Comparison of Evapotranspiration Estimates from Remote Sensing, Crop Coefficient Models, and Eddy Covariance Stations for Celery in California

California's agricultural community is facing the combined challenges of increasing interannual variability in precipitation and limited availability of water resources for irrigation. Recent policy changes have led to increased interest in the use of evapotranspiration (ET) data in irrigation scheduling to increase on-farm water use efficiency. Remote sensing offers a promising approach for rapid, continuous wide-area mapping of ET, but there is a need to validate remotely-sensed ET estimates across the diversity of crops grown in California.

In this study, we evaluated remotely sensed estimates of crop ET and associated crop water requirements from NASA's Satellite Irrigation Support (SIMS) system, and the UCANR CropManage irrigation and nutrient management software for a celery crop in Soledad, CA. We measured actual evapotranspiration (ET_a) from the field using two flux stations: an open path eddy covariance (OPEC) system and an energy balance residual (REB) station. We also measured applied irrigation and soil moisture at the site. We compared ET data from the two flux stations with daily ET data from SIMS, which uses a remote sensing approach, and CropManage, which uses modeled fractional cover to drive a crop-coefficient model and calculate daily ET.

We present results from the comparison between the measured and modeled ET, and describe implications for future research to improve accuracy of remotely sensed ET. Results to date from REB and OPEC measurement show good agreement, with a daily average difference of 4%. The mean daily energy balance closure for the OPEC station is 85%.

This study highlights the potential utility of remote sensing approach to monitor ET and provides another reference point for the community working to develop accurate and cost-effective tools to support growers in optimizing irrigation management.