Satellite-Based near Real-Time Assessment of Water Requirement of California Crops.

A common method for estimating crop water requirement is to multiply daily reference evapotranspiration (ETo) calculated from weather stations by a generic crop coefficient (Kc) from historical data.  The adoption rate of this method by California farmers is low because the historical Kc values may not adequately reflect the actual crop conditions.  A near real-time determination of water status of crops on the ground is needed.  The objective of this research was to develop an approach for near real-time assessment of crop water demands using both satellite and ground-based climate data.  Field experiments were carried out from April to October 2008 in a 27 by 35 km area southwest of Fresno, CA with measurements taken for fractional canopy cover of 20 row, tree, and vine crops from five large farming operations on the same days of LandSat overpass.  Correlations were made between the canopy cover data and NDVI values from the LandSat –5 imagery.  Using Kc and canopy cover relationships from an earlier study, Kc values for each measurement field on each measurement date were estimated.  Actual crop ET (ETc) values were computed by multiplying Kc with daily ETo estimates from a local weather station.  The initial results on predicted ETc were reasonable compared to historical norms.  Remaining challenges for the satellite-based ET method include validations over large areas such as using an eddy covariance method; limitations on satellite resolutions, e.g. temporal - 16 day/30m LandSat vs. 2 day/250m for MODIS; improved accounting for soil evaporation and stomatal regulation; further examination of canopy cover-Kc relationships for a variety of representative crop types.  Final adoption by farmers will require convenient information retrieval mechanisms and additional effort from outreach and extension specialists.