure orifice. The abrupt increase in pressures near day 43 was causedwhen a fishing boat struck and overturned the tripod

case, the rapid increase in energy was genuine, and thespikey appearance was caused by the plotting of manyweeks of data on one plot. An examination of the indi-vidual pressure records (Figure 5-21) reveals how rapidlythe energy increased in only a few hours (a characteristicof Great Lakes storms). This example demonstrates thatthe method of displaying wave statistics can have a majorinfluence on the way the data are perceived by an analyst.Additional examples and quality control procedures forvalidating wave data are discussed in Morang (1990).

(4) Analysis procedures and terminology.

(a) Wave data analysis can be broadly subdivided intonon-directional and directional procedures. Although thelatter are considerably more complex, the importance ofdelineating wave direction in coastal areas is usually greatenough to justify the extra cost and complexity of tryingto obtain directional wave spectra. The types of wavestatistics needed vary depending on the application. Forexample, a geologist might want to know what the aver-age wave period, height, and peak direction are along astretch of the shoreline. This information could then beused to estimate wave refraction and longshore drift. An

engineer who is building a structure along the shorewould be interested in the height, period, and approachdirection of storm waves. He would use these values tocalculate stone size for his structure. Table 5-11 listscommon statistical wave parameters.

(b) Table 5-11 is intended to underscore that waveanalysis is a complex procedure and should be undertakenby coastal researchers with knowledge of wave mechanicsand oceanography. In addition, researchers are urged tobe cautious of wave statistics from secondary sources andto be aware of how terms have been defined and statisticscalculated. For example, “significant wave height” isdefined as the average height of the highest one-third ofthe waves in a record. How long should this record be?Are the waves measured in the time domain by countingthe wave upcrossings or downcrossings? The two meth-ods may not produce the same value ofHs. Might it notbe better to estimate significant wave height by perform-ing spectral analysis of a wave time series in the fre-quency domain and equatingHs = Hm0? This is theprocedure commonly used in experiments where largeamounts of data are processed. The latter equivalency is

5-44

EM 1110-2-181031 Jan 95

Figure 5-19. Example of a single wave burst of 1,024 pressure points from the same gauges which produced therecords in Figure 5-18. The data from the plugged gauge (the upper curve) is not only reduced in amplitude butalso shifted in phase. It is essentially impossible to correct the plugged data and recreate even an approximationof the original

usually considered valid in deep and intermediate waterbut may not be satisfactory in shallow water (Horikawa1988).

(c) Directional wave statistics are also subject to mis-interpretations depending upon the computation method.At sea, very rarely do the waves come from only onedirection. More typically, swell, generated by distantstorms, may approach from one or more directions, whilethe local wind waves may have a totally different orienta-tion. Researchers need to distinguish how the waveenergy is distributed with respect to both direction andperiod (i.e., the directional spectral density,S(f,Θ)). Thedirectional distribution of wave energy is often computedby a method developed by Longuet-Higgins, Cartwright,and Smith (1963) for use with floating buoys in deepwater. Other distribution functions have been proposedand used by various researchers since the 1970’s(Horikawa 1988). Although the various methods do notproduce the same directional wave statistics under somecircumstances, it is not possible to state that one methodis superior to another.

(d) The user of environmental data must be aware ofthe convention used to report directions. Table 5-12 liststhe definitions used at CERC; other institutions may notconform to these standards.

(e) Some oceanographic instruments are sold withsoftware that performs semi-automatic processing of thedata, often in the field on personal computers. In someinstruments, the raw data are discarded and only theFourier coefficients saved and recorded. The user ofthese instruments is urged to obtain as much informationas possible on the mathematical algorithms used by thegauge’s manufacturer. If these procedures are not thesame as those used to analyze other data sets from thearea, the summary statistics may not be directly compara-ble. Even more serious, this author has encountered com-mercial processing software which was seriously flawedwith respect to the calculation of directional spectra. Inone field experiment, because the original raw data hadnot been archived in the gauge, the data could not bereprocessed or the errors corrected. As a result, themulti-month gauge deployment was rendered useless.

(f) In summary, it is vital that the user of wave databe aware of how wave statistics have been calculated andthoroughly understand the limitations and strengths of thecomputational methods that were employed.

(5) Display of wave data and statistics.

(a) In order to manage the tremendous amount ofdata that are typically acquired in a field experiment,

5-45

EM 1110-2-181031 Jan 95

Figure 5-20. Analyzed wave data from Burns Harbor, Indiana. Spikey appearance is caused by plotting almost3 months of data on one plot

perform quality control, and interpret the results, wavedata should be analyzed as soon as possible. In addition,there is often an urgent need to examine the raw data toascertain whether the gauges can be redeployed or mustbe repaired.

(b) Figures 5-17 and 5-19 are examples of pressureplotted against time. The value of this form of displayfor quality control purposes has been demonstrated, butthese plots are of limited value in revealing informationabout the overall nature of the wave climate in the studyarea.

(c) To review the data from an extended deployment,the summary statistics must be tabulated or plotted.

Figure 5-22 is an example of tabulated directional wavedata from a Florida project site. These same data aregraphically displayed in Figure 5-23. The upper plotshows Hm0 wave height, the center peak period, and thelower peak direction. Although other statistics could havebeen plotted on the same page, there is a danger of mak-ing a display too confusing. The advantage of the tabula-tion is that values from individual wave bursts can beexamined. The disadvantage is that it is difficult to detectoverall trends, especially if the records extend over manymonths. As data collection and processing proceduresimprove, and as more and more data are acquired at fieldprojects, it will be increasingly difficult to display the

5-46

EM 1110-2-181031 Jan 95

Figure 5-21. Pressure data from Burns Harbor, Indiana, April 6, 1988. This plot shows how dramatically the energycan increase in only a few hours

results in a useful and flexible format that does notoverwhelm the end user but yet also does not over-simplify the situation.

(6) Applications of wave data. One important use ofwave climate data in coastal engineering is in the con-struction of wave refraction diagrams. These demonstratehow nearshore bathymetry influences the direction ofwaves approaching the shoreline. This information can beused to estimate mass transport and longshore transport ofsediment, which, in turn, can be used to predict morpho-logic changes under both natural and structurally influ-enced coasts. Wave refraction analyses can also be usedfor hypothetical scenarios, such as predicting the effectson incident waves of dredging an offshore shoal or dump-ing dredged materials offshore.

c. Water level records.

(1) Changes in water levels along coastlines have pro-found influence on the geology, the natural ecology, andhuman habitation in these regions. Predicting and under-standing these changes can guide coastal planners indeveloping rational plans for coastal development and inthe design, construction, and operation of coastalstructures and waterways. Causes of sea level changesalong open coasts have been discussed in Chapter 2.

(2) Tide gauge records may be analyzed for spatialinterpolation and for assessing temporal variations such assurges, tides, seasonal changes, and long-term trends.Discrepancies between the predicted tide at one site andthe actual tide measured only a short distance away maybe considerable. A method for adjusting between pre-dicted tides at a station and those at a nearby study areausing only limited field measurements is discussed byGlen (1979). Other analysis methods are discussed inEM 1110-2-1414.

(3) For engineering projects, assessments of short-term water level changes range from simple plotting ofdata to more sophisticated mathematical analyses. In somecases, some of the components that drive water levelchanges can be isolated. To assess longer (multiyear)trends, it is important to dampen or separate the effects ofyearly variability so that the nature of the secular trendsbecomes more pronounced. Least-squares regressionmethods are typically inadequate because the seculartrends often show pronounced nonlinearity (Hicks 1972).It may also be important to examine long-term periodiceffects in a long data record such as the 18.6-year nodalperiod, which Wells and Coleman (1981) concluded wasimportant for mudflat stabilization in Surinam.

5-47

EM 1110-2-181031 Jan 95

5-48

EM 1110-2-181031 Jan 95

Table 5-12Reporting Conventions for Directional EnvironmentalMeasurements

Type Convention Example

Wind FROM WHICH wind isblowing

North wind blows from0 deg

Waves FROM WHICH wavescome

West waves come from270 deg

Unidirectionalcurrents

TO WHICH currents areflowing

East current flowing to90 deg

(4) Historic water levels have been used by Hands(1979, 1980) to examine the changes in rates of shoreretreat in Lake Michigan and to predict beach/nearshoreprofile adjustments to rising water levels. Additionalresearch is being sponsored by the International JointCommission to model how changing water levels affecterosion of various bluff stratigraphies and the nearshoreprofile.

d. Current records. Current data are often criticalfor evaluating longshore and cross-shore sediment trans-port and for evaluating hydraulic processes in inlets andother restricted waterways. Currents, which are generatedby a variety of mechanisms, vary greatly spatially andtemporally in both magnitude and direction. Four generalclasses of unidirectional flow affect coastal environmentsand produce geologic changes. These include:

• Nearshore wave-induced currents, including long-shore and rip currents.

• Flow in tidal channels and inlets, which typicallychanges direction diurnally or semi-diurnally,depending on the type of tide along the adjacentcoast.

• River discharge.

• Oceanic currents, which flow along continentalland masses.

This section will briefly discuss the first two of thesetopics and present data examples. The third and fourthare beyond the scope of this manual, and the reader isreferred to outside references for additional information.

(1) Nearshore wave-induced currents.

(a) In theory, one of the main purposes for measuringnearshore, wave-induced currents is to estimate longshoretransport of sediments. At the present level of technology

and mathematical knowledge of the physics of sedimenttransport, the direct long-term measurement of longshorecurrents by gauges is impractical. Two main reasonsaccount for this situation. First, deployment, use, andmaintenance of instruments in the nearshore and the surfzone are difficult and costly. Second, the mechanics ofsediment transport are still little-understood, and no onemathematical procedure is yet accepted as the definitivemethod to calculate sediment transport, even when cur-rents, grain size, topography, and other parameters areknown. An additional consideration is how to monitor thevariation of current flow across and along the surf zone.Because of the extreme difficulty of obtaining data fromthe surf zone, neither the cross-shore variations of cur-rents nor the temporal changes in longshore currents arewell known.

(b) Longshore (or littoral) drift is defined as:“Material (such as shingle, gravel, sand, and shell frag-ments) that is moved along the shore by a littoral current”(Bates and Jackson 1984). Net longshore drift refers tothe difference between the volume of material moving inone direction along the coast and that moving in theopposite direction (Bascom 1964). Along most coasts,longshore currents change directions throughout the year.In some areas, changes occur in cycles of a few days,while in others the cycles may be seasonal. Therefore,one difficulty in determining net drift is defining a perti-nent time frame. Net drift averaged over years or decadesmay conceal the fact that significant amounts of materialmay also flow in the opposite direction.

(c) Because net longshore currents may vary greatlyfrom year to year along a stretch of coastline, it would bedesirable to deploy current meters at a site for severalyears in order to obtain the greatest amount of data possi-ble. Unfortunately, the cost of a multi-year deploymentcould be prohibitive. Even a long deployment might notdetect patterns that vary on decade-long scales, such asthe climatic changes associated with El Niño. At a mini-mum, near-shore currents should be monitored at a fieldsite for at least a year in order to assess the changes asso-ciated with the passing seasons. Coastal scientists andengineers must be aware of the limitations of field currentdata and recognize that long-term changes in circulationpatterns may remain undetected despite the best fieldmonitoring efforts.

(2) Flow in tidal channels and inlets.

