Leveraging Freely Available Remote Sensing and Ancillary Datasets for Semi-Automated Identification of Potential Wetland Areas Using a Geographic Information System (GIS)

Author:
Felton, Benjamin, Civil Engineering - School of Engineering and Applied Science, University of Virginia
Advisor:
Goodall, Johnathan, Civil and Environmental Engineering, University of Virginia
Abstract:

Conducting environmental assessments for federal and state agencies is an integral part for many transportation construction projects. Wetlands are a particular environmental feature that could potentially be affected by construction projects. The identification of wetland locations can be accomplished in a variety of ways, ranging from less involved, lower accuracy methods to highly involved, higher accuracy methods. Past efforts to develop wetland identification methods are lacking in one or more of the following ways: inadequate use of ancillary data, little automation, not leveraging freely available data, excessive computation times, or requiring software not typically available within transportation agencies. This study aims to address these issues by developing a Geographic Information System (GIS)-based wetland identification tool. The aim of the tool is screening for potential wetland areas that can be further investigated by more detailed wetland identification and survey methods. Therefore, the tool is designed to minimize the number of false negatives, which are cases where the tool incorrectly designates an area as non-wetland when a wetland does in fact exist. Applying the tool to a study region with detailed wetland delineations available shows that the tool was able to identify potential wetland locations with 69.3% agreement, 24.3% false positives, and only 6.4% false negatives. Decision makers can use the prediction confidence levels generated by the tool to balance tradeoffs between the size of the area determined to be potential wetland area and the percentage of false positive prediction errors within the study area.

Degree:
MS (Master of Science)
Keywords:
wetland, mapping, GIS, remote sensing, identification
Language:
English
Rights:
All rights reserved (no additional license for public reuse)
Issued Date:
2015/07/13