AD-AI17 316 ARMY ENGINEER INST FPR WATER RESOURCES FORT BELVOIR VA F/6 5/3

A COMPSJTER-BASEO INTERACTIVE MODEL FOR INDUSTRIAL LAND USE FORE--ETCIU)JUN 82 A SOICOECIIEA, M R KROUSE

UNCLASSIFIED

"

*-I ND

111 oQ 18U0&6.1 . 2 11.2

36

11125 114 M

Ill 0 O E

MICROCOPY RESOLUTION" TEST CHART

JUN

GOICOECHEA & KROUSE

A COM4PUTER-BASED INTERACTIVE MODELMR INDUSTRIAL LAND USE FORECASTING

AMBROSE GOICOECHEA, Ph.D.ym4 International Water Resources Institute

School of Engineering andApplied Sciences

George Washington UniversityWashington, DC10

MICHAEL R. KROUSEInstitute for Water ResourcesU. S. Army Corps of Engineers

Fort Belvoir, Virginia

1. INTRODUCTIONAn industrial engineering activity that is growing in relevance and

receiving due attention in the literature is that of identifying land areassuitable for future industrial use. As cities expand and multiply, thevarious activities that reflect the social-economic makeup of a community(e.g., industrial, commercial, residential, agricultural, etc.) competewith each other for use of the same fixed resource--land. It then becomesnecessary and meaningful to consider the science and art (e.g., economicand behavioral aspects) of land use forecasting.

Land use forecasting has long been a planning activity of interest tothe various Federal and State agencies, particularly those with mandatesfor the development of land and water projects. Certainly this is the caseat the U.S. Army Corps of Engineers, where land use forecasting has longbeen applied to the evaluation of economic benefits resulting from engi-neering measures and associated land uses. Over the last 50 years a numberof research efforts have been funded by the Corps relating to the develop-ment of analytical land use methodologies and, in some cases, the design ofcomputer-based forecasting models.

.1The purpose of this paper, bb*, is to review briefly the progressmade in the analytical and behavioral development of land use forecasting

-" models, to point to the modeling functions of special relevance to

EDILEIUTOI STAM W"T~S JUL 2 1 IM , I Db d VkhiI

B 82 07 19 271.. .J . . . . . . . . ... o - . .. . . . ... .... .. .. .,-..-. . . . . . . ... .... . . .,.-,-.-.-- -, , ,,_,.,:.. _ --

GOICOECHEA & KROUSE

industrial land uses, and describe a new interactive computer model beingdeveloped at the Institute for Water Resources (IWR) of the U.S. Army Corpsof Engineers.

2. BACKGROUNDA substantial number of efforts to develop land use forecasting models

have been undertaken over the last 30 years. This section compiles a listof over 50 models created during that time period that cover a wide rangeof forecasting activities, and that represent the extent of the modelingeffort in the private and public sectors.

The beginnings of land use forecasting in the United States are to befound in the schools of city planning created at Harvard University in1929, and at the Massachusetts Institute of Technology (MIT) in 1931.These two schools spearheaded the tremendous development that resultedthereafter. Kilbridge et al.[1] presents a classification of 20 urbanplanning models by land use (e.g., industrial, commercial, residential,agricultural, etc.), function (e.g., projection, allocation, and deriva-tion), theory (e.g., behavioral, gravity, trend, and growth index), andmethod (e.g., regression, input-output, markov process, linear programming,and simulation) that span the time period 1959-1967. Table 1 extends andupdates that classification by identifying 28 other models that are consid-ered most significant and that cover the time period (1962-1979). As Table1 reveals, these models offer a wide range of forecasting capabilities, usediverse analytical and behavioral approaches, and have been applied to agood number of cities in the U.S. Also, as the reader can observe, indus-trial land use forecasting is well integrated and represented in many ofthese models (see refs. 2 through 29 and 38).

The remainder of this section reviews some of the previous V rk thatled to the development of the Alternative Land Use Forecasting (ALUF) modelof the Institute for Water Resources, U.S. Army Corps of Engineers.

