Please use this identifier to cite or link to this item:
|Title:||Comparisons of digital terrain data for wetland inventory on two Alaskan army bases|
|Authors:||Environmental Laboratory (U.S.)|
Strategic Environmental Research and Development Program (U.S.)
Melloh, Rae A.
Racine, Charles H.
Sprecher, S. W.
Greeley, Nancy H.
Weyrick, Patricia B.
Landsat Thematic Mapper
Fort Wainwright, Alaska
Fort Richardson, Alaska
|Publisher:||Cold Regions Research and Engineering Laboratory (U.S.)|
Engineer Research and Development Center (U.S.)
|Series/Report no.:||Special report (Cold Regions Research and Engineering Laboratory (U.S.)) ; 99-15.|
Abstract: The nation’s military installations encompass undeveloped lands that have become increasingly important as wildlife habitats. Resource managers of the installations need wetland inventories to improve stewardship of these lands. Digital geographic data are readily available to land managers. The use of these data to inventory wetlands has not been demonstrated. As part of a project to integrate wetlands into the ITAM (Integrated Training Area Management) program for managing Army lands, wetland inventory methods using existing digital geographic information for two terrains on Army installations in Alaska were explored: (1) glacial moraine depressions and estuarine marsh on Fort Richardson, and (2) discontinuous permafrost and taiga forest on Fort Wainwright’s Yukon Command training site. Our results show that (1) existing geographic data used to infer wetland locations (Landsat Thematic Mapper [TM], National Wetland Inventory [NWI] maps, and hydric soil maps) only partly agree, and (2) optimum Landsat TM band combinations for wetland inventory vary on a site-specific basis. Landsat TM classifications (unsupervised) of Fort Richardson wetlands compared reasonably well (0.73 Kappa Index of Agreement [KIA]) with the NWI map as long as the band combinations included at least one visible and the near-infrared wavelength band (e.g., bands 3, 4, and 5 or bands 2, 3, and 4). The Fort Richardson hydric soils map indicates more extensive wetlands than indicated by the NWI (0.64 KIA). The Landsat TM classification could be made to agree fairly well with the NWI map (0.73 KIA). At Fort Wainwright, use of the thermal wavelength band (6, 4, and 2 composite) improved Landsat TM classification agreement with the NWI (0.67 KIA) because of warmer apparent brightness temperatures of lowland wetland sites compared to upland forested sites. Topographic position in the taiga forest plays a strong role in determining soil moisture, dominant vegetation, and whether or not the site is underlain by permafrost; therefore, a wet terrain map derived from a digital elevation model agreed nearly as well to the NWI map (0.64 KIA) as did the Landsat TM classification (0.67 KIA). Existing geographic information can serve as an initial wetland map. However, accurate wetland maps will require field mapping.
|Rights:||Approved for public release; distribution is unlimited.|
|Appears in Collections:||Special Report|
Files in This Item:
|SR-99-15.pdf||2.85 MB||Adobe PDF|