Comparison of Alternative Approaches for Wetlands Mapping: A Case Study from three U.S. National Parks
2016
Sharpe, Peter J. | Kneipp, Gregg | Forget, Andrew
We performed an assessment of U.S. Geological Survey/National Park Service (USGS/NPS) vegetation mapping versus National Wetland Inventory (NWI) estimates of wetland occurrence and extent for three national parks, each having a different NWI mapping scale (1:40,000, 1:58,000, and 1:80,000). Our prediction was that the USGS/NPS mapping would be significantly more effective than NWI in predicting total wetland area within each park, and would commit fewer errors of omission and commission. For use as a control group, each park had recent wetland field determination data collected in accordance with U.S. Army Corps of Engineers protocols. Contrary to our prediction, mean NWI wetland area estimates were more accurate than USGS/NPS mapping at the 1:40,000 and 58,000 scales. At the 1:80,000 scale, NWI and USGS/NPS estimates were similar. NWI wetland area estimates were not significantly different (α = 0.05) from the control data at two of the study parks, whereas USGS/NPS estimates were significantly larger than the control group at two of the three parks. This research highlights the relative strength of NWI mapping for landscape level wetland analysis, and the need to support remote sensing data by allocating field resources for accuracy assessment in specific areas based on management goals.
显示更多 [+] 显示较少 [-]