Land Use Forecasting at the Corps of Engineers. Essential to the taskof project development and evaluation is the determination of "withproject" and "without project" future economic conditions. The calculationof these economic benefits has provided, in fact, the motivation for muchof the effort on land use forecasting at the Corps. But substantial andcontinuous as this effort has been over the last decade, the need stillexists to develop a computer package that offers a satisfactory balance ofsound methodological framework, data base and computer time requirements.Some of the methodologies proposed in the past, although analyticallycorrect and based on sound methodological frameworks, were inadequate forsubsequent implementation for several reasons. At times the theoreticaldevelopment was valid and well researched, but the computer model wasincomplete. More frequent was the case, however, where the proposed

Avail ea/0o(~2) Dist speciallop . I

GOICOECHEA & KROUSE

Table 1. A Classification of Land-Use ForecastinX Nodeft

PROJIT AWN FORECASTING AIALTICAL TORS0 NCML VAUI CASE STUDT CAPAIILT1lS IPLOITD 8EtruE]Cg bam

I OLSS-iVP Area Sm Francis a plomen CS year e Regreslon Canter for real INu -SiutAtie4 by Ara &se 0 Judgeottal estate and wuhm 19

A0 Popation Usighting eenmoate 1* Industrial Land Uqiversity of

Use (21) California, larkelyo Residential Housing Also. N.J. Breaw

Location (1972)a Gov. mpioyuent Zid. (2)

2 PLUM - San Francisco 0 Household Location o Regression W. Golner (1%S) 1961Projeetive say Area o Population-Serving Functions (3)Land Use Model Employment o Subjective

I ndustry Location Probabilityo Land Uaeo Regional Employment0 Ra RonWl Population

3 Puget 3oad Puget Sound 0 CoD Eployment o 16 Setoer C.N. Grove (11) -Regional Regional Flan- a Industrial Location Input-Output (4) 1970Traprtaton Commision o Population ModelStudy Sattle. Vash. a Retail o Judgmetal

* Pop. by county •eightlng* plommt by

Industry

SEWIPC Sout.neastara, o Regional fEployment o Linear Pro- S. V'iscorst I566Wisconsin o Land Use 6ram.int Regional Planningeglonal Plan- Residential o Judgmental CAm.-MISSOA

sing Comnission Industrial (L.?.) o Input-Output Tch. get. 3Agriculture (5)

TALUS Detroit Regional o Employment by o R gression Rubia. J.J. (1564) 196"Transportation District a Gravity Accesa (6)and Lan Uae a Householda byStudy District

o Land Useo E omplolmento population

A HARVARD Southuust Sector o Industrial Location • Computer Steinte ad 197).of the Boton o RAsidential. Simulation togers (011o) isRegion. Esper. Recreation, Open a Universal Preposal for YearStudy (16 week) Space Trisve-se Fou- (. 7i)

o Trnn:portation .ercator Aole. !.L. NtanILrg(UT") Grid (19)o Grid h -tuork (7),(

MANGROVE poomery Day Land Environmental Planning o Ecological Cenf'er for Urban 19?3Use !tudles. Stratelies: Syste rtudies. Pliancollie' County o Canal Concept ModelnIng University. FL.Florita 0 Resource ur tev a Judgmental A.P. Veet et.a-

9 Filling of elnla ns(13)* Lang Presv1ation (9)

a AIRFONT Airponrt e coordinaoted Repint office of Itntre- 1910Enireas: of Lan Use Plainm poluin PlomingLad vie Central Controls. Rase and Developmet

R0veds olv ireamentalPlnnineg 2tvietenwashingeto, Ia'tO)

GOICOECHEA & KMOUSE

neodplasm raa- * Land Use Ale.-tlem e Useer Pre- Iva Pas T7%4 191!Ocelo ste a 'Wth and withot, grannis W.5Z anG MayModel: 11111S Aalysis k0 Regres (sion3liver, e • ZTwosome ncremenltal al (IAM =00 a Economic Aalysti

* Population Dlstrlbo-

im

10 uSI ,t. Louis 1109" * Determine Lmw o Factor laim Paw 714,PS 1975fur Industrial 0 Disinat Coras melocatlon In nood Analysis Py"r (1974)plate. 0 tettstioal (12)

Analysis

11 1 MLJUI- Cleveld o Populatiun o Crid Network eader #al 1971Dynamlo Land Study* o Employ.4n,. o ' roliulty Crave (1971)use allocation o Land uze F.ctors (12)model o Ch angcs In Zn(ra- o 'rafalo Display

12 OAX D3Z3 LAS. oO0-sq a I A c:- 1...l:d-use 0 ",.-f analysis r.A. Leasel. 1976

Ite.ion in a r.rn model 0 A~tractiveneas (14)Tenes ee Data ba .e f pop disc. jcore

labor availihie. *sie 4 a Staiocigalfreq of inl,,trieo n170i

a Cuulativ.e distribution ' lysisof exisrinc induscrie a Dphi techniquee

o £mploymonc by zone.

