Short URL

Hydrogeomorphic Mapping of Arkansas Wetlands

 

CAST and the Arkansas Multi-Agency Wetland Planning Team (MAWPT) have collaborated for a number of years on a GIS-based analysis of wetlands around the state, principally utilizing data layers that serve as indicators of vegetation type, inundation, and soil type. This approach has generated a binary "wetland / not wetland" result, which has proved to be useful in basic land use management. Over the last few years, the MAWPT has also worked on developing and applying a more precise, site-specific analysis of wetlands. The Hydrogeomorphic (HGM) approach to assessment of wetland functions relies upon site-specific evaluation of geomorphic setting, water source, and hydrodynamics. In this project, CAST worked with biologists and geomorphologists of the MAWPT to model the classes and subclasses of wetlands, as defined by the HGM approach, of the portion of the the White River watershed known as the Western Lowlands. CAST used geospatial technology and improved data to determine the location and extent of these wetland classes and subclasses.

3- and 5-year floodplains

Since hydrodynamics are a key element of the HGM classification scheme, it was critical to be able to accurately identify areas of frequent inundation. Although high-density transect-based models would produce the most accurate understanding of inundation patterns, the cost is prohibitive. Instead, satellite imagery scenes from Landsat 5 and 7 were correlated with historic river gage station data, using the Point Over Threshold Method (POTM) to determine flooding frequencies. Since members of the MAWPT HGM team had empirically linked 2- and 5-year flood frequencies to distinct changes in wetland classes, these were the frequencies which were mapped.

A vital step in correlating imagery scenes to gage station data is determining the "bank full" stage of the gage, i.e. the level at which the river exceeds its banks. Unfortunately, many of the gage stations in this study area do not have these data available, and thus could not be used. As these data become available, the overall accuracy of these particular data layers could be improved.

A flow chart diagramming the  processing of the imagery can be viewed here.

HGM classification of wetland classes and subclasses The actual HGM mapping process was an iterative one. Along with the inundation data, SSURGO soils, geomorphology, and additional hydrology data sets were incorporated into a geographic information system (GIS). Classification keys provided by the MAWPT scientists were used to create rule-based GIS models, which were provided in return to the MAWPT team. After ground-truthing the data, comments and updated classification keys were sent back to CAST for the next round of modeling. Overall, this process went through five cycles to produce a series of maps that, given the limits of available data, reflect the classes and subclasses of wetlands found on the ground.

Although this project was completed in 2007, another round of HGM mapping is under progress for the St. Francis River Wetland Planning Area. Completed reports for the work in the Western Lowlands can be found at the MAWPT web site: http://www.mawpt.org