(a) An inlet is “a small, narrow opening in a shore-line, through which water penetrates into the land” (Bates

5-49

EM 1110-2-181031 Jan 95

Figure 5-22. Example of tabular summary of wave data from offshore Fort Walton Beach, Florida

5-50

EM 1110-2-181031 Jan 95

Figure 5-23. Plots of wave height, peak period, and peak direction from offshore Fort Walton Beach, Florida

5-51

EM 1110-2-181031 Jan 95

and Jackson 1984). Inlets range in size from short, nar-row breaches in barrier islands to wide entrances of majorestuaries like Chesapeake Bay. Many geologic and engi-neering studies concern flow through tidal inlets in sand-dominated barriers, particularly when the inlets serve asnavigation channels connecting harbors to the open sea.

(b) Inlets allow for the exchange of water betweenthe sea and the bay during each tidal cycle. Therefore,currents in tidal inlets are typically unidirectional, chang-ing direction diurnally or semidiurnally, depending uponthe tides along the adjacent open coast. Flow through theinlets can be complicated by the hydrodynamics of theinland bay, especially if there are other openings to thesea.

(c) Various numerical and conceptual models havebeen developed to describe flow through inlets and allowresearchers to predict the effects of changing inlet dimen-sions, lengths, and orientations (Aubrey and Weishar1988; Escoffier 1977; Seelig, Harris, and Herchenroder1977; Shore Protection Manual1984). Most models,however, benefit from or require calibration with physicalmeasurements made within the inlet and the generalvicinity. The required field measurements are usuallyeither tidal elevations from the open sea and within theadjacent bay or actual current velocities from within theinlet’s throat.

(d) Display of tidal elevation data is relativelystraightforward, usually consisting of date or time on thex-axis and elevation on the y-axis. Examples of tidalelevations from a bay and an inlet in the Florida Panhan-dle are shown in Figure 5-6. Although the overall enve-lope of the curves is similar, each one is unique withrespect to the heights of the peaks and the time lags. Thecurves could be superimposed to allow direct comparison,but, at least at this 1-month-long time scale, the resultwould be too complicated to be useful.

(e) Display of current meter measurements is moredifficult because of the large quantity of data usually col-lected. An added difficulty is posed by the changingcurrents within an inlet, which require a three-dimensionalrepresentation of the flow which varies with time. Cur-rent measurements from East Pass, Florida, collectedduring three field experiments in the mid-1980’s are pre-sented as examples. Currents were measured with manualPrice type AA meters deployed from boats and withtethered Endeco 174 current meters. The manual meas-urements were made hourly for 24 hr in order to observea complete tidal cycle. The measurements were madeacross the inlet at four stations, each one consisting of a

near-surface, a mid-depth, and a near-bottom observation(Figure 5-24). Therefore, 12 direction and velocity datavalues were obtained at each hour (Figure 5-25). Oneway to graphically display these values is to plot thevelocities on a plan view of the physical setting, as shownin Figure 5-24. This type of image clearly shows thedirections and relative magnitudes of the currents. In thisexample, the data reveal that the currents flow in oppositedirections in the opposite halves of the inlet. The disad-vantage of the plan view is that it is an instantaneoussnapshot of the currents, and the viewer cannot follow thechanges in current directions and magnitudes over timeunless the figure is redrawn for each time increment.Temporal changes of the currents can be shown on dualplots of magnitude and direction (Figure 5-26). Unfortu-nately, to avoid complexity, it is not reasonable to plot thedata from all 12 measurement locations on a single page.

Therefore, measurements from the same depth are plottedtogether, as in Figure 5-26, or all measurements from onesite can be plotted together (top, middle, and bottom).

(f) In summary, current data can be displayed in theform of instantaneous snapshots of the current vectors oras time series curves of individual stations. Many plotsare usually needed to display the data collected from evenshort field projects. It may be advantageous to presentthese plots in a data appendix rather than within the textof a report.

(3) Error analysis of current data.

(a) Error analysis of current records can be broadlydivided into two categories. The first concerns calibra-tions of the actual current sensing instruments. A userneeds to know how closely the numbers reported by aparticular instrument represent the water motions that it ispurported to be measuring. This information is importantfor both evaluation of existing data sets and for planningof new field experiments, where some instruments may bemore suitable than others.

(b) The second broad question pertains to whetherthe measurements which have been gathered adequatelyrepresent the flow field in the inlet or channel that isbeing examined. This second problem is exceedinglydifficult to evaluate because it raises the fundamentalquestions of “How much data do I need?” and, “Can Iafford to collect the data that will really answer myquestions?” The user is typically tempted to respond thathe wants just as much data as possible, but this mayprove to be counterproductive. For example, if the

5-52

EM 1110-2-181031 Jan 95

Figure 5-24. Current measurement stations in East Pass Inlet, Destin, Florida, during October, 1983. Measurementswere made hourly from small boats. At 02:10 CST, currents were flowing to the northwest along the west side ofthe inlet and to the southeast along the center and east sides of the inlet. Station 2 was in the mixing zone

5-53

EM 1110-2-181031 Jan 95

Figure 5-25. Example of handwritten field notes listing times and data values of East Pass current measurements.The data are efficiently presented but difficult to visualize

5-54

EM 1110-2-181031 Jan 95

Figure 5-26. Time series plots of East Pass current speed (bottom) and direction (top)

5-55

EM 1110-2-181031 Jan 95

currents in an inlet are being measured to determinevariations in the tidal prism over time, will a densegridwork of sampling stations in an inlet provide moreuseful data? Or might the excess data reveal unnecessarydetails about turbulence and mixing in the inlet? Theseare intrinsically interesting questions, but may not begermane to the engineering problems that must beaddressed. Although the dense grid pattern of data can beused to evaluate overall flow, the collection, analysis, andmanagement of the excess data can be costly and time-consuming. The money used on management of this datamight be better spent extending a simpler sampling pro-gram for a longer period at the site. Unfortunately, thereare no firm guidelines to planning current studies andplacing instruments.

(c) Analysis of error from various types of currentsensors has been the subject of extensive study in the last30 years. Numerous types of error can occur, both duringfield deployment of the instrument and during data pro-cessing. These can result from instrument calibration,clock time errors, and data recording and playback. Inaddition, the user is cautioned that each of the many typesand brands of current meters is capable of recordingaccurately only a segment of the spectrum of watermotions because of the influence of the mooring assem-bly, type of velocity sensor used, and recording scheme ofthe instrument (Halpern 1980). Halpern’s (1980) paperlists many references pertaining to the results of tests ofmoored current meters.

(d) Manufacturers of current meters publish accuracystandards in their literature. These standards may beoptimistic, especially under the adverse conditionsencountered in many coastal settings. In addition, thetype of mooring used for the instrument affects the qualityof the measured data (Halpern 1978). For these reasons,the user of existing data is urged to obtain as much infor-mation as possible regarding the specifics of the deploy-ment and the type of mooring in order to try to assess theaccuracy of the results. Ultimately, successful use ofcurrent gauges is critically dependent upon the planningof the experiment and upon the care and skill of the tech-nicians who maintain and deploy the instruments.

(4) River discharge.

(a) River outflow has a major effect on some coast-lines, particularly where massive deltas have formed (e.g.Mississippi Delta). Even if a study area is not located ona delta, coastal researchers must be aware of the potentialimpact of rivers on coastal processes, especially if the

study region is affected by freshwater runoff at certainseasons or if longshore currents carry river-derived sedi-ment along the shore.

(b) River discharge data are available for manycoastal rivers. A cursory examination of the annualhydrograph will reveal the seasonal extremes. Because ofthe episodic nature of coastal flooding, annual dishargefigures may be misleading. A useful parameter to esti-mate river influence on the coast is the hydrographic ratio(HR), which compares tidal prism volume with fluvial dis-charge volume (Peterson et al. 1984).

(5) Oceanic currents.

(a) Major oceanic currents intrude onto some conti-nental shelves with enough bottom velocity to transportsandy sediments. The currents operate most effectivelyon the outer shelf, where they may transport significantvolumes of fine-grained sediments but presumably con-tribute little if any new sediment (Boggs 1987). Alongmost coastlines, ocean currents have little direct effect onshoreline sedimentation or erosion. Even off southeastFlorida, where the continental shelf is narrow, the westernedge of the Gulf Stream flows at least 1/2 km offshore.However, in some locations where currents approach thecoastal zone, sediment discharged from rivers is trans-ported and dispersed along the adjacent coastline. Thisprocess may arrest the seaward progradation of the deltafront, while causing extensive accumulations of riverine-derived clastics downdrift of the river mouth (Wright1985).

(b) In shallow carbonate environments, reefs thrivewhere currents supply clean, fresh ocean water. Reefsstabilize the bottom, provide habitat for marine life, pro-duce carbonate sediment, and sometimes protect theadjoining shore from direct wave attack (i.e., the GreatBarrier reef of Australia). In the United States, live reefsare found in the Gulf of Mexico off Texas and westFlorida and in the Atlantic off Florida. Coral islands arefound in the Pacific in the United States Trust Territories.For geologic or engineering studies in these environments,there may be occasional need to monitor currents. Proce-dures of deepwater current measurement are presented inAppell and Curtin (1990) and McCullough (1980).

(c) In summary, the effect of tide or wave-inducedcurrents is likely to be much more important to mostcoastal processes than ocean currents. Measurement ofocean currents may occasionally be necessary for geologicstudies in deltaic or carbonate environments.

5-56

EM 1110-2-181031 Jan 95

e. Shoreline change mapping1.

(1) Introduction.

(a) Maps and aerial photographs can provide a wealthof useful information for the interpretation of geologiccoastal processes and evolution. Maps and photographscan reveal details on:

• Long-term and short-term advance or retreat ofthe shore.

• Longshore movement of sediments.

• The impact of storms, including barrier islandbreaches, overwash, and changes in inlets, vegeta-tion, and dunes.

• Problems of siltation associated with tidal inlets,river mouths, estuaries, and harbors.

• Human impacts caused by construction ordredging.

• Compliance with permits.

• Biological condition of wetlands and estuaries.

(b) The use of maps and aerial photographs to deter-mine historical changes in shoreline position is increasingrapidly. Analyzing existing maps does not require exten-sive field time or expensive equipment, therefore oftenproviding valuable information at an economical price.This section summarizes the interpretation of shorelineson photographs and maps and corrections needed to con-vert historic maps to contemporary projections and coordi-nate systems.

(c) Many possible datums can be used to monitorhistorical changes of the shoreline. In many situations,the high water line (hwl) has been found to be the bestindicator of the land-water interface, the coastline(Crowell, Leatherman, and Buckley 1991). The hwl iseasily recognizable in the field and can usually beapproximated from aerial photographs by a change incolor or shade of the beach sand. The datum printed onthe NOS T-sheets is listed as “Mean High Water.” Fortu-nately, the early NOS topographers approximated hwlduring their survey procedures. Therefore, direct

_____________________________1 Material in this section adapted from Byrnes and Hiland(1994) and other sources.

comparisons between historical T-sheets and modernaerial photographs are possible. In order to calculate thegenuine long-term shoreline change, seasonal beach widthvariations and other short-term changes should be filteredout of the record. The best approach is to use only mapsand aerial images from the same season, preferably sum-mertime, when the beach is exposed at its maximumwidth.

(d) A crucial problem underlying the analysis of allhistorical maps is that they must be corrected to reflect acommon datum and brought to a common scale, projec-tion, and coordinate system before data from successivemaps can be compared (Anders and Byrnes 1991). Mapsmade before 1927 have obsolete latitude-longitude coordi-nate systems (U.S. datum or North American (NA)datum) that must be updated to the current standard ofNAD 1927 or the more recent NAD 1983. To align mapsto a specific coordinate system, a number of stable andpermanent points or features must be identified for whichaccurate and current geographic coordinates are known.These locations, called primary control points, are used bycomputer mapping programs to calculate the transforma-tions necessary to change the map’s projection and scale.The most suitable control points are triangulation stationswhose current coordinates are available from the NationalGeodetic Survey.