13 EMPItIC ioston Arce o Population by Zone o Llnesa Dief- Peat @3, oi. (19M1) 1911(Plus a lozen a Employment by Zone Orence equatilos (is)other Clies) • Lad use a StatLitleel

analyis* Mo-behovlerol,

14 Harvard The Interacticn o update vacant land bloom a"d brown 1919odel between urboiva- values)

timn. Lana Cuauiitaid pultty

Landsape - Land value mdelArchieetctreReseareh

harvard15 Unzvyersity - Housing d-del o Reside.al 1-onS Use vlklra et. al. 1979

. bSe. Atthelsd ty S'.nl...i c i;e- (1)cmputer taally ztriucturs.reaout)

- Pubic Inotitutlets Tidl and Oroa 1979

t1moel (Is)

?- TransportationM e trvel Deand oi Tyler ad Comalas is"

000e1 Trasportation facilities (19)

IS - :r4ustrial Model # )ndustrial isa Siting Gottry eL. as. 19"

BAsed on:SlopeDepth to w4roelZoning

o Useloguent coas uoedam econoie criletis

9 - Publi ellpen- . Land use hunge bNaed Itrll e0. Ga. 199

ditures model upon p" le loeo oXpo- (21)Mules,

1GOCOECHEA & ROUSE

* - Voter 4mtty a Voter demard Rgers en1d0

md uality . Coirom Count etuiek* Carbomaeeou o (n2)SDissolved Qeag=

a Salinity

11 - Veteatlie-uild e Podmce land Glass- t Iwolife Mel trieatioe syst ()

a ftaree Evaluation* Decription Hthsa

Commercial o A developers 2eropee- Wilkins sd (2' If"easel tlvt in estiating e (24)

locatl or and size ofoeoerctal cenlter

Mar,.d(caint.)

23 - Solid Wst@ a Landfilling and 0 RPereasio8 Rogers ad 1anagement ezlart-out-technolog sl o Land use McClellan

on a town-by-town exclusion (25)basis crtLeria

24 - Lsoal/Implel e Land ue alleatiee Gatetstaner, 195aostatlee msdst In aceordance with Steimit

state. toeral. sad (26)local land use controls

- istorisol o valuates the speelfio, Stelaits and 1918

Sa ea Id unique, often Gajlitative oagluadVolues of amoe am (in)

- Reeeatlee Model o Identifies aits suitable Steluits and 197826 for recreational development buglas

and ranks tha. (28)

a! - Slls Hadel Idettficatiom of eilleroio ionem.

o transport of sed!nztto accumulation areas

* Estimation of cstsssociated withmitigating procedures

28 - Land use o Cover characteristics Way. D.S. 1978Descriptors o Construction (29)

0 Cost

computer model required vast amounts of input data, the exogeneousparameters themselves were difficult to estimate (e.g., spatial population

distributions) or the amount of time required to apply the model would havebeen unreasonably large (in the order of months).

An alternative course of action is delineated here. Essentially, someof the computer subroutines in program RIA are combined with an economicdata bank file and a search procedure to allocate land uses. Optimal land

allocations are not sought; instead, "near optimal," feasible land alloca-tions are desired.

. . . .. . . .. .

GOICOECHEA & KROUSE

3. ALTERNATIVE LAND USE FORECASTING (ALUF) PROGRAMThe development of a grid cell data file requires that each variable

map be individually encoded and geographically registered to a common baseand stored, along with data variables in the data bank, on a computer stor-age device.

The IWR package consists of two computer programs which are used inconnection with a grid cell spatial data base as shown in Figure 1. Themain program, Alternative Land Use Forecasting (ALUF) does the actualallocation of future land uses to specific grid cells. The Existing LandUse Analysis Program (ELUA) is provided to help identify significant landuse location factors for the allocation process based on the relationshipbetween land use locations and other data available in the grid cell databank.

The final program output is a new data variable written into the database file for each grid cell, indicating projected future land use. Theprograms are written in FORTRAN IV for the CDC 6600/7600 series computers.

