Optical Remote Sensing As a Tool for Determining Crop Coefficients and Reducing Uncertainty in Evapotranspiration Estimates.

Measured or modeled values of evapotranspiration (ET) are routinely used to quantify crop water use and define irrigation needs. There are a variety of ways to obtain ET values for a given canopy. One common method is to calculate grass reference ET (ET0), then, assuming well-watered conditions, ET0 values are converted to actual ET (ETa) by multiplication of a scalar. This scalar is referred to as a crop coefficient (Kc), where Kc = ET0/ETa. Meteorological variables required for calculating ET0 are readily available from user-friendly and relatively inexpensive environmental sensors. Kc on the other hand can be more difficult to quantify because it varies depending on canopy architecture, height, ground cover, phenology, and species, among other variables. Empirically-derived Kc values exist for a variety of different crop species; however, determining the correct Kc value requires several assumptions about canopy properties, which can be especially difficult during time periods when canopy characteristics are rapidly changing. Furthermore, the “look-up table” approach assumes uniformity of canopy characteristics across the spatial domain. Alternatively, several studies have demonstrated that remotely sensed optical measurements, especially of surface reflectance in visible and near infrared regions of the electromagnetic spectrum, offer a way to characterize crop dynamics that are related to variations in Kc. Proximal optical sensors have limited spatial coverage, but offer near-continuous measurement of canopy reflectance so that variables such as growth stage can be monitored. Satellite, and increasingly UAV, sensors have limited temporal resolution, but offer unmatched spatial coverage and opportunities for upscaling ground-based point measurements. Here we provide an overview of the progress that has been made using optical remote sensing for obtaining Kc estimates and for parameterizing ET models. Finally we discuss ways that this technology can further improve ET measurement and modeling in the future.