(e) Maps that were originally printed on paper havebeen subjected to varying amounts of shrinkage. Theproblem is particularly difficult to correct if the shrinkagealong the paper’s grain is different than across the grain.Maps with this problem have to be rectified or discarded.In addition, tears, creases, folds, and faded areas in papermaps must be corrected.

(f) Air photographs, which are not map projections,must be corrected by optical or computerized methodsbefore shore positions compiled from the photos can bedirectly compared with those plotted on maps. The dis-tortion correction procedures are involved because photosdo not contain defined control points like latitude-longitude marks or triangulation stations. On manyimages, however, secondary control points can beobtained by matching prominent features such as thecorners of buildings or road intersections with theirmapped counterparts (Crowell, Leatherman, and Buckley1991). Types of distortion which must be correctedinclude:

• Tilt. Almost all vertical aerial photographs aretilted, with 1 deg being common and 3 deg notunusual (Lillesand and Kiefer 1987). The scaleacross tilted air photos is non-orthogonal,

5-57

EM 1110-2-181031 Jan 95

resulting in gross displacement of featuresdepending upon the degree of tilt.

• Variable scale. Planes are unable to fly at a con-stant altitude. Therefore, each photograph in aseries varies in scale. A zoom transfer scope canbe used to remove scale differences betweenphotos.

• Relief displacement. Surfaces which rise abovethe average land elevation are displaced outwardfrom the photo isocenter. Fortunately, most U.S.coastal areas, especially the Atlantic and Gulfbarriers, are relatively flat and distortion causedby relief displacement is minimal. However,when digitizing cliffed shorelines, control pointsat about the same elevation as the feature beingdigitized must be selected.

• Radial lens distortion. With older aerial lenses,distortion varied as a function of distance fromthe photo isocenter. It is impossible to correct forthese distortions without knowing the make andmodel of the lens used for the exposures(Crowell, Leatherman, and Buckley 1991). Ifoverlapping images are available, digitizing thecenters, where distortion is least, can minimizethe problems.

Fortunately, most errors and inaccuracies from photogra-phic distortion and planimetric conversion can be quanti-fied. Shoreline mapping exercises have shown that if careis taken in all stages of filtering original data sources,digitizing data and performing distortion corrections, theresulting maps meet, and often exceed, National MapAccuracy Standards (Crowell, Leatherman, and Buckley1991).

(g) In order to accurately document shoreline positionand generate shoreline position maps, several steps areneeded to quantify shoreline change. These steps includeassembling data sources, entering data, digitizing coordi-nates, analyzing potential errors, computing shorelinechange statistics, and interpreting shoreline trends. Basedon shoreline change studies conducted at universities andFederal, State, and local agencies, a brief summary of therecommended techniques and procedures is given below.

(2) Data sources. Five potential data sources exist forassessing spatial and temporal changes in shoreline posi-tion. These include United States Geological Survey(USGS) topographic quadrangles, National Ocean Service

(NOS) topographic sheets, local engineering surveys,near-vertical aerial photographs, and GPS surveys. Eachdata source addresses a specific mapping purpose, asdescribed below.

(a) USGS topographic maps. The most commonmaps used for documenting changes along the coast areUSGS topographic quadrangle maps. These maps arecreated at a range of scales from 1:24,000 to 1:250,000(Ellis 1978). The primary purpose of these maps is toportray the shape and elevation of the terrain above agiven datum, usually the mean high water line. Accuratedelineation of the shoreline was not a primary concern onthese land-oriented maps. However, shoreline positionroutinely is revised on 1:24,000 topographic maps usingaerial photographic surveys. Many shoreline mappingstudies have used these maps for quantifying changes inposition, but more accurate and appropriate sources shouldbe employed if available.

(b) NOS Topographic Maps. Another type of topo-graphic map is that produced by the U.S. Coast andGeodetic Survey (USC&GS) (now National Ocean Sur-vey - NOS). Because this agency is responsible for sur-veying and mapping topographic information along thecoast, topographic map products (T-sheets) have beenused in the study of coastal erosion and protection, andfrequently in law courts in the investigation of land own-ership (Shalowitz 1962). Most of these maps are plani-metric in that only horizontal position of selected featuresis recorded; the primary mapped feature is the high-watershoreline. From 1835 to 1927, almost all topographicsurveys were made by plane table; most post-1927 mapswere produced using aerial photographs and photogram-metric methods (Shalowitz 1964). NOS shoreline positiondata are often used on USGS topographic quadrangles,suggesting that T-sheets are the primary source for accu-rate shoreline surveys. Scales of topographic surveys aregenerally 1:10,000 or 1:20,000. These large-scale prod-ucts provide the most accurate representation of shorelineposition other than direct field measurements using sur-veying methods.

(c) Large-scale engineering surveys. In areas ofsignificant human activity, engineering site maps oftenexist for specific coastal regions. However, surveyedareas often are quite limited by the scope of the project;regional mapping at large scale (greater than 1:5,000) issparse. If these data do exist, they potentially provide themost accurate estimates of high-water shoreline positionand should be used. These data are valuable for rectify-ing aerial photography for mapping shoreline position.

5-58

EM 1110-2-181031 Jan 95

(d) Near-vertical aerial photography. Since the1920’s, aerial photography has been used to record shore-line characteristics in many coastal regions. However,these data cannot be used directly to produce a map.Aircraft tilt and relief may cause serious distortions thathave to be removed by rectification. A number of graphi-cal methods and computational routines exist for remov-ing distortions inherent in photography (Leatherman 1984;Anders and Byrnes 1991). Orthophotoquadrangles andorthophotomosaics are photomaps made by applying dif-ferential rectification techniques (stereo plotters) toremove photographic distortions. Users are warned thatconversion of photographic images to map projections isnot a trivial procedure, despite the availability of moderncartographic software. Ease of data collection and thesynoptic nature of this data source provide a significantadvantage over most standard surveying techniques.

(e) GPS surveys. During the late 1970’s through the1980’s, significant advances in satellite surveying weremade with the development of the Navigation SatelliteTiming and Ranging (NAVSTAR) Global PositioningSystem (GPS). The GPS was developed to support mili-tary navigation and timing needs; however, many otherapplications are possible with the current level of technol-ogy. This surveying technique can be very accurate undercertain conditions; however, signal degradation throughselective availability causes significant positional errors ifonly one station is being used (Leick 1990). DifferentialGPS provides the capability for accurately delineatinghigh-water shoreline position from ground surveys.

(3) Data entry.

(a) Frequently, shoreline maps have variable scalesand use different datums and coordinate systems. Shore-line maps should be corrected to reflect a common datumand brought to a common scale, projection, and coor-dinate system before data from successive maps can accu-rately be compared. There are several computercartographic systems, consisting of a high-precision digi-tizing table and cursor and computer interface, available.Ideally, a GIS system should be used to digitize variousdata sources and store the information in data layers thatcan be linked to a relational database. Most systems havea table or comment file associated with each data layer todocument the original map source, cartographic methods,and potential errors.

(b) Digitized shoreline points are commonly enteredinto an X-Y data file for each shoreline source. Datatransformation to a common surface can be convertedusing standard mapping software packages. A header or

comment line should be incorporated into the digitalrecord of the cartographic parameters, i.e., map scale,projection, and horizontal and vertical datums. Datumshifts can account for significant changes in historicalmaps generated in the 1800’s.

(c) Another transformation is the projection of theearth’s spherical shape onto a two-dimensional map. Themethod by which the earth’s coordinates are transferred toa map is referred to as the map projection. Commonlyused projections include:

• Lambert projection - scaled to correct along twostandard parallels; base for state planecoordinates.

• Mercator projection - scaled along uniformlyspaced straight parallel lines; used for NOST- and H-sheets.

• Transverse Mercator projection - essentially thestandard Mercator rotated through 90 deg; usedfor all large scale maps such as USGS 7½′quadrangles.

Additional projections are described in Ellis (1978).

(d) Before the shoreline is digitized, triangulationpoints should be digitized for each shoreline map. Thetriangulation stations provide control points, which arecrucial when using older maps or a multitude of differentmap sources (Shalowitz 1964). Older maps may containmisplaced coordinate systems. If there is not enoughinformation on the coordinate system or triangulationstation, the map should not be used for quantitative data.A useful source of available U.S. triangulation stations isDatum Differences(USC&GS 1985).

(e) Media distortion can be eliminated by usingmaps drawn on stable-base materials such as NOS T- andH-sheets. Most USACE District project maps are madefrom Mylar film. The original map or a high-qualityMylar copy should be used as opposed to black-line, blue-line, or other paper-based medium. However, if papermaps are used, and distortion from shrinking and swellingis significant, the digitizer setup provides some degree ofcorrection by distributing error uniformly across the map.In addition, rubber-sheeting and least-squares fit computerprograms allow the user to define certain control pointsand correct for distortion errors as much as possible. It isalso important to remember that data in digital formacquire no new distortions, whereas even stable-basemaps can be torn, wrinkled, and folded. Scale distortion

5-59

EM 1110-2-181031 Jan 95

from optical methods of map reproduction are also cor-rected by bringing all maps to a 1:1 scale.

(4) General digitizing guidelines. Cartographic meth-ods and map handling should be consistent within a proj-ect and organization. Shoreline digitizing guidelines(summarized by Byrnes and Hiland, in preparation)include:

(a) All shorelines are digitized from stable-basematerials. If possible use NOS T- and H-sheets onMylar, or on bromide if Mylar is not available for a par-ticular map. Shorelines mapped from rectified aerialphotography are drawn onto, and digitized from, acetatefilm.

(b) To prevent curling and wrinkling of maps, storecartographic and photographic materials flat or vertical.Bromide-based maps that are shipped in a map tubeshould be kept flat for several days before digitizing.

(c) When attaching a map to a digitizer table, the areabeing digitized is always perfectly flat. Any wrinkles cancause that portion of the map to move during digitizing,creating positional errors. High-quality drafting tape ormasking tape is used to attach the map. One corner istaped first, then the map is smoothed diagonally and theopposite corner is taped securely; this procedure isrepeated for the other two corners. Once the corners aresecured, the map is smoothed from the center to the edgesand taped along each edge.

(d) High-precision equipment must be used for accu-rate shoreline change mapping. Digitizer tables andcursors with a precision of 0.1 mm are recommended.This magnitude of change equates to 1 m of ground dis-tance at a scale of 1:10,000. The center bead or crosshairshould ideally be smaller than the width of the line beingdigitized; the smallest pen width generally available is0.13 mm. The width of the crosshair of a high-precisioncursor is approximately 0.1 mm.

(e) When digitizing, use manual point input asopposed to stream input. Stream input places points at aspecified distance as the user traces over the line beingdigitized. This procedure tends to make a very uniformand smooth line. However, it could miss some curvaturein the line if the specified distance is too large; likewise,it could accept more points than are needed if the speci-fied distance is too small, resulting in extremely largefiles, as well as storage and display problems. In addi-tion, if the user’s hand slips during the digitizing process,

stream digitizing will continue to place points in the erro-neous locations. These can be difficult and time-consuming to correct. Manual digitizing allows the userto place points at non-uniform distances from each other,and therefore allows the user to represent all variations inthe shoreline.

(f) The seaward edge of the high-water shoreline andthe center point of the printed bathymetric soundingshould be used as the reference positions for data capture.

