Publications

Torbick, Nathan; Salas, William (2015). Mapping agricultural wetlands in the Sacramento Valley, USA with satellite remote sensing. WETLANDS ECOLOGY AND MANAGEMENT, 23(1), 79-94.

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

DOI:
10.1007/s11273-014-9342-x

ISSN:
0923-4861