Abstract
Surface water is crucial to ecosystems in Alaska. Precisely mapping the dynamic surface water extent in this region is required by a wide range of environmental studies. However, most existing inundation products cannot reveal the distribution of surface water in mainland Alaska at high spatiotemporal scales. To bridge this gap, this study developed a framework to generate subpixel surface water fraction (SWF) maps from the 8-day VIIRS surface reflectance composites at the 1 km resolution through a random forest regression. Assessment of map accuracy resulted in an r2 value of 0.839 and a root mean square error (RMSE) of 12.17%. With the adoption of a proper terrain shadow preprocessing procedure and more training samples in rugged terrain, the developed framework can be easily extended to produce accurate time-series SWF maps over the entire Arctic-Boreal region on a weekly basis, particularly during the summer season.
Original language | English (US) |
---|---|
Pages | 6120-6123 |
Number of pages | 4 |
DOIs | |
State | Published - 2021 |
Event | 2021 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2021 - Brussels, Belgium Duration: Jul 12 2021 → Jul 16 2021 |
Conference
Conference | 2021 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2021 |
---|---|
Country/Territory | Belgium |
City | Brussels |
Period | 7/12/21 → 7/16/21 |
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- General Earth and Planetary Sciences
Keywords
- Alaska
- SWF
- VIIRS
- random forest
- surface reflectance