(5) Potential errors. It is important that all availableprocedures be used as carefully as possible to capture mapdata; however, no matter how cautious the approach, acertain amount of error will be generated in all measure-ments of digitized horizontal position. Potential errors areintroduced in two ways.Accuracyrefers to the degree towhich a recorded value conforms to a known standard. Inthe case of mapping, this relates to how well a position ona map is represented relative to actual ground location.Precision,on the other hand, refers to how well a meas-urement taken from a map or an aerial photograph can bereproduced. Table 5-13 lists the factors affecting themagnitude of error associated with data sources and meas-urement techniques. Both types of error should be eval-uated to gage the significance of calculated changesrelative to inherent inaccuracies. The following discus-sion addresses these factors in terms of data sources,operator procedures, and equipment limitations.

(6) Cartographic sources.

(a) Shoreline measurements obtained from historicalmaps can only be as reliable as the original mapsthemselves. Accuracy depends on the standards to whicheach original map was made, and on changes which mayhave occurred to a map since its initial publication. Fieldand aerial surveys provided the source data used to pro-duce shoreline maps. For T- and H-sheets at a 1:10,000scale, national standards allow up to 8.5 m of error for astable point (up to 10.2 m of error at 1:20,000), but thelocation of these points can be more accurate (Shalowitz1964; Crowell, Leatherman, and Buckley 1991). Non-stable points are located with less accuracy; however,features critical to safe marine navigation are mapped toaccuracy stricter than national standards (Ellis 1978). Theshoreline is mapped to within 0.5 mm (at map scale) oftrue position, which at 1:10,000 scale is 5.0 m on theground.

(b) Potential error considerations related to fieldsurvey equipment and mapping of high-water shoreline

5-60

EM 1110-2-181031 Jan 95

Table 5-13Factors Affecting Potential Errors Associated with Cartographic Data Sources

Accuracy

Maps and Charts Field Surveys and Aerial Photographs Precision

ScaleHorizontal DatumShrink/StretchLine ThicknessProjectionEllipsoidPublication Standards

Location, Quality, and Quantityof Control Points

Interpretation of High-Water LineField Surveying StandardsPhotogrammetric StandardsAircraft Tilt and PitchAircraft Altitude ChangesTopographic ReliefFilm Prints Versus Contact Prints

Annotation of High-Water LineDigitizing EquipmentTemporal Data ConsistencyMedia ConsistencyOperator Consistency

(After Anders and Byrnes (1991))

position were addressed by Shalowitz (1964; p. 175) asfollows:

With the methods used, and assuming the normalcontrol, it was possible to measure distances withan accuracy of 1 meter (Annual Report, U.S. Coastand Geodetic Survey 192, 1880) while the positionof the plane table could be determined within 2 or3 meters of its true position. To this must beadded the error due to the identification of theactual mean high water line on the ground, whichmay approximate 3 to 4 meters. It may thereforebe assumed that the accuracy of location of thehigh-water line on the early surveys is within amaximum error of 10 meters and may possibly bemuch more accurate than this. This is the accu-racy of the actual rodded points along the shoreand does not include errors resulting from sketch-ing between points. The latter may, in some cases,amount to as much as 10 meters, particularlywhere small indentations are not visible to thetopographer at the planetable.

The accuracy of the high-water line on early topo-graphic surveys of the Bureau was thus dependentupon a combination of factors, in addition to thepersonal equation of the individual topographer.But no large errors were allowed to accumulate.By means of the triangulation control, a constantcheck was kept on the overall accuracy of thework.

(c) In addition to survey limitations listed byShalowitz (1964), line thickness and cartographic errors(relative location of control points on a map) can beevaluated to provide an estimate of potential inaccuracy

for source information. Although it can be argued thatsurveys conducted after 1900 were of higher quality thanoriginal mapping operations in the 1840’s, an absolutedifference can not be quantified. Consequently, theparameters outlined above are assumed constant for allfield surveys and provide a conservative estimate ofpotential errors. For the 1857/70 and 1924 T-sheets,digitizer setup recorded an average percent deviation of0.02, or 4 m ground distance at a 1:20,000 scale. Linethickness, due to original production and photo-reproduc-tion, was no greater than 0.3 mm, or 6 m ground distancefor this same scale.

(d) A primary consideration with aerial surveys isthe interpreted high-water shoreline position. Becausedelineation of this feature is done remotely, the potentialfor error is much greater than field surveys and is a func-tion of geologic control and coastal processes. Dolanet al. (1980) indicated that average high-water line move-ment over a tidal cycle is about 1 to 2 m along the mid-Atlantic coast; however, accurate delineation of the line issometimes difficult due to field conditions, knowledge ofhuman impacts, and photographic quality. Although themagnitude of error associated with locating the high-waterline is unknown, on gently sloping beaches with largetidal ranges (i.e. Sea Islands, Georgia/South Carolina),significant horizontal displacement can occur with a smallincrease in elevation.

(e) For H-sheets, a topographical survey of the coastwas often conducted before the bathymetric survey. Thecontrol points established along the shoreline were thenused for positioning of the survey vessel offshore. Due tothe nature of triangulating distances and angles frompoints on land, horizontal positions plotted for the vesselbecame less accurate as it moved away from shore.

5-61

EM 1110-2-181031 Jan 95

When the vessel was out of sight of the triangulationpoints along the coast, positioning was done by deadreckoning. Therefore, horizontal positions of some off-shore soundings on early H-sheets may be suspect.

(7) Digitizer limitations. Another source of errorrelates to equipment and operator accuracy and precision.As stated earlier, the absolute accuracy (accuracy andprecision) of the digitizing tables used for this study is0.1 mm (0.004 in). At a scale of 1:10,000, this convertsto ± 1 m. Furthermore, the precision with which an oper-ator can visualize and move the cursor along a line canlead to much greater errors (Tanner 1978). To evaluatethe magnitude of operator error associated with digitizingshoreline position, at least three repetitive measurementsshould be compared.

(8) Analysis of shoreline change data.

(a) In most instances, data pairs are generated fromshoreline locations relative to some arbitrary axis system.A comparison of these data pairs is used to calculatemean shoreline movements, variations in the rate anddirection of movements, and maximum net movements(Anders, Reed, and Meisburger 1990). Generally thecoastline is divided into segments based on the generalorientation of the shoreline, as shown in Figure 5-27.Baselines should be chosen based on segments that areparallel to the shoreline. Usually a standard Cartesiancoordinate system is assigned to each segment with thepositive x-axis directed generally north to south and thepositive y-axis lying orthogonally seaward. The resultingdata pairs include the x-value and the y-value, whichrepresents the perpendicular transect.

(b) Three primary statistics are generally used forshoreline change computations. They include the samplemean, sample standard deviation, and maximum shorelinemovement. The sample mean is defined as a measure ofcentral tendency for a set of sample observations and isexpressed as follows:

(5-1)x

n

i 1

xi

n

where xi = sample observations fori = 1 to n and n =total number of observations. The sample standard devia-tion s is a measure of sample variability about the mean.

Figure 5-27. Coastline is divided into segments basedon the general orientation of the shoreline

(5-2)

s

n

i 1

(xi x)2

n 1

The maximum shoreline movement represents the differ-ence in the most landward and seaward position. It alsorepresents the end points for shoreline change inclusive ofall the data sets. Identifying areas of maximum shorelinemovement is useful with beach fill projects.

(c) Comparisons of calculated shoreline change ratesare generally grouped by specific time periods or byalongshore segments (i.e. geomorphic features represent-ing spatial trends). A case example, shown in Fig-ure 5-28, of distinctive spatial shoreline trends is locatedin northern New Jersey, where the shoreline is part of abarrier spit complex including an active compound spit(Sandy Hook, New Jersey to Sea Bright), barrier penin-sula (Sea Bright to Monmouth Beach), and a headlandcoastline (Monmouth Beach to Shark River Inlet)(Gorman and Reed 1989).

(9) Interpretation of shoreline change.

5-62

EM 1110-2-181031 Jan 95

Figure 5-28. Distinctive spatial shoreline trends along the northern New Jersey shore

5-63

EM 1110-2-181031 Jan 95

(a) Historical shoreline positions have been recog-nized as a primary data source for quantifying rates oferosion and accretion. Coastal scientists, engineers, andplanners often use this information for computing rates ofshoreline movement for shore protection projects, deter-mining a project rate of retreat for shore protection, esti-mating the magnitude and direction of sediment transport,monitoring engineering modifications to a beach, examin-ing geomorphic variations in the coastal zone, establishingcoastal erosion set-back lines, and verifying numericalshoreline change models.

(b) Relevant published studies that quantify shorelinemovement for key U.S. coastlines are listed in Table 5-14.Usually, the alongshore shoreline is subdivided based ongeomorphic features or human modifications (Anders,Reed, and Meisburger 1990; Gorman and Reed 1989).Another criterion used is to identify end points of shore-line segments where little or no net change was measured(Knowles and Gorman 1991). Recently, a blocking tech-nique was used by Byrnes and Hiland (1994) to evaluatethe spatial shoreline trends based on areas with similardirection of change producing variable-length shorelinecells.

f. Beach and nearshore profiles.

(1) Background. Evaluation of continuous andrepeated beach and nearshore profiles documents theentire active profile envelope and provides a completepicture of the response of the profile to coastal processes.Because storms are an important factor in coastal sedi-mentary processes, it is important to assess profilechanges after major storm events. Collecting field data assoon as possible after storms and comparing these profilesto the most recent pre-storm ones provides a measure ofareas of erosion and accretion and the volume changesthat have occurred.

(2) Accuracy criteria.

(a) Elevation resolution for a typical project profileis estimated to be about 0.012 m at a maximum range.

(b) As described in section 5-3, water bodies areoften surveyed by a sled which is towed by boat out intothe water from about +1.5 m to closure depth. Thisresults in overlap between the onshore rod survey and thesled survey to assure that the two systems are recordingthe same elevations. If offshore surveys are conducted byboat-mounted echosounder, overlap with the rod survey isusually not possible.

(c) Comparison of sled/Zeiss systems and boat echo-sounder systems has shown sled surveys to have a highervertical and horizontal accuracy (Clausner, Birkemeier,and Clark 1986). Echosounder surveys are limited by theindirect (acoustic) nature of the depth measurement, theeffects of water level variations and boat motions, and theinability to survey the surf zone due to wave action andtidal range. In summary, there are quality advantages inusing sled surveys offshore, but operational limitations areimposed by wave heights, water depth, seafloor obstruc-tions, and the maneuvering needed to keep the sled online.

(d) All profile surveys must be referenced to thesame elevation datum. This can especially be a problemwhen echosounder surveys are conducted by differentagencies or contractors over time (Figure 5-29). Meticu-lous field notes must be kept to record datums, correc-tions, equipment calibrations, and other information that isneeded for accurate data reduction.

(3) Analysis techniques.

Table 5-14Selected Shoreline Movement Studies

Author Location Method

Byrnes and Hiland 1994 Cumberland-Amelia Islands,Georgia/Florida

Georeferenced in GIS

Gorman and Reed 1989 Northern New Jersey Cartographic techniques, map overlays

Anders, Reed, and Meisburger 1990 South Carolina Cartographic techniques, map overlays

McBride et al. 1991 Louisiana Georeferenced in GIS

Morton 1979 Texas Map overlays

Everts, Battley, and Gibson 1983 Cape Henry - Cape Hatteras Map overlays

Leatherman 1984 Maryland Metric mapping

5-64

EM 1110-2-181031 Jan 95

Figure 5-29. Example of vertical offset between two offshore profile surveys due to use of different datums

(a) Profile analysis reveals the variability in cross-shore elevation patterns and volume change that occuralong a profile line. With multiple profiles, the along-shore variability in profile response is documented. Witha long-term monitoring program, seasonal variations andthe impact of storms are identified.

(b) Profile data recorded in the field are typicallyprocessed in the laboratory using computer software pack-ages. The CERC Interactive Survey Reduction Program(ISRP) plots and compares both spatial and temporalprofile sets (Birkemeier 1984). The program allows theplotting of field data sets at various scales and verticalexaggerations from baseline (X) and elevation (Y). Anunlimited number of profiles can be plotted on a singleaxis to compare profile change and determine profileenvelopes and closure areas. The most frequent analysisuses profiles of successive dates to compare morphologyand volume changes. CERC’s Beach Morphology andAnalysis Package (BMAP) contains many analysis tools,including generation of synthetic profiles (Sommerfeldet al. 1994).

(c) Vertical elevations of important morphologicfeatures found on profiles are usually referenced to

NGVD or another datum specified for a particular project.All horizontal distances should be measured from thedesignated baseline position, preferably located behind theprimary dunes for safety (i.e., survival during majorstorms). Volume change calculations can be made fromthe baseline to a common distance offshore (usually theshortest profile) to normalize volume change betweensurvey dates. Profile volume calculations should be basedon the minimum distance value. Survey distances off-shore often vary in length due to the wave conditions atthe time of the sled survey.

(4) Profile survey applications.

(a) General. Interpretation of beach response tocoastal processes can be done with geometric and volu-metric comparison of beach profile sets. If the profilesets cover a long period, information on both the cross-shore and alongshore evolution of a coastline can be made(i.e., dunes versus seawalls, position of the berm crest,and closure depth). Several types of beach parameterscan be measured from profile data, including the width ofthe subaerial beach, location and depth of the inner bar,and beach and nearshore profile slope. Comparisonsbetween successive profiles can be used to quantify

5-65

EM 1110-2-181031 Jan 95

shoreline position change, volumetric change, and sea-sonal profile response. Numerous studies (Hands 1976,Wright and Short 1983) have documented the cyclicnature of beach topography in response to seasonal shiftsin the local wind and wave climate. In addition to normaleffects, profile surveys can also be used to measurechange caused by short-term episodic events (Chiu 1977;Savage and Birkemeier 1987).

(b) Linear Measurements. Selected parameters can beused to define cross-shore morphologic features within astudy area. General location and limits of features in thebeach and nearshore zone used for linear profile computa-tions are shown in Figure 5-30.

• The most variable beach parameter is beachwidth, which is usually measured between thebase of the dune and mean low water (mlw).

• Beach slope can be calculated between the baseof the dune and mlw.

• The zone from mlw out to the nearshore slopebreak is generally considered as the area wherethe nearshore slope is computed.

• Alongshore changes of the inner bar position area useful guide of the surf zone breaker height andbottom slope. The inner bar position is measuredfrom 0.0 m (NGVD) to the bar crest (Hands1976; Gorman et al. 1994).

• If shoreline change or aerial photography mapsare not available, shoreline position can be esti-mated from the location of a specified elevationpoint on a profile line. An approximate positionof the high-water shoreline should be selectedbased on local tidal information. A commonelevation referenced for this type of analysis is0.0 (NGVD) (USAE District, Jacksonville 1993).However, this position constitutes a highly vari-able measure due to the movement of the bar orridge and runnel features along the lower beach.

Figure 5-30. Features within the beach and nearshore zone used for linear profile computations

5-66

EM 1110-2-181031 Jan 95

(c) Volumetric analysis. Volume analysis of mostlong-term profile data sets will provide temporal andspatial documentation of profile volume change due tooverwash processes, storm impacts, and nearshore barevolution. Computer programs such as ISRP can providequantitative information on profile shape change andvolume of sediment gained or lost between two or moresurvey dates (Birkemeier 1984). Figure 5-31 shows ananalysis of the Ocean City, Maryland, beach fill project.Based on volume computations, this type of analysis pro-vided a time history of fill placed on the beach and thesubsequent readjustment of the fill material. Typicalprofile response showed erosion on the dry beach aboveNGVD and accretion in the nearshore area after fill place-ment as the shoreface adjusted to a new equilibriumprofile.

(d) Seasonality. Winter erosional beach profiles canbe characterized as having concave foreshore areas and awell-developed bar/trough in the nearshore. During fair-weather summer conditions, the bar moves landward andwelds onto the foreshore, producing a wider berm with alower offshore bar and flatter trough. Profile response tothe seasonal cycle is a function of storm frequency andintensity. When trying to determine the extent of theprofile envelope, at least 1 year of data should be used.The profile envelope of an East Coast beach system isshown in Figure 4-29, with the characteristic winter andsummer berm profiles. Because there are frequent localstorm surges during the winter months, the berm and dunecrest often retreat; however, in most areas sand recoverytakes place during the summer months as littoral materialmoves onshore and longshore. Along a well-defined ridgeand runnel system, significant sediment exchange canoccur between the summer and winter months(Figure 3-21).

Great Lakes beaches also display summer/winter patterns,often characterized by considerable bar movement (Fig-ure 4-30). At some Great Lakes sites, the mobile sandlayer is quite thin and seasonal patterns can be difficult todetect (Figure 5-32).

g. Bathymetric data.

(1) Introduction. Analysis and examination oftopographic and bathymetric data are fundamental inmany studies of coastal engineering and geology. Whenassembling bathymetric surveys from a coastal area, aresearcher is often confronted with an immense amount ofdata which must be sorted, checked for errors, redisplayedat a common scale, and compared year by year or surveyby survey in order to detect whether changes in bottom

topography have occurred. This section will discuss threegeneral aspects of geographic data analysis:

• Processing of bathymetric data using mappingsoftware.

• Applications and display of the processed results.

• Error analyses.

(2) Bathymetric data processing - data preparationand input.

(a) Most historical bathymetric data sets consist ofpaper maps with printed or handwritten depth notations(Figure 5-33). Occasionally, these data are available onmagnetic media from agencies like NOAA, but often areseacher must first digitize the maps in order to be ableto perform computer-based processing and plotting. Ifonly a very limited region is being examined, it may bemore expedient to contour the charts by hand. The disad-vantage of hand-contouring is that it is a subjective proce-dure. Therefore, one person should be responsible for allof the contouring to minimize variations caused by differ-ent drawing styles or methods of smoothing topographicvariations.

(b) In order to be able to manipulate 3-dimensional(X, Y, and Z) data, display and plot it at different scales,and compare different data sets, it is necessary to use oneof the commercial mapping programs such as GeoQuestCorporation’s Contour Plotting System 3 (CPS-3) orGolden Software’s Surfer. These are comprehensivepackages of file manipulation, mapping algorithms, con-touring, and 2- and 3-dimensional display. Their userequires considerable training, but they are powerful anal-ysis tools.

(c) The raw data used by mapping programs consistsof data in X-Y-Z form. As described in the previoussection, if the data are derived from old maps, they mustfirst be corrected to a common datum, map projection,and coordinate system. For small files, visual examina-tion of the data may be worthwhile in order to inspect forobviously incorrect values. Because it is laborious toreview thousands of data points, simple programs can bewritten to check the raw data. For example, if all thedepths in an area are expected to be between +2.0 and-12.0 m, the program can tag depths that are outside thisrange. The analyst can then determine if questionablepoints are erroneous or represent genuine but unexpectedtopography. The X and Y points should typically

5-67

EM 1110-2-181031 Jan 95

Figure 5-31. Analysis of the Ocean City, MD, beach fill project. Upper plot shows profile before beach fill and thelarge quantity of sand placed on the beach during the summer of 1988. Lower plot shows erosion of the upperprofile during a storm in early 1989. Sand from the beach moved offshore to the region between 400 and 900 ftfrom the benchmark

5-68

EM 1110-2-181031 Jan 95

Figure 5-32. Profile envelope from St. Joseph, Michigan. The horizontal platform from 1,000 to 2,500 ft offshore isan exposed till surface. Most shoreface sand movement appears to be confined to the zone landward of the tillplatform, although it is likely that thin veneers of sand periodically cover the till (previously unpublished CERCdata)

represent Cartesian coordinates, which is the case if theoriginal maps were based on State Plane coordinates.X and Y points which are latitude and longitude must beconverted by the program.

(3) Gridding operations.

(a) Gridding is a mathematical process in which acontinuous surface is computed from a set of randomlydistributed X, Y, and Z data.1 The result is a data struc-ture (usually a surface) called a grid. Note that the grid isan artificial structure. It is based on the original data, butthe grid points are not identical to the original surveypoints (Figures 5-34 and 5-35). Because the grid repre-sents the surface that is being modeled, the accuracy of

________________________________1 Examples in this section were prepared using CPS-3mapping software from GeoQuest Corporation. However,the overall concepts and procedures discussed are general,and other software packages perform similar functions.

the grid directly affects the quality of any output based onit or on comparisons with other grids generated fromother data sets. Computing a grid is necessary beforeoperations such as contouring, volume calculation, profilegeneration, or volume comparison can be performed. Theadvantage of a grid is that it allows the program tomanipulate the surface at any scale or orientation. Forexample, profiles can be generated across a channel evenif the original survey lines were not run in these direc-tions. In addition, profiles from subsequent surveys canbe directly compared, even if the survey track lines werevery different.

(b) Several steps must be considered as part of thegrid generation. These include:

• Selecting a gridding algorithm.

• Identifying the input data.

• Specifying the limits of the grid coverage.

5-69

EM 1110-2-181031 Jan 95

Figure 5-33. Example of a hand-annotated hydrographic map from a Florida project site. The depths have beencorrected for tide and are referenced to mlw. (Map courtesy of USAE District, Mobile)

• Specifying gridding parameters.

• Specifying gridding constraints.

• Computing the grid.

The choice of a gridding algorithm can have a majoreffect on the ultimate appearance of the grid. Softwarecompanies have proprietary algorithms which they claimare universally superior. Often, however, the type ordistribution of data determines which procedure to use,and some trial and error is necessary at the beginning of aproject. Because a computed grid is an artificial structure,often it is a subjective evaluation whether one grid is“better” than another. For subaerial topography, anoblique aerial photograph can be compared with acomputer-generated 3-dimensional drawing oriented at thesame azimuth and angle. But for a subaqueous seafloor,how can a researcher really state that one surface does not

look right while another does? Even comparing a griddedsurface with a hand-contured chart is not a valid testbecause hand-contouring is a very subjective procedure.

(4) The fundamental challenge of a gridding algo-rithm is to estimate depth values in regions of sparse data.The procedure must attempt to create a surface whichfollows the trend of the terrain as demonstrated in theareas where data do exist. In effect, this is similar to thetrend-estimating that a human performs when he contoursbathymetric data by hand. The other challenge occurs incomplex, densely sampled terrains. The algorithm mustfit the surface over many points, but genuine topographicrelief must not be smoothed away! Along a rocky coast,for example, high pinnacles may indeed project above thesurrounding seafloor.

(a) Gridding algorithms include:

5-70

EM 1110-2-181031 Jan 95

Figure 5-34. Digitally collected hydrographic data from a Florida project site. The track lines are obvious, as is thefact that the soundings are not uniformly distributed throughout the survey area. (Data courtesy of USAE District,Mobile)

5-71

EM 1110-2-181031 Jan 95

Figure 5-35. Surface grid computed by CPS-3 based on the data shown in Figure 5-31. The nodes are uniformlyspaced compared with the locations of the original soundings. A grid does not necessarily have to be square,although this is common

5-72

EM 1110-2-181031 Jan 95

• Convergent (multi-snap) (CPS-3 software).

