Highlighting the Use of Remote Sensing of Evapotranspiration for Agricultural Drought Early Warning and Water Resources Management

Evapotranspiration (ET) is an essential hydrological component for many applications including, more recently, agricultural drought early warning and, traditionally, agricultural water resources management. This analysis highlights the use of the ET in these applications. For the former, the Evaporative Stress Index (ESI) has been recently developed as an indicator to monitor agricultural drought. The ESI was developed based on ALEXI model as the ratio of actual to reference ET estimates. In a collaborative effort, many countries in the Middle East and North Africa (MENA) region that include Morocco, Tunisia, Jordon, and Lebanon, are currently seeking to improve their ability to monitor agricultural drought for adaptation and planning purposes. The ESI is currently being evaluated for its ability to provide a clear depiction of the onset, duration and severity of historic drought events in Tunisia. Compared to other indicators such as the Standardized Precipitation Index (SPI). The performance of the ESI is evaluated against crop yield data, precipitation, and in-country interviews with drought monitoring experts and agricultural producers.

For the latter, ET can be used in providing estimates of consumptive water-use. There is presently many different remote sensing of ET models that provide ET estