A Simple Drainage Enforcement Procedure for Estimating Catchment Area Using DEM Data David Nagel, John M. Buffington, and Charles Luce U.S. Forest Service, Rocky Mountain Research Station Boise Aquatic Sciences Lab NW GIS User Conference, Oct. 17-21, Boise, ID October 20, 2011
Study Area Middle Fork Salmon River 10,000 km of rivers and streams ~ 1,000 km used by salmon
Our Purpose Estimate catchment area Discharge Grain size Salmon spawning habitat
Data 10 m NED DEM dataset http://ned.usgs.gov/ NHD stream lines http://nhd.usgs.gov/
Problem NHD stream DEM flow line Flow accumulation lines from DEMs do not match vector stream lines
Drainage Enforcement Objective Non-enforced Enforced Recondition the DEM so all cells drain toward the stream and downstream
Foundation of Watershed Analysis 1) Fill 2) Flow direction 3) Flow accumulation Fill Direction Accumulation
Flow Accumulation 1,000 12,000 Counting up the number of cells that contribute to a location on the stream = catchment area
Raster Accumulation to Vector Raster accumulation 12,000 cells Attribute corresponding vector
Mismatch Between Raster and Vector Without Enforcement Non-enforced Enforced Can t attribute vector if data are mismatched
Why the DEMs and Stream Lines Do Not Match in Flat Valleys
DEM Production Process 1) Aircraft 2) Aerial photo 3) Stereo plotter 4) Map production 5) Scan and tag 6) LT4X LT4X, Infotec Development, Inc.
Original Contours and 10 m DEM Model 500 m LT4X Original 40 contours Blue box = 100 m x 100 m 2 m contours derived from 10 m DEM
Original Contours and 10 m DEM Model With Streams LT4X Original 40 contours 2 m contours derived from 10 m DEM
LT4X made the DEMs from the original contours There were no contours in flat valleys, so there wasn t information for generating the DEM data Cartographers drew stream lines where they saw them, independent of the LT4X model
Why Not Use LiDAR? Site scale vs. landscape scale
Drainage Enforcement Algorithms 1) ANUDEM - ArcGIS 2) AGREE Arc Hydro Tools 3) IDDEA Forest Service
ANUDEM Implemented by ArcGIS (Topo to Raster) Not designed for reconditioning DEMs directly - requires contour lines or point input Contour file sets may become too large Avoids trenching Hutchinson, M.F., 1989. A New Procedure for Gridding Elevation and Stream Line Data with Automatic Removal of Spurious Pits. Journal of Hydrology, 106:211-232.
AGREE Algorithm Implemented by Arc Hydro Tools Dewald, T., NHDPlus User Guide, U.S. EPA and USGS, April 29, 2008 Hellweger, F., 1997. AGREE DEM surface reconditioning system. Center for Research in Water Resources
Drainage Enforcement with AGREE Original Reconditioned Arcs must point downstream Smooth drop may modify watershed boundaries Uses trenching
IDDEA Method Inverse Distance Drainage Enforcement Algorithm 1) Grid stream lines 2) Generate Euclidean distance from all stream lines 3) Drop stream by constant value (e.g. 200 m) and 4) Invert distance, multiply by constant: (1/d) * 1000 5) Subtract result from original DEM Where c ij, 200, else e ij ((1 / d ij ) * 1000) c ij is a stream channel cell at raster location ij e ij is the elevation at ij d ij is the Euclidean distance at ij
1) Grid Stream Lines Vector to raster conversion
2) Euclidean Distance from Streams
3) Inverse Euclidean Distance (1/d) * 1000 200 100 50 10 0.1 Value decreases (unitless) 200 0.02 Force constant drop (200 m) at stream channel Enforcement decreases away from the channel
DEM Minus Inverse Distance minus =
Profile Results Preserves relative topography and watershed boundaries
Run Flow Accumulation Nonenforced Enforced
Attributing Vectors Non-enforced Enforced Higher probability of accurately attributing vectors with the correct contributing area
Results Comparison Catchment Area Comparison y = 0.9984x + 7.6833 R² = 0.9999 Non-enforced Catchment Area (HA) 1000 900 800 700 600 500 400 300 200 100 0 0 200 400 600 800 1000 IDDEA Catchment Area (HA)
Disadvantages of the IDDEA Method 1) Trenches the DEM 2) Meander bends smaller than cell size get cut off and reroute flow Trenching Cut-offs
Advantages of the IDDEA Method 1) Works relatively quickly at landscape scales 2) Preprocessing of raster or vector data is not required 3) Preserves relative topography and watershed boundaries
Thank you
Selected References Buffington, John M., David R. Montgomery, and Harvey M. Greenberg, 2004. Basin-scale availability of salmonid spawning gravel as influenced by channel type and hydraulic roughness in mountain catchments. Canadian Journal of Fisheries and Aquatic Science 61:2085-2096, doi: 10.1139/F04-141. Crystal, Roger E., and Jason Underwood, 2002. Hydrologically Enhanced, High-Resolution DEMs. Geospatial Solutions, April 1, 2002. Hellweger, F., 1997. AGREE DEM surface reconditioning system. Center for Research in Water Resources, University of Texas at Austin. URL: http://www.ce.utexas.edu/prof/maidment/gishydro/home.html Hutchinson, M.F., 1989. A New Procedure for Gridding Elevation and Stream Line Data with Automatic Removal of Spurious Pits. Journal of Hydrology, 106:211-232. Simley, Jeff, 2004. USGS National Hydrography Dataset Newsletter, vol. 3, no. 4, February 2004. URL: nhd.usgs.gov/newsletter_list.html. Simley, J.D., Carswell Jr., W.J., 2009. The National Map Hydrography: U.S. Geological Survey Fact Sheet 2009-3054, 4 p. USGS, 1997. Overview of DEM Production History, Historical DEM II Production Processes, September 1997. URL: http://rockyweb.cr.usgs.gov/hazards/demmtg97/demhist.html USGS and EPA, 2008. NHDPlus User Guide, URL: http://www.horizon-systems.com/nhdplus. April 29, 2008.