• Least squares with smoothing.

• Moving average.

• Trend.

• Polynomial.

The convergent procedure often works well for bathyme-tric data. It uses multiple data points as controls forcalculating the values at nearby nodes. The values areblended with a distance-weighting technique such thatclose points have more influence over the node than dis-tant points. Several iterations are made, with the firstbeing crude and including many points, and the finalbeing confined to the closest points. The least-squaresmethod produces a plane that fits across several pointsnear the node. Once the plane has been calculated, theZ-value at the node is easily computed. The reader mustconsult software manuals to learn the intricacies of howthese and other algorithms have been implemented.

(b) Another important parameter that must be chosenis the gridding increment. This is partly determined bythe algorithm chosen and also by the data spacing. Forexample, if survey lines are far apart, there is little pur-pose in specifying closely spaced nodes because of thelow confidence that can be assigned to the nodes locatedfar from soundings. In contrast, when the original dataare closely spaced, large X- and Y-increments result in anartificially smoothed surface because too many data pointsinfluence each node. Some programs can automaticallycalculate an increment that often produces good results.

(4) Applications and display of gridded data.

(a) Contouring of an area is one of the most commonapplications of mapping software (Figure 5-36). Not onlyis this faster than hand-contouring, but the results areuniform in style across the area and precision (i.e. repeat-ability) is vastly superior.

(b) The power of mapping programs is best demon-strated when analyzing different surveys. If at all pos-sible, the different data sets should be gridded with thesame algorithms and parameters in order that the resultsbe as comparable as possible. Difficulty arises if earliersurveys contain much sparser data than later ones. Underthese circumstances, it is probably best if the optimumgrid is chosen for each data set. A simple application is

to plot a suitable contour to demonstrate the growth overtime of a feature like a shoal (Figure 5-37). Computationof volumetric changes over time is another application(Figure 5-38). This can graphically demonstrate howshoals develop or channels migrate.

(c) Volumetric data can be used to estimate growthrates of features like shoals. As an example, using all18 of the 1,000-ft squares shown in Figure 5-37, theoverall change in volume of the East Pass ebb-tidal shoalbetween 1967 and 1990 was only 19 percent (Fig-ure 5-39). Although the shoal had clearly grown to thesouthwest, the minor overall increase in volume suggeststhat considerable sand may have eroded from the innerportions of the shoal. In contrast, when plotting thechange in volume of nine selected squares, the growthover time was 600 percent. This underscores how criti-cally numerical values such as growth rates depend uponthe boundaries of the areas used in the calculations. Theuser of secondary data beware!

(5) Error analysis of gridded bathymetry.

(a) A crucial question is how much confidence can aresearcher place on growth rates that are based on bathy-metric or topographic data? Unfortunately, in the pastmany researchers ignored or conveniently overlooked thepossibility that error bars may have been greater thancalculated trends, particularly if volumetric computationswere based on data of questionable quality.

(b) This section outlines a basic procedure that canbe used to calculate volumetric errors provided that esti-mates of the vertical (∆Z) accuracy are available. If∆Zvalues are unavailable for the specific surveys, standarderrors of ± 0.5, ± 1.0, or ± 1.5 ft, based on the class ofthe survey, can be used (Table 5-4). For coastal surveysclose to shore, this method assumes that errors in posi-tioning (∆X and ∆Y) are random and have insignificanteffect on the volumes compared with possible systematicerrors in water depth measurements, tide correction, anddata reduction. For older historic surveys, positioningerror may be important, requiring a much more comp-licated analysis procedure. Positioning accuracy of hydro-graphic surveys is discussed in EM 1110-2-1003 andNOAA (1976).

(c) The error in volumetric difference between sur-veys can be estimated by determining how much theaverage depth in each polygon changes from one surveyto another and then calculating an average depth changeover all polygons. Maximum likely error (MLE) is:

5-73

EM 1110-2-181031 Jan 95

Figure 5-36. Contoured bathymetry of the same area shown in Figures 5-34 and 5-35. Depths in feet below mlw

5-74

EM 1110-2-181031 Jan 95

Figure 5-37. Overall growth of an ebb-tidal shoal over 24 years is shown by the advance of the 15-ft isobath. Thisisobath was chosen because it represented approximately the mid-depth of the bar front. The 1000-ft squares arepolygons used for volumetric computations (Morang 1992a)

5-75

EM 1110-2-181031 Jan 95

Figure 5-38. Isopach map showing overall changes in bottom configuration between 1967 and 1990 at East Pass,Florida. Red contours represent erosion, while green represent deposition (both colors at 2-ft interval). The blackcontour line represents the zero line (no erosion or deposition). The migration of the channel thalweg to the east isobvious, as is the growth of scour holes at the jetties. Map computed by subtracting June 1967 surface fromFebruary 1990 surface. (Morang 1992a)

5-76

EM 1110-2-181031 Jan 95

Figure 5-39. Growth of the ebb-tidal shoal at East Pass, FL. Areas used in the computations are shown inFigure 5-37. Growth rates are dramatically different depending upon which polygons are included in the volumetriccomputations

2 × ∆Z∆Zave

For example, if∆Z = 0.15 m and∆Zave = 1.0 m, thenMLE is:

0.30 m1.00 m

0.30 30 percent

Note that this is for a Class 1 survey; many offshoresurveys are not conducted under such tight specifications.If ∆Z = 0.46 m for Class 3, then MLE for the above

example = 91 percent. Under these circumstances, itbecomes meaningless to say that an area has changed involume by a certain amount ± 91 percent.

(d) The size of the polygons used in the calculationof ∆Zave can influence the MLE. A particular polygonthat covers a large area may average∆Z of only 0.3 or0.6 m, but water depths from spot to spot within the poly-gon may vary considerably more. Therefore, by usingsmaller polygons,∆Z will typically be greater and MLEcorrespondingly less. However, the use of smaller poly-gons must be balanced against the fact that positioningerrors (∆X and ∆Y) become correspondingly moresignificant.

5-77

EM 1110-2-181031 Jan 95

(e) More research is needed to quantify errors asso-ciated with various types of offshore surveys and to iden-tify how these errors are passed through computedquantities. They mustnot be neglected when analyzinggeologic data, particularly if management or policy deci-sions will be based on perceived trends.

h. Sediment grain size analyses.

(1) Introduction1.

(a) The coastal zone is comprised of many dynamicmorphologic features that frequently change their formand sediment distribution. Although a beach can displaya large range of sizes and shapes, each beach is character-ized by particular texture and composition representingthe available sediment (Davis 1985). Textural trendsalongshore and cross-shore are indicative of the deposi-tional energy and the stability (or instability) of the fore-shore and nearshore zones.

(b) Because of natural variability in grain size distri-butions, a sampling scheme should adequately sample thenative beach in both the cross-shore and alongshore direc-tions. Sediment sampling needs to coincide with surveyprofile lines so that the samples can be spatially locatedand related to morphology and hydrodynamic zones.Consideration of shoreline variability and engineeringstructures should be factored into choosing samplinglocations. A suggested rule of thumb is that a samplingline be spaced every half mile, but engineering judgmentis required to define adequate project coverage. On eachline, it is recommended that samples be collected at allmajor changes in morphology along the profile, such asdune base, mid-berm, mean high water, mid-tide, meanlow water, trough, bar crest, and then 3-m intervals todepth of closure (Figure 5-40) (Stauble and Hoel 1986).

(2) Grain size analysis statistics.

(a) Sediments should be sieved using U.S. Standardsieves at 1/4-phi (φ) unit intervals. Phi (φ) is defined asthe negative logarithm of the grain dimension in millime-ters to the base 2. The equation for the relationship ofmillimeters to phi scale is:

(5-3)φ log2(dmm)

where

dmm = particle diameter in millimeters

_____________________________1 Text condensed from Gorman et al. (in preparation)

Figure 5-40. Recommended sampling locations at atypical profile line

(b) Grain-size analyses should include grain-sizedistribution tables, statistics and graphics of frequency,cumulative frequency and probability distribution (see“Calculation of Composite Grain Size Distribution” in theAutomated Coastal Engineering System(ACES) ofLeenknecht, Szuwalski, and Sherlock (1992) and ASTMStandard D 2487-92. Standard grain-size distributionstatistics include:

• Median grain size or d50 - the particle size in thecenter of the population.

• Mean grain size or average grain size.

• Standard deviation or the spread of the distribu-tion about the mean - defines the concept ofsorting.

• Skewness or measure of symmetry of the distribu-tion around the mean.

• Kurtosis or measure of the peakedness of the fre-quency distribution.

Each of these statistical parameters provides informationon the grain-size distribution and its depositional environ-ment. The mean is the most commonly used statistic tocharacterize the average grain size of the distribution.The median value can be read directly off a cumulativecurve and is near-normal to the mean in a normal distri-bution but differs if the distribution is non-normal. Thesorting gives the spread of the various grain sizes in thedistribution. A well-sorted distributioncontains a limitedrange of grain sizes and usually indicates that the deposi-tional environment contains a narrow range of sedimentsizes or a narrow band of depositional energy. Apoorlysorted distributioncontains a wide range of grain sizes

5-78

EM 1110-2-181031 Jan 95

indicating multiple sources of sediment or a wide range ofenergies of deposition. Positive skewnessindicates anexcess of fine grain sizes, whereasnegative skewnessindicates an excess of coarser grain sizes. Thekurtosismeasures the ratio between the sorting in the tails of thedistribution relative to the central portion (sand size) ofthe distribution.

(c) These statistical parameters are commonly calcu-lated by two different methods. Thegraphic methodusesspecific percentiles of a grain-size distribution (i.e., 5, 16,25, 50, 75, 84, and 95) that are read from graphical dataplots (Folk 1974) or can be calculated from sieve data.The values are used in simple equations to produce theapproximate statistical parameters. Phi values are used tocalculate these parameters, and only the mean and medianshould be converted to millimeter values. Themethod ofmomentsuses the entire grain-size distribution values tomathematically produce the statistical parameters (Fried-man and Sanders 1978). This procedure is more accurate,but was time-consuming to calculate before the use ofcomputers; for this reason, older sediment statistical dataare commonly based on the Folk graphic method. Addi-tional consideration for the user of grain size statistics arelisted below:

• The graphical and moment methods arenotdirectly comparable. Because sediment statisticsfor many projects have historically been calcu-lated by the graphic method, for uniformity itmay be best to continue using the graphicmethod.

• The graphic and moment procedures have advan-tages and disadvantages. These are summarizedin Table 5-15.

• Note that calculated statistical parameters are onlyan indication of the characteristics of the sedimentin the field. The user must not assume that thewhole population has exactly these characteristics.

• Accurate sediment grain-size statistics are depen-dent on adequate sample size. Recommendationsfor field sampling have been listed in Table 5-9.

The following sections list equations and provide verbaldescription of sediment grain-size parameters for both thegraphic method and the method of moments. The equa-tions are identical to those used in the USACE ACESsoftware (Leenknecht, Szuwalski, and Sherlock 1992).

(d) Mean grain size. Table 5-16 lists formulas anddescriptive criteria for classifying the mean grain size of asample.

(e) Standard deviation (sorting). The standard devia-tion or measure of sorting uses the equations and verbaldescriptors listed in Table 5-17.

(f) Skewness. The skewness or measure of symme-try shows excess fine or coarse material in the grain-sizedistribution. Table 5-18 lists equations used for thegraphic method and method of moments, with the rangeof verbal descriptors.

