Development of Improved Stream Mapping for NC Presentation for Soil Science Society of North Carolina NC Division of Water Quality January 21, 2009
What are stream maps used for? Riparian buffer rules Water supply protection rules Siting of Solid Waste landfills Local buffer protection programs Land use planning Site planning Transportation improvements alternatives selection
Which maps are used now? Mainly USGS 1:24,000 topographic maps blue line streams Also streams shown on county soil surveys only rarely digital Web Soil Survey unfortunately does not show all streams on original maps Some counties have separate mapping using computer based models
USGS Topographic Map 1:24,000 scale
County (SCS or NRCS) Soil Survey Harnett County 1994
Problems with Existing Maps Variability between (and within!) maps Not all digital (especially County Soil Surveys) Varying accuracy Some local maps often uncertain how they were derived In general, no known accuracy levels associated with any of these maps
Issues in Map Accuracy of Stream Networks Variation among quad sheets in how small streams are depicted Some quad sheets show both intermittent and perennial reaches, some show only perennial streams. Stream networks among quad sheets vary in drainage density (miles of streams/square mile of area) Streams often do not connect across quad sheet boundaries When topographic maps were produced from aerial photos in the 40s, 50s, and 60s, USGS had no firm standards for mapping streams. Individual cartographers used own judgment.
Issues in Map Accuracy of Stream Networks USGS topographic maps usually greatly underestimate the number and total lengths of headwaters streams, except in the Coastal Plain where ditches were often mapped as streams. NRCS soil survey maps usually overestimate the number and total lengths of headwaters streams. NRCS has different stream mapping convention than USGS. Most recent NRCS soil surveys only show streams shown on USGS maps.
Contiguous USGS Maps Tributaries shown on one map and not the other Headwater stream shown on one map and not the other McDaniel Bald and Marble Quadrangles, Cherokee County, NC
Drainage densities different between adjacent maps
Table 2. Percent error in total stream length within the study watersheds as depicted on NCRS and USGS maps compared to onsite determinations. Region NRCS Maps USGS Maps Coastal Plain +29 +31 Piedmont +3-25 Northern Mountains +25-48 Central Mountains +25-30 Southwestern Mountains +35-54 NCDWQ Teams Mapped streams with GPS in randomly selected study watersheds of about 0.5 mi 2 in area
Stream definition Basic Problem One underlying past problem is lack of consistent stream definition Class taught by NCSU (Dr. Jim Gregory) and DWQ staff for past 10 years about 40 times Developed and refined stream identification method so useable across entire state Four day class with field and written exams Method now standard in North Carolina and widely used in southeastern U.S. Identification Methods for the Origins of Intermittent and Perennial streams NC Division of Water Quality Version 3.1 Effective Date: February 28, 2005 Available on website at h2o.enr.state.nc.us/ncwetlands
What is the solution? Goal - A GIS-based map of streams in the state based on field data with known levels of accuracy. NC OneMap will be the new state stream map NC Center for Geographic Information and Analysis (CGIA) as main contact DWQ as technical support, field data and modeling
http://www.ncstreams.org Current NC One map Stream data availability currently only in NC mountains Catawba Lower Little Tennessee/Upper Little Tennessee/Tugaloo Nolichucky/Watauga Pigeon Seneca Tuckasegee Upper Broad Upper French Broad Upper New/Upper Yadkin
DWQ Headwater Stream Spatial Dataset Will augment NC ONE stream map Accessible through NC ONE stream map Keys to success Continue field data collection GIS support Stream ID spatial database (web-based) Continued coordination with CGIA
APPROACH AND OBJECTIVES Use LiDAR data from the NC Floodplain Mapping Program to create new topographic maps at a resolution of 5 meters (16.4 feet) Conduct field identification and mapping of 1st and 2nd order intermittent and perennial headwater stream origins in selected watersheds Build a geodatabase of study watersheds to include field data, DEMs and DEMderived data and other GIS data, e.g., slope, contributing drainage area, curvature, soils, precipitation and Ecoregion Conduct spatial analysis using GIS data and field data Use field data and other GIS data to create predictive, statistical models of streams and origins
Existing stream mapping data from Pilot Study
Contributing Drainage Areas for Stream Origins (Ephemeral to Intermittent/Perennial) Eastern Blue Ridge Foothills EPA Level IV Ecoregion Northern Carolina Triassic Outer Slatebelt Basins Piedmont Acres Quantiles Acres Acres Acres Min 0.23 0.05 1.55 10% 1.40 1.47 2.53 25% 3.32 4.56 4.87 Median 4.96 8.58 10.37 75% 6.54 29.21 16.35 90% 8.46 73.12 27.29 Max 15.85 322.27 115.95 Rolling Coastal Plain Acres 0.10 0.16 1.23 7.64 1.96 12.26 3.49 27.43 7.36 68.28 12.86 128.13 21.12 343.66
Crucial question What is map accuracy? Three accuracy measures Presence/absence - percentage of streams occupying the same valley as the field-mapped streams Length - percentage of modeled stream length by stream order that is the same as the fieldmapped stream length by stream order Flow duration - percentage of intermittent and perennial stream cells that are coincident with the same intermittent and perennial stream cells as the field-mapped streams.
Headwater Stream Model Stream location accuracy = 84% Stream length accuracy 1 st order = 58% 2 nd order = 96% 3 rd order = 98% Overall = 83% Flow duration ~ 30%
Next Steps Refinement of statistical models for Slate Belt Ecoregion Focus on flow duration accuracy Work with CGIA and DOT on a pilot study in Slate Belt Ecoregion Select several permitted projects Develop accuracy assessment based on existing field data Select proposed project Work with regulators and DOT staff on selection of preferred alternative Prepare maps for project Document cost savings with reduced field time to examine alternative alignments
Next Steps (cont.) Once document utility of process, then expand statewide Supplement existing data by Ecoregion Mountains determine accuracy of existing enhanced maps Work with local governments and other sources to collect data Lincoln County work underway Permitted projects where DWQ has confirmed stream origins such as PCS Phosphate in Beaufort County Collect field data for remaining Ecoregions Prioritize data collection by DOT, CGIA and DWQ goals Develop predictive models with known levels of accuracy Incorporate results into NC OneMap with CGIA
Questions?