Mapping agricultural wetlands in the Sacramento Valley, USA with satellite remote sensing

摘要: Information on rice paddy practices and habitat is required by decision makers monitoring protocol to manage the agricultural wetlands of Sacramento Valley, California. Satellite remote sensing imagery have been able support these needs with accurate maps extent delineation winter flooding. However, habitat, hydroperiod, spatiotemporal dynamics not thoroughly carried out in region. The objective this research application was evaluate cost-efficient synthetic aperture radar optical for mapping hydroperiod management. Ground-truth field data from Glenn-Colusa Irrigation District used test a suite indices differentiate among wetland types irrigation practices. A Classification And Regression Tree approach that utilized random forest algorithm built map Valley. Results show sensitive vegetative development surface water (Landsat NDVI LSWI) along Synthetic Aperture Radar backscatter values (PALSAR HH sigma nought σ° dB) were most useful across habitats. Overpass timing management key factors sensor index selection. Nearly 93 % (185,494 ha) paddies Valley underwent wet seeding half (90,168 ha) actively cultivated area flooded during winter. Eight percent (16,420 ha) received at least one crop season. challenging separate fallow different frequencies inundation due permitting policies duck promotion. Overall, results emphasize utility satellite making as cost-effective tools supporting Monitoring, Reporting, Verification protocol.