(g) Kurtosis. Kurtosis or measure of the peakednessof the grain-size distribution relates sorting of the tailscompared to sorting of the central portion of the distribu-tion. The equations listed in Table 5-19 are used for thegraphic method, which centers around graphic kurtosisKG = 1.00, and the method of moments, which centersaround the moment kurtosis k = 3.00. The range of

Table 5-15Comparison of graphic and moment procedures for calculating grain-size statistics

Method Advantages Disadvantages

Graphic Can be calculated from almost all distribution data Does not use all data from all sieves

Resistant to sampling and laboratory errors (i.e., a singlefaulty sieve does not invalidate the calculated statistics)

Can use open-ended samples (more than 5 percent of sampleweight on either tail)

Moment Uses formula that has a greater number of parameters Parameters have to be established in laboratory

Uses data from all sieves Parameters should be important to the application;otherwise may be more than needed or useful

Open-ended distributions (more than 5 percent ofsample in either tail) must be excluded, thereforelosing the geologic information that these samplesmight reveal

5-79

EM 1110-2-181031 Jan 95

Table 5-16Mean Grain Size

Graphic Mean, M:

(5-6)M φ16 φ50 φ843

Where: φn = grain size of nth weight percentile in phi units

Moment Mean, x:

(5-7)xf mφ

100

Where: f = frequency weight percentmφ = midpoint of size class

Descriptive Criteria:

Grain size (mm) Grain size (Phi) Wentworth Classification

1.00 - 2.00 0.0 - -1.0 Very Coarse Sand

0.50 - 1.00 1.0 - 0.0 Coarse Sand

0.25 - 0.50 2.0 - 1.0 Medium Sand

0.125 - 0.25 3.0 - 2.0 Fine Sand

0.0625 - 0.125 4.0 - 3.0 Very Fine Sand

Table 5-17Sample Standard Deviation (Sorting)

Graphic Sorting, σ:

(5-8)σ φ84 φ164

φ95 φ56.6

Moment Sorting, σ:

(5-9)σ

f (mφ x)2

100

12

Descriptive Criteria:

Sorting Range (Phi) Description of Sorting

<0.35 Very well sorted

0.35 - 0.50 Well sorted

0.50 - 0.71 Moderately well sorted

0.71 - 1.00 Moderately sorted

1.00 - 2.00 Poorly sorted

2.00 - 4.00 Very poorly sorted

> 4.00 Extremely poorly sorted

5-80

EM 1110-2-181031 Jan 95

Table 5-18Sample Skewness

Graphic Skewness, Sk:

(5-10)Sk φ16 φ84 2(φ50)2(φ84 φ16)

φ5 φ95 2(φ50)2(φ95 φ5)

Moment Skewness, Sk:

(5-11)Skf (mφ x)3

100 σ3

Descriptive Criteria:

Skewness Range Description of Skewness

+1.0 to +0.3 Very fine-skewed

+0.3 to +0.1 Fine-skewed

+0.1 to −0.1 Near-symmetrical

−0.1 to −0.3 Coarse-skewed

−0.3 to −1.0 Very coarse-skewed

Table 5-19Sample Kurtosis

Graphic Kurtosis, KG:

(5-12)KGφ95 φ5

2.44 (φ75 φ25)

Descriptive Criteria for Graphic Method:

Graphic Kurtosis Range Description of Kurtosis

< 0.67 Very platykurtic (flat)

0.65 to 0.90 Platykurtic

0.90 to 1.11 Mesokurtic (normal distribution)

1.11 to 1.50 Leptokurtic

1.50 to 3.00 Very leptokurtic

> 3.00 Extremely leptokurtic (peaked)

Moment Kurtosis, k:

(5-13)kf (mφ x)4

100 σ4

Descriptive Criteria for Moment Method:

Moment Kurtosis Range Description of Kurtosis

< 3.00 Platykurtic (flat)

Around 3.00 Mesokurtic (normal distribution)

> 3.00 Leptokurtic

5-81

EM 1110-2-181031 Jan 95

verbal descriptors of peakedness is based on the platy-kurtic (flat) curve versus the leptokurtic (peaked) curve,with a mesokurtic curve as normal.

(3) Composite sediments1. Combining samples fromacross the beach can reduce the high variability in spatialgrain size distributions on beaches (Hobson 1977). Com-posite samples are created by either physically combiningseveral samples before sieving or by mathematically com-bining the individual sample weights to create a newcomposite sample on which statistical values can be cal-culated and sediment distribution curves generated. Sam-ples collected along profile sub-environments can becombined into composite groups of similar depositionalenergy levels and processes as seen in Figure 5-41. Inter-tidal and subaerial beach samples have been found to bethe most usable composites to characterize the beach andnearshore environment area. After comparing severalcomposite groups, Stauble and Hoel (1986) found that acomposite containing the mean high water, mid-tide, andmean low water gave the best representation of the_____________________________1 Text adapted from Stauble (1994).

foreshore beach. They found that nearshore sample com-posite sediment distributions changed little over time.This suggests that active sorting and sediment transportoccur on the active beach face and bar area and that near-shore sands remain uniform over time.

(4) Seasonal variability. There can be a wide vari-ability in grain size distribution on a native beach betweenwinter high wave periods and summer fair weather peri-ods. This variability can be a problem in choosing arepresentative native beach. The winter grain size distri-bution usually is coarser and more poorly sorted than thesummer distribution (due to the higher frequency ofstorms in the winter). The concept of the seasonal beachcycle is based on the frequency of storm-induced erosionand fair weather accretion. Extreme events, such as hurri-canes that occur in the summer or early fall, as well asmild winters with few extratropical storms, may causeperturbations on the seasonal cycle. A sampling strategyto characterize the seasonal variability should take intoaccount the recent local storm climate.

Figure 5-41. Combination of samples into composite groups of similar depositional energy levels and processes(example from CERC Field Research Facility, Duck, NC)

5-82

EM 1110-2-181031 Jan 95

(5) Sediment data interpretation.

(a) Grain size distributions of beach sediment varywith both time and space. Because of the daily wave andtidal influence on sediment deposition on the beach fore-shore, swash processes create an ever-changing foreshoresediment distribution. The use of composites helps tosimplify the analysis and interpretation of these changes.The bar/trough area also experiences a wide variety ofenergy conditions and thus displays a variety of grain sizedistributions over time. Dune and nearshore grain sizedistributions have less variability due to the lower energyconditions that affect these areas. The dune is primarilyinfluenced by wind transport, which limits change to thefiner grain sizes except under extreme wave conditions,when the waves actually impact on the dune. The near-shore zone is dependant on regional and local coastalprocesses.

(b) An example of composite grain size distributioncurves for the beach at the Field Research Facility, Duck,NC, is shown in Figure 5-42. Using the entire distribu-tion from coarse to fine sizes shows the changes in sizeclasses for various depositional regimes. The beach groupcomposite is illustrated because it displayed the greatestvariability in distribution during the study. The bimodalnature of the distribution can be seen, with increases inthe coarse mode fraction after storms or when samples onthe foreshore contained granule-size lag deposits. Thecoarsest material was present early in the study periodduring the winter storm period. Later, the distributionshifted to the finer mode except during July, 1985, whena coarse fraction was present. Swash processes of uprushand backwash are the principal transport mechanism inthis area.

(c) Spatial variation along a beach is more complex.Analysis of grain size data from six profiles at OceanCity, MD, shows the influence of beach fill placementand storm processes. Figure 5-43 shows the change inmean grain size of the foreshore composites (high tide,mid-tide and low tide samples) for the six profiles locatedalong the central section of the beach fill project.Between the pre- and post-fill sampling, the meansbecame finer and the volume of the profile increased onfive of the six locations as the fill was placed. Stormprocesses caused the foreshore means to become coarser,but a return to finer foreshore mean was found with stormrecovery. From these studies, a general trend to coarser(and more poorly sorted) sediment grain size distributionoccurred after high wave conditions. High wave powervalues and, to a lesser extent, wave steepness values

correlated with times when the means became coarse.The shift to finer means occurred as the wave parametersdecreased.

i. Coastal data display and analysis usingGeographic Information Systems.

(1) Definitions.

(a) Geographic Information Systems - known as GIS- are information-oriented computerized methods designedto capture, store, correct, manage, analyze, and displayspatial and non-spatial geographic data (Davis and Schultz1990). Using computer-based technology, GIS methodol-ogy has revolutionized many of the traditional, manualmethods of cartographic analysis and display. GIS is anoutgrowth of many existing technologies: cartography,spatial analysis, remote sensing, computer mapping, anddigital database management.

(b) GIS is based on the manipulation of spatial data.The termspatial data refers to “any data or informationthat can be located or tied to a location, regardless of theoriginal form (tabular, map, image, or some other form).Essentially, spatial data possess attributes or characteris-tics that are linked to location.” (Davis and Schultz 1990).

(2) Components.

(a) Many of the data manipulation operationsrequired to analyze bathymetric data, aerial photographs,and historic maps can be accomplished with GIS. Fivemajor components of GIS include:

• Geographic database.

• Software.

• Hardware.

• User interface.

• Support for equipment and structure (people,organization, training).

(b) The major functions of GIS include:

• Collection.

• Storage.

• Retrieval.

5-83

EM 1110-2-181031 Jan 95

Figure 5-42. Plots of beach composite grain size distribution curves over time from the CERC Field Research Facil-ity at Duck, NC

5-84

EM 1110-2-181031 Jan 95

Figure 5-43. Change in mean grain size of the foreshore composites (high tide, mid-tide, and low tide samples) forsix profiles located along the central section of the beach fill, Ocean City, MD

5-85

EM 1110-2-181031 Jan 95

• Transformation.

• Analysis.

• Modelling.

• Display or output.

(c) The key operations that accompany the abovefunctions are:

• Data capture and entry.

• Database management.

• Data manipulation, correction, and analysis.

• Reporting and map production.

(d) Although initially GIS was considered a subset ofremote sensing and cartography, it has grown in recentyears into a discipline with its own theories, approaches,techniques, and interests. The topic is too complex tocover in detail in this manual, and the reader is urged torefer to the pertinent literature. A readable primer (con-taining an extensive bibliography and glossary) is pre-sented by Davis and Schultz (1990). More detailedcoverage of GIS concepts and discussion of ARC/INFO®

software is covered in Environmental Systems ResearchInstitute (1992).

(3) Management and use of GIS. Rapidly declininghardware costs have made GIS affordable to an increas-ingly wider range of agencies. In addition, many scien-tists and planners are realizing that GIS may be the onlyeffective way to interpret, display, and make more under-standable vast quantities of geologic and terrain data.GIS, however, is not a panacea for all of an agency’s dataanalysis problems. Use of a GIS in an organizationrequires a major commitment in training, funding, andmanagerial skill because the technology is relatively newand unfamiliar. Most agencies must accept new proce-dures in archiving and organizing data, performing qualitycontrol, updating and supporting software and hardware,and assigning key personnel to training and long-termpractical projects. The latter point is critical - users can-not simply sit down at a terminal, experiment with thesoftware for a few hours, and have any likelihood ofproducing effective or trustworthy data products. Thedecision to purchase and develop a GIS should not betaken lightly!

(4) Coastal data suitable for GIS. Some uninitiatedusers have the impression that GIS is a magic box thatcan display all kinds of geologic and marine data. Intheory, this may be true. In reality, cost of hardware anddata management is always a limiting factor. The moredata that a particular database contains, the more costlythe management, maintenance, and quality control of thatinformation. Table 5-20 summarizes some of the types ofcoastal data that can be included in a GIS.