The ALUF program incorporates the HEC RIA Attractiveness modelingprogram and RIA Distance Determination package. These were adapted for usein this process so that land use locator scores can be developed accordingto user specified criteria, as well as location criteria derived from thestatistical findings.

The kinds of data variables commonly used as a basis for allocatingfuture land use include:

A. Access (Distance)1. Transportation2. Central Business Districts or Regional Centers3. Dependent Activities

B. Proximity to Compatible Land UsesC. Physical Land Attributes (Developability)

1. Slope2. Drainage3. Type of Cover4. Soils

D. Infrastructure1. Sewers and Water2. Gas and Power3. Mass Transit

E. ZoningF. OwnershipG. Land Prices

GOICOECHEA & KROUSE

SIATUCATIV3Im

X1 2 mcLtncAi Rm.S

L DISTANCE SATDSTORCIYC

u ZN DTADATA

LA SS1 PANDMETESE

L ITNC ADDS ORANGE SDTRINTO SCOR - --- LADE VA-Lt8I1

Im TO I DATA BAWK

VALUES9 BE TDlE ATAaiEACH mAND US

DISTANCE VARIABLES,BANK

Vv L. Curnet Stv'utuTo Of COmte: Proarm ALUP

GOICOECHEA &KROUSE

4. CALCULATION OF THE ATTRACTIVENESS INDICESTo demonstrate the computation of the raw attractiveness indices as

performed in the computer program ALUF, consider the land use grid shown inFigure 2. Purposely, the grid is small and contains only 27 cells, so asto render the exercise workable (typically grid representatives of regionsof interest may require 5,000-50,000 cells).

Listed in Figure 2 is the legend used to represent the various landuses. e.g., (1) natural vegetation, (2) developed open space, (3) low den-sity residential, etc. In this manner, we can see that grid cell (i,j) =(1,4) is currently allocated to low density residential. A railroad tracktraverses the grid network, as shown.

As program ALUF is structured currently, a matrix arrangement isavailable to the analyst to identify the variables (topographic) of inter-est, as shown in Table 2. The analyst-user then is required to: (1) de-sgnate topographic variables, (2) assign relative weights to thevariables, and (3) specify a shading intensity for each value of eachdesignated variable. A matrix must be filled in for each land use (e.g.,activity) being considered. For illustrative purposes, Table 2 alone isshown with the matrix values for industrial use.

We continue our illustrative computation of the attractiveness indicesfor industrial use with the specification of two variables only: (1) dis-tance to Seaboard Railroad (variable 023), and slope (variable #8). Infor-mation on these two variables must be built into the data bank file priorto running the program. For our example, this information would appear asshown in Figures 3 and 4. With reference to location (ij) a (1,4), wenotice that the slope value of 2 corresponds to a "2 to 6 percent slope"(Table 5, variable 8, Appendix), and the distance to the railroad tracks isthree cell units. The actual computation of the raw Attractiveness Indexproceeds as follows:

Distance to R.R. (1) x (2) : 2Slope z 2 (8) x ) 8

Shading IntensityRelative WeightAttractiveness

In a similar manner, indices (also called scores) for the remaining cellsare computed in Table 3 and again shown in Figure 5.

There remains the matter of using the attractiveness scores to allo-

cate a land use to each grid cell. Currently, the program assigns landuses according to the priority identified by the analyst in the Data Deck;

GOICOECHEA At KROUSE

.j..~I~j.~I~i1.T;D In rm"5~.

0 u tv ' j

3 .a 0 S

2~ ~ ~~~m "I e st

&M aa s

ftedo : ft _~at 040,3

-- A--

I 3 21'IanI

CA WO

OOICOECHEA & CROUSE

Table 2. Attractiveness Matrix for Industrial Use

Shading IntensityTopographic ImportanceVariable 0 1 2 3 4 5 Weight

(23) Distance to R.R. -0 1-0 -5 -1 7 2(8) Slope -1 10 8 2 0 1

Existing Land Use 0 10 -1 -1(17) Distance to E.R. 0

Distance to E.I. 10 10 9NOTE: Shading intensity values range from -1 to 10. A value of 10 is

assigned if variable is of most significance to land use being con-sidered. A value of -1 is assigned if variable is to be excludedcompletely from further consideration.

