Catalog Item

S&T Project 1867 Final Report: Seasonal/Temporary Wetland/Floodplain Delineation using Remote Sensing and Deep Learning

Seasonal wetlands are an important habitat for many aquatic species, including juvenile anadromous fish. In the past, the delineation of seasonal wetlands has been very limited and often inaccurate by traditional methods used by Reclamation. A new methodology was created to automatically delineate seasonal wetlands from satellite imagery using machine learning methods. While additional research is needed to verify the accuracy of the results, this methodology has the potential to identify seasonal wetlands over a large area at a much lower cost than traditional methods.
Download File Opens in new window
Generation Effort S&T Project 1867: Seasonal/Temporary Wetland/Floodplain Delineation using Remote Sensing and Deep Learning
Location Name Central Valley
Type Uploaded file(s)
File Type PDF
Publisher Bureau of Reclamation
Publication Date Friday, March 12th, 2021
Update Frequency not planned
Last Update Tuesday, May 18th, 2021

Disclaimer

The findings and conclusions of this work are those of the author(s) and do not necessarily represent the views of the Bureau of Reclamation.