(5) Data quality.

(a) It is critical that only the highest quality data,whether original discrete points or interpreted results (e.g.,shorelines extrapolated from aerial photographs) bearchived in a GIS. The erroneous impression has spreadthat GIS automatically means high quality and high accu-racy. GIS has the insidious effect that the output usuallylooks clean and sophisticated, and large numbers of chartsand summary statistics can be quickly generated.

(b) Unfortunately, recipients of computer-drawnmaps are often far removed from the assumptions andcorrections used to enter and analyze the original data.Table 5-21 lists a series of steps involved in creating amodern GIS map from historic field data. Data manipula-tion and interpretation occur at least five times betweenthe field operation and the completed map. Users of GISmaps must be appraised of the steps and assumptionsinvolved in analyzing their particular data. As with anyform of computerized analysis, garbage for input meansgarbage as output.

a. Coastal data interpretation with numericalmodels.

(1) Introduction.

(a) The use of numerical models in assessingchanges in coastal geomorphology is rapidly increasing insophistication. Models are designed to numerically simu-late hydrodynamic processes or simulate sedimentresponse on beaches, offshore, and in inlets. Specifictypes include models of wave refraction and longshoretransport, beach profile response, coastal flooding, andshoreline change and storm-induced beach erosion (Birke-meier et al. 1987; Komar 1983; Kraus 1990). The judi-cious use of prototype data and models can greatly assistthe understanding of coastal processes and landforms at astudy site. Because models should be tested and cali-brated, field data collection or mathematical simulation of

5-86

EM 1110-2-181031 Jan 95

Table 5-20Coastal data suitable for GIS

1. Index of all available geographic coastal data

2. Bathymetric• Original soundings - valuable for numerous purposes.• Gridded surface data - needed for volumetric computations (original data usually also retained).

3. Shoreline position - derived from:• Historic maps.• Recent field surveys.• Aerial photographs - photos require interpretation by an experienced analyst.

4. High-resolution seismic• Images of original records? No - too much data to store; not useable by most people without geophysical training.• Interpreted seismic results:

1. Depth to reflectors.2. Sediment type.3. Channels, faults, gas, features.

5. Side-scan sonar• Images of original records? No - too much data to store; not useable by most people without geophysical training.• Interpreted sonogram results:

1. Geohazards - debris, pipelines, shipwrecks.2. Surficial sediment - rock, sand, cohesive.3. Bedform orientation.

6. Surficial sediment (grab samples)• Mean grain size and other statistics (grain size distribution curve if possible).• Color - need standard nomenclature.• Organic content.• Carbonate content.• Engineering properties.

7. Core data• Photographic image of core? No - too much data to store.• Image of core log? Possible.• Grain size distribution and statistics at various depths.• Depths to interfaces (boundaries).• Organic content and other properties at various depths.• Engineering properties at various depths.

8. Oceanographic properties of the water column (temporal in nature)• Salinity, other seawater chemistry.• Currents at spot locations (temporal - vary greatly with time).• Suspended sediment concentration.

9. Biological• Bottom type if coral or reef.• Species diversity.• Individual species counts.• Pollutant concentrations.

10. Cultural (man-made) features• Shore protection.• Oil platforms, pipelines.• Underwater cables.• Piers, jetties, structures.• Real estate, roads, parking lots.

5-87

EM 1110-2-181031 Jan 95

Table 5-21Interpretation and data manipulation required to convert historic data to GIS product

1. Original structure or feature interpreted and measured by survey team in the field.• Field party skilful and methodical?• Best survey procedures used?• Equipment calibrated and maintained?

2. Information recorded onto paper charts or log books.

3. Additional interpretation occurs if data smoothed or contoured.

4. Historic maps and field logs interpreted by analyst many years later.• Changes in nomenclature?• Unusual datums?• Logs or notes incomplete? in same language?• Logs and maps legible?• Determination of date (old maps often display several dates).

5. Data translated into digital form for modern use.• Technician careful?• Appropriate corrections made for old datums or navigation coordinates?• Paper charts torn, stretched, or faded?• Adjoining maps same year? If not, use separate layers?• Digitizing or scanning equipment working properly?

6. Digital data incorporated into GIS database.

7. GIS maps (layers) interpreted by end user.• Data from different years correctly overlain?• Valid to compare data sets of greatly differing quality?

waves, tides, and winds at a project site is usuallyrequired.

(b) The advantage of tools like numerical models isthat they can simulate phenomena only rarely observed,can generate complex and long-duration changes, and canincorporate judgements and measurements from manysources. The use of numerical models is a highly special-ized skill, requiring training, an understanding of theunderlying mathematics, and empirical (“real world”)experience of coastal processes. This section summarizestypes of models and introduces some of their strengthsand limitations.

(2) Types of models1.

(a) Coastal experience/empirical models. This repre-sents the process by which an understanding or intuitivefeeling of coastal processes and geomorphology is adaptedand extrapolated from a researcher’s experience to a

_____________________________1 Material in this section has been summarized fromKraus (1989).

specific project. Prediction through coastal experiencewithout the support of objective quantitative tools hasmany limitations, including severe subjectivity and a lackof criteria to use for optimizing projects. Complete reli-ance on coastal experience places full responsibility forproject decisions on the judgment of the researcher with-out recourse to testing the “model” with alternate tools.An empirical model is always necessary before chosing anumerical model.

(b) Beach change numerical models. Figure 5-44summarizes the time ranges and spatial coverage ofnumerical models used by CERC. Summaries of thecapabilities of the models follow:

• Analytical models of shoreline change. These areclosed-form mathematical solutions of simplifieddifferential equations for shoreline change derivedunder assumptions of steady wave conditions,idealized initial shoreline and structure positions,and simple boundary conditions. Because of themany simplifications needed to obtain closed-formsolutions, these models are too crude to use fordesign.

5-88

EM 1110-2-181031 Jan 95

Figure 5-44. Classification of beach change models (Kraus 1989) error analysis of gridded bathymetry

• Profile change/beach erosion models. These areused to calculate sand loss on the upper profilecaused by storm surge and waves. The modelsare one-dimensional, assuming that longshorecurrents are constant. Extra work needs to bedone to extend their use to simulate major mor-phological features such as bars and berms.

• Shoreline change models. These models general-ize spatial and temporal changes of shorelinesanalytically in response to a wide range of beach,wave, coastal structure, initial and boundary con-ditions. These conditions can vary with time.Because the profile shape is assumed to remainconstant, onshore and offshore movement of anycontour can be used to represent beach change.

5-89

EM 1110-2-181031 Jan 95

These models are sometimes referred to as“one-contour line” or “one-line” models. Therepresentative contour line is usually taken to bethe shoreline (which is conveniently measured oravailable from a variety of sources). TheGENESIS model has been extensively used atCERC (Hanson and Kraus 1989).

• Multi-contour line/schematic three-dimensional(3-D) models. These models describe theresponse of the bottom to waves and currents,whose intensity and geologic influence can varyboth cross-shore and alongshore. The fundamen-tal assumption of constant shoreline profile, nec-essary for the shoreline change models, is relaxed.The 3-D beach change models have not yetreached wide application. They have been limitedby their complexity and their large requirementsfor computer resources and user expertise. Inaddition, they are still limited by our ability topredict sediment transport processes and waveclimates.

(3) Calibration and verification.

(a) Model calibration is the use of a model to repro-duce changes in shoreline position measured over a cer-tain time interval. Verification is application of a modelto reproduce beach changes over a time interval differentthan the one used for the model’s calibration. Successfulverification means that the model’s predictions are inde-pendent of the calibration interval. However, if empiricalcoefficients or boundary conditions change (for example,by the construction of an entrance channel which inter-rupts sand transport) the verification is no longer valid.Therefore, a modeler must be aware of any changes inphysical conditions at the study site that could affect thevalidity of his model.

(b) Unfortunately, in practice, data sets are usuallyinsufficient to perform rigorous calibration and verifica-tion of a model. Typically, wave gauge data are missing,and historical shoreline change maps are usually spotty orunsuitable. In situations where data are lacking, coastalexperience must be relied upon to provide reasonableinput parameters. This underscores that considerablesubjectivity is part of the modeling procedure, even if themodel itself is mathematically rigorous.

(4) Sensitivity testing.

(a) This refers to the process of examining changes inthe output of a model resulting from intentional changes

in the input. If large changes are caused by minorchanges in the input, the overall results will dependgreatly upon the quality of the verification. Unfortu-nately, for many practical applications, there is somedegree of doubt in the verification (Hanson and Kraus1989) (Figure 5-44). If a model is oversensative to smallchanges in input values, the range of predictions will betoo broad and will in essence provide no information.

(b) In summary, numerical models are a valuablecomplement to prototype data collection and physical(scale) models of coastal processes. However, usefulnumerical models require empirical input during thecalibration and may be based on incomplete data sets.

Therefore, the reader is urged to be cautious of the outputof any model and to be aware of the results of the verifi-cation and sensitivity tests.

5-6. Summary

a. Before initiating detailed field, laboratory, oroffice study, a thorough literature review and investigationof secondary data sources must be conducted. Existingsources of data are numerous, including information onprocesses such as waves, water levels, and currents, infor-mation on geomorphology such as geologic, topographic,and shoreline change maps, as well as information thathas been previously interpreted in the literature or has yetto be interpreted such as aerial photography. If such asearch is not conducted, assessment of geologic history islikely to be less reliable, field studies may be poorlyplanned, and considerable expense may be wasted becauseof duplication of existing information.

b. A wide variety of techniques and technologies areavailable for data collection, analysis, and interpretation ofthe geologic and geomorphic history of coasts. Onemeans of acquiring coastal data is through field data col-lection and observation. These data may be numerical ornon-numerical, and may be analyzed further in the labora-tory and office depending upon the type of data collected.Laboratory studies are used to analyze geological proper-ties of data collected in the field, such as grain size ormineralogy, or to collect data through physical modelexperiments, such as in wave tanks. Office studies arepart of most investigations, in that they involve the analy-sis and/or the interpretation of data collected in the fieldand laboratory, from primary and secondary sources.Typically, the best overall understanding of environmentalprocesses and the geologic history of coasts is acquiredthrough a combination of techniques and lines of inquiry.

5-90

EM 1110-2-181031 Jan 95

A suggested flowchart for conducting studies of coastalgeology is illustrated in Figure 5-45.

c. Many recent developments and techniques areused in the analysis of coastal data sets. The evaluationof geologic and geomorphic history is largely dependentupon the availability and quality of research equipment,techniques, and facilities. New techniques are constantlybeing introduced, and it is important that the coastal geol-ogist and engineer stay abreast of new techniques andmethods, such as remote sensing and geophysicalmethods, computer software and hardware developments,and new laboratory methods.

d. In addition to keeping up with recent develop-ments, the coastal scientist or engineer has the serious

responsibility of making accurate interpretations of thegeologic and geomorphic history of coasts. It is vital thatthe important research problem and objectives be clearlydefined, that important variables be incorporated in thestudy, and that the inherent limits and errors of theresearch techniques and technologies be recognized,including problems and assumptions involved in datacollection and analysis. To some extent, the coastal sci-entist or engineer can make some adjustments for varioussources of error. However, because of the geologic andgeomorphic variability of coasts, extreme caution shouldbe taken in extrapolating the final interpretations andconclusions regarding geologic history, particularly fromdata covering a short time period or a small area. Forthese reasons, assessment of the geologic and geomorphichistory of coasts is an exceptionally challenging endeavor.

5-91

EM 1110-2-181031 Jan 95

5-92