Table 3. Attractiveness Scores for Industrial Use

CELL NO. SCORES(ii) DIST R.R. SLOPE DIST R.R. SLOPE TOTAL1,4 3 - (1)(2)=2 8 )(1 )=8 101,5 3 2 (1)(2)=2 (8)(1)=8 102,3 2 2 (5)(2)=10 (8)(1)=8 182,4 2 3 (5)(2)=10 (2)(1)=2 122,5 2 3 123,2 1 1 (10)(2)=20 (10)(1)=10 303,3 1 2 (10)(2)=20 (8)(1)=8 283,4 1 3 (10)(2)=20 (2)(1)=2 223.5 2 33,6 1 3 N4,2 0 1 0 (10)(1)=1O 104,3 0 2 0 (8)(1)=8 8

4,4 0 2 0 (8)(1)=8 84,5 0 3 0 (2)(1)=2 24,6 0 3 0 (2)(1)=2 25,2 1 1 (10)(2)=20 (10)(1):1O 305,3 1 1 " " 305,4 1 1 305,5 1 4 (10)(2)=20 (0)(1)=0 205,6 1 4 " 206,2 2 2 (5)(2)=1O (8)(1)28 186,3 2 1 (5)(2)=10 (10)(1)=10 206,4 2 1 (5)(2)z10 (10)(1)210 206,5 2 4 (5)(2)z10 (0)(1)ZO 107,2 3 2 (1)(2)u2 (8)(1)*8 107,3 3 2 " 107,4 3 0 (1)(2)22 (-1).".REJECT -

. . .. . . . . . . .. .. . .. . . ~ , . .I

" /

GOICOECHEA & KROUSE

that is, if the desired priority is industrial, followed by high densityresidential, low density residential, commercial, etc., then the analystphysically places data cards for industrial at the top of the "Data Deck,"followed by data cards for high density residential, and so on. In thatmanner, given a request for 13 cells, say, for industrial, the programassigns a land use (Legend Code 7) to the 13 cells that exhibit the highestindustrial attractiveness score. A similar allocation rationale is thenused for high density residential, and so on down the priority list. Forour example then, the cells allocated to industrial use are shown In Figure6. Note that for cell(6,5) there corresponds a slope value of 4 (i.e., 10

to 15 percent grading) and that Table 3 shows a shading intensity of zero;the slope variable, then, contributes a value of zero to the attractivenessscore, e.g. (0)(1.0) = 0.0. Cell(7,4), on the other hand, has a slopevalue 0.0 (i.e., water body) and since an intensity value of -1 has beenassigned to it, the cell is excluded from industrial use.

5. AN ILLUSTRATIVE COMPUTER APPLICATION: TRAIL CREEK PILOT STUDYNow that the computation of the raw attractiveness scores has been

illustrated in a step-by-step manner, the application of the procedure to areal-world situation is demonstrated using computer program ALUF. Theregion of interest is the Trail Creek study area shown in Figure 7, and itexhibits variety and complexity of roads, railroad track, river lengths,urban center nearby, etc. Current land use of this area is as shown inFigure 7, with adopted dimensions for each rectangular cell of 200 and333.3 feet.

The interactive computer mode of the program was then used to fill inthe attractiveness matrices. This time it is noted that the exercise wasextended beyon 'he industrial land use stated requirement to includeresidential and commercial. The number of cells required for each use was900, 800 and 200, respectively.

Finally, shown in Figure 8 is the computer printout of the computedfuture land use pattern. Only the left half of the pattern is used, as theother half would be of a similar nature. The actual computer printout doesyield the two halves, however. Let us now compare existing and future landuse of a particular cell, say cell(35,55). It is observed that Figure 7identifies the current use as being agricultural (i.e. code number 6), andnow the future use is projected to be industrial (i.e. code number 7), asgiven in Figure 8.

6. SUMMARY AND CONCLUSIONSThis paper discusses the architecture and use of a new land use fore-

casting model labeled ALUF, Alternative Land Use Forecasting. The modelmakes use of information on current land uses, topographic characteristics,and preferences elicited from the planners to forecast future land uses.

j .._ .. ....IId ..... .. ..n -" " ..

)ECHEA &KROUSE

........ .I 61 ileis

3 pin~.....gI*... ... ts..01*1 ... .. ..... ..... ,,.... 666

~~I'~*I*IjiessmmI.st..ue~.et...~r .......of .. 38 .ri cells .. . 6.... ....

S r *.~~~~~~~~~~~~....... .. . .** ..... nm....6 ..... ~ ,.....*t.

.. ...74 "m ,' .. S.t~~,II')iS~kl

A te .....*It5 m h('. c.-9b....9ll~99 6666I31I46* 156*I lt "0t.bbT~ 51

0-211211111111.u f ...... us...j&

I~ ~ ~ ~ ~~~~~~~. so s....,.S.i.4L u,*Suuu.

ai: go. .91 atist*~'41.,, =4usto is is s~l i I~St.

Iase ~ ss i so5.gSssitoo5SS 99I

.. . . n tm m :.N. s s S ~ I . . .S I

iiiitt'.5iiiiiin*: ....h~

s 5. ., sa uu mus m .. . ...... 6

:: M M I .''. ~ gglgt$$$ Soso I let.-mul ....

a ~ ~U32 USj CATISOZUS

11-oo" &Ssowms

600felt*

£ TRAIL CREEKmismeN #I Fixu 7. EXISTING LAND USE PATTERN

GOICOECIIEA &KROUSE

pr eftg no. a n* I4~~j cis* u~i-Prit Il3lli6 IIV3 pum rS S U146.l 4~~ii

3UMTOI~kL@ acoi*Taalo lllMOCSI

S333:12666911915111833 aacc~aiIIaU

Zci~ccacaliiong333 aa6 aaaaaaa

&4666:6 6"21121120913SiS

I£2966"66 & a~c6 66060216&994"S

g cccl iiaii~~~ aaccllIcc4"

I cccch~~~cccca * l6c6 Sclcu£ cc~cecccucia acaca1M ISSccccl,

* ~ ~ 96 "661 Stc~cccccccil it"U Iccc

g l~c4 asek*.c Is &&accl 1987#7 MI NM l?* accaacS" hcc1I aaccc7d"a7"l33c37cccc

II

8781"77713iaai A allllccl=p?1P2233flP=187119"977113 7,i11 a a aCS a aZ ZI aas7?3 2 2 1c 1

I It* lgaa.3a.i,1333?ffla77Pfcd64

119 1 i aS 777I~~~~~~~~~~z I33aa1 a~an7,77Sia1cccc

9 4= 9111 stiltas :?a,7?l~sia~l?1cac

7 01.1171S IIIsSAll I .1P?71 .3334Si ts59 s711? a tIa S~sS3SuS7?S:2~:.9???lh1c~cc11a3flP7.1131 5333* u5 9,aS SSS? 7S52 S27f 1117 2cS~d~lI SSS~ 31:11 a

11122=1a22422222.2211 2:aas~a1l1l

911 &&2N "a 7,1aa222aaa:laaaalsaaaa

a wa

3 03

GOICOECHEA & KROUSE

In the process, it calculates the economic benefits to be derived from aproposed engineering measure or zoning policy. The model is currentlyoperational and it is available to Corps personnel and general city plan-ners involved in project development and evaluation. Also, it is hopedthat industrial engineering practitioners will find it useful in theirdialogue with city planners as new industrial enterprises in growing commu-nities are discussed.

REFERENCES

Kilbridge, M., R. O'Block, and P. Teplitz, "A Conceptual Frameworkfor Urban Planning Models," Management Science, Vol. 15, No. 6, 1969.

Brown, H. C., J. R. Ginn, F. J. James, J. F. Kain, and M. R.Straszheim, Empirical Models of Urban Land Use: Suggestions onResearch Objectives and Organization, Exploratory Report 6, NationalBureau of Economic Research, New York, 1972.

Goldner, W., Projective Land Use Model (PLUM), Bay AreaTransportation Study Commission, BATSC Technical Report 219, SanFrancisco, California, September 1968.

Hydrologic Engineering Center, Resource Information and AnalysisUsing Grid Cell Data Banks, Computer Program 4DI-X6-L7590; UserManual, U.S. Army Corps of Engineers, Davis, California, September1978.

Hydrologic Engineering Center, An Investigation of Concepts andMethods for Broadened Scope Flood Plain Information Studies, PhaseI , Oconee Basin Pilot Study, U.S. Army Corps of Engineers, Davis,California, September 1975.

Jolissaint, C. H., Issues Related to Interfacing Water ResourcePlanning and Land Use Planning: Development and Application ofQuantitative Procedures, INTASA, Inc., Menlo Park, California, May1977.

Note - Space limitations preclude a complete listing of references.Please write to authors requesting such listing.

iF!