evaluating generalizations of hydrography in differing ... · c. brewer et al. icc2009, november...
TRANSCRIPT
C. Brewer et al. ICC2009, November 2009
Evaluating Generalizations of Hydrographyin Differing Terrains for The National Map
of the United States
Cynthia Brewer, Pennsylvania State UniversityBarbara Buttenfield, University of Colorado—Boulder
E. Lynn Usery, CEGIS, U.S. Geological Survey
Research Assistants:Chelsea Hanchett and Paulo Raposo, Penn State
Mamata Akella, PSU→ESRIChris Anderson-Tarver and Jochen Wendel, CU-Boulder
TNM data prep: NHD programming: Much advice:Tom Hale, USGS Larry Stanislawski, USGS Charlie Frye, ESRI
C. Brewer et al. ICC2009, November 2009
Scope
Multiscale topographic mappingGeneralization of National Hydrography Dataset
(1:24,000 High Resolution NHD)Multiscale topographic map designFor USGS web delivery portal The National Map (TNM)
Scale range 1:20,000 – 1:200,000 (1:5K – 1:2M planned)
Why focus on hydrography? Highly sensitive to scale and landscape differencesCommonly displayed on most base maps
C. Brewer et al. ICC2009, November 2009
Objectives
Work toward full automationMinimize anticipated times for manual editingUse COTS software or USGS available code
Maximize efficiency Make no more generalized data versions than necessaryLevel of detail (LoD) data
Preserve important hydrographic/cartographic differencesTerrain differences – flat, hilly, mountainousNext: climate impacts – humid, dry
C. Brewer et al. ICC2009, November 2009
Humid —Flat: Upper Suwannee, FL-GAHilly: Pomme de Terre, MOMountainous: South Branch Potomac, W V
Dry —Flat: Lower Beaver, UTHilly: Lower Prairie Dog Town Fork Red, TXMountainous: Piceance-Yellow, CO
Urban — St. Louis, MO; Atlanta, GA
Landscape Types
Subbasin Sample
Four in Iowa:
North and South Raccoon, Middle Des Moines, and Lake Red Rock
C. Brewer et al. ICC2009, November 2009
Evaluating the solutionsContextual
– Map series across range of scales– Critique by domain experts (hydrologists, cartographers)
Validation– Compare against 100K Medium Resolution NHD
Metric– Summary statistics on retained geometry– Channel length, network local density– Catchment areas, upstream drainage
C. Brewer et al. ICC2009, November 2009
Building an LoD — 3 case studies
A. Isolate a single, continuous centerline– Intersect NHD Flowlines with NHD Areas– Dissolve on GNIS name– Simplify
B. Identify NHD Flowlines by local density– Prune differentially– Simplify, concatenate
C. Isolate local textures from NHD Areas / Waterbodies– Analyze, aggregate, simplify or smooth– May require treating feature codes individually
C. Brewer et al. ICC2009, November 2009
Case Study A:Isolate a single
continuous centerline
Pomme de Terre River,
Missouri—
humid - hilly
We are almost there…
C. Brewer et al. ICC2009, November 2009
Pomme de Terre, MO — NHD 24Khumid - hilly
3. Prune and simplify flowlines
2. Primary and secondary centerlines
1. Select and simplify waterbodies and areas
C. Brewer et al. ICC2009, November 2009
Draft result — MO 50K LoD mapped at 100K
Derived 50K LoDHigh resolution (24K NHD)
C. Brewer et al. ICC2009, November 2009
Validation
Derived 50K LoDMedium resolution (100K NHD)
C. Brewer et al. ICC2009, November 2009
Missouri series with data from
The National Map– 24K map
ArcGIS to JPEG to screen capture
C. Brewer et al. ICC2009, November 2009
Missouri, 50K map with 24K hydro
Missouri50K map with
24K hydro
1
3
24
C. Brewer et al. ICC2009, November 2009
Missouri, 50K map with 50K LoD
Missouri50K map with
50K LoD hydro
Visual evaluation of hydro in map
context
1
3
24
C. Brewer et al. ICC2009, November 2009
Missouri 50K LoD and original 24K hydro
C. Brewer et al. ICC2009, November 2009
Case Study B: Identify NHD
Flowlines by local density
South Branch Potomac River,West Virginia
—humid -
mountainous
C. Brewer et al. ICC2009, November 2009
What is meant by local density …
C. Brewer et al. ICC2009, November 2009
S. Br. Potomac, WV — NHD 24Khumid - mountainous
3. Isolate non-centerline flowlines
1. Prune differentially, merge, build diffs_file
2. Intersect NHD flowlines with NHD areas / waterbodies to delineate centerlines
4. Simplify
Pruning criterion: Summed remaining stream channel lengthSummed original catchment area
C. Brewer et al. ICC2009, November 2009
1. Prune differentially, merge, build diffs_file
C. Brewer et al. ICC2009, November 2009
Draft result — WV 50K LoD mapped
At 100K: With original 24K hydro With derived 50K LoD
C. Brewer et al. ICC2009, November 2009At 100K: With original 100K hydro With derived 50K LoD
Draft result — WV 50K LoD mapped
C. Brewer et al. ICC2009, November 2009
West Virginia series from TNM – 24K
West Virginia series with data from
The National Map– 24K map
C. Brewer et al. ICC2009, November 2009
West Virginia, 80K
West Virginia 80K
C. Brewer et al. ICC2009, November 2009
West Virginia, 200K
West Virginia 200K
C. Brewer et al. ICC2009, November 2009
Case Study C:Isolate local textures from NHD Areas and
Waterbodies
Upper Suwannee River,
Florida-Georgia—
humid - flat
C. Brewer et al. ICC2009, November 2009
Upper Suwannee, FL-GA — NHD 24Khumid - flat
1. Swamp / Marsh
2. Lakes / Ponds
C. Brewer et al. ICC2009, November 2009
1. Swamp/ Marsh
C. Brewer et al. ICC2009, November 2009
1. Swamp/ Marsh
C. Brewer et al. ICC2009, November 2009
Florida-Georgia, 80K map with 24K hydro
Florida-Georgia, 80K map with
24K hydro
C. Brewer et al. ICC2009, November 2009
Florida-Georgia, 80K map with 50K LoD
Florida-Georgia, 80K map with
50K LoD hydro
C. Brewer et al. ICC2009, November 2009
The National Map: eight themes
• elevation• land use / land cover• boundaries• transportation• structures• hydrography• geographic names• orthoimagery
nationalmap.gov New Viewer announced for December
C. Brewer et al. ICC2009, November 2009
Practical considerations in design development
• No custom edits on TNM data (no clean up)• Simple geometric symbols, so no missing fonts or
pictures on export• Regular fonts for export and file sharing (vs. USGS look
with Souvenir and Univers)• Lots of group layers to easily turn off categories while
evaluating appearance• All rasters and layers with transparency at bottom of
TOC so export retains editable vectors and type• No over/under passing on bridges and ramps (user may
query this detail using GeoPDF click out to Google)
C. Brewer et al. ICC2009, November 2009
Color contrasts
• Red roads vs brown contours – use gray contours
C. Brewer et al. ICC2009, November 2009
Color contrasts
• Red roads vs brown contours – use gray contours• All colors lighter than black labels – few halos
(only blue hydro labels and green reservation labels use halos)
C. Brewer et al. ICC2009, November 2009
Color contrasts
• Red roads vs brown contours – use gray contours• All colors lighter than black labels – few halos
(only blue hydro labels and green reservation labels use halos)
• This is not a road map – do not use whole contrast range on road categories
C. Brewer et al. ICC2009, November 2009
Color contrasts
• Red roads vs brown contours – use gray contours• All colors lighter than black labels – few halos
(only blue hydro labels and green reservation labels use halos)
• This is not a road map – do not use whole contrast range on road categories
• Contrasting outlines on point symbols separate from each other and background
C. Brewer et al. ICC2009, November 2009
Color contrasts
• Red roads vs brown contours – use gray contours• All colors lighter than black labels – few halos
(only blue hydro labels and green reservation labels use halos)
• This is not a road map – do not use whole contrast range on road categories
• Contrasting outlines on point symbols separate from each other and background
• Leave contrast available for update and overlay of operational information– magenta could be used for additions if no magenta symbols
C. Brewer et al. ICC2009, November 2009
Systematic color categories for symbols
Rd
YG
Pu Cy
Bu
Or
Yl
Gntransportationbuilt-up areas
wooded areasforest reservesparks
hydrography
points:emergencyhospitalsschools
Gy
boundaries
Br
Human themes: Natural themes:admin reserveshillshade
C. Brewer et al. ICC2009, November 2009
Systematic color categories for symbols
Rd
YG
Pu Cy
Bu
Or
Yl
Gntransportationbuilt-up areas
wooded areasforest reservesparks
hydrography
points:emergencyhospitalsschools
Gy
boundaries
Br
Human themes: Natural themes:admin reserveshillshade
C. Brewer et al. ICC2009, November 2009
100K
24K
Multiscale durability:good point overlaps
C. Brewer et al. ICC2009, November 2009
Hydrography
• Label with a paired layer, classed differently (not symbolized)
• More level-of-detail (LoD) databases in development(if 50K LoD covers 50-200, does 200K LoD cover 200-800?)
• Tapered by symbolizing upstream drainage area (stream order not useful)
• Intermittent and perennial symbolized
C. Brewer et al. ICC2009, November 2009
Representing stream hierarchy
Class upstream drainage area attribute and represent by width and color
C. Brewer et al. ICC2009, November 2009
Stream tapering
Width (pts) RGB
lighter0.50 100,180,2000.75 100,180,200
med blue1.00 50,165,2001.35 50,165,200
darker1.70 0,150,2002.00 0,150,200
C. Brewer et al. ICC2009, November 2009
Multicolor hillshade with transparency
C. Brewer et al. ICC2009, November 2009
Additional workGeneralization sequences for additional landscapes: settlement patterns (rural vs. urbanized) and coastal
Tailor generalization algorithms, parameters, sequences to landscape types
Decide the smallest number of tailored solutions that will apply to any generalization situation, nationally.Confirm they are not interchangeable through testing
Test topo designs in multiple formats (ArcGIS, GeoPDF, cached tiles for web services, print) and multiple resolutions (print, 96 ppi desktop, 130+ ppi laptop)
C. Brewer et al. ICC2009, November 2009
Funding acknowledgements
USGS Center of Excellence for Geospatial Information Science (CEGIS)
Funded through Department of Interior CESU program, January 2007 to present
ResourcesProject resources: ScaleMaster.orgLynn’s Center: cegis.usgs.govCindy’s website: www.personal.psu.edu/cab38babs’ Meridian Lab: greenwich.colorado.edu