H11E-1381
Floodplain Mapping and Wetland Connectivity to Lotic Aquatic Systems of Indiana and North Dakota, USA

Monday, 14 December 2015
Poster Hall (Moscone South)
Charles Lane1, Alexander Hall2, Ellen D'Amico2, Nikhil Sangwan3 and Venkatesh Merwade3, (1)Environmental Protection Agency Cincinnati, Cincinnati, OH, United States, (2)CSS-Dynamic Corporation, Cincinnati, OH, United States, (3)Purdue University, West Lafayette, IN, United States
Abstract:
In a recent literature review of over 1,350 publications, the U.S. Environmental Protection Agency concluded that wetlands in floodplains are physically, chemically, and biologically integrated with rivers and can improve downstream water quality. However, current knowledge on the extent of floodplains is limited, with national-scale maps unavailable with the exception of coarse-scale Federal Emergency Management Agency (FEMA) Flood Insurance Rate Maps (FIRMs). Recently, novel approaches to rapidly and inexpensively map floodplains using widely available soil survey data have been developed, as have approaches using hydrogeomorphic metrics. To better understand wetland connectivity to other aquatic systems, we contrasted FIRMs with soil-based, hydrogeomorphic, and simple distance-based metrics in watersheds of Indiana and North Dakota. Soil-based, hydrogeomorphic, and distance-based measures were particularly useful in determining connectivity. For example, in the White River of Indiana, these three measures connected 74-86% of the watershed’s wetland area while the FEMA FIRMs only connected 58% of the watershed’s wetland area, with omissions coming from stark demarcations of FIRMs at seemingly arbitrary locations. Overall, we found that FEMA FIRMs were generally unsuitable for use in determining watershed-scale floodplains, as FIRMs tended to focus on urban areas at the expense of agricultural (or non-urban) regions. Distance-based connectivity measures likely over-estimated floodplain extent through the “creation” of non-existent floodplains on low-order streams. As with any method, the accuracy of the hydrogeomorphic and soil-based measures was highly dependent on the resolution and quality of the input data. However, these two methods in particular appear to hold great promise for accurately extending our knowledge of floodplains throughout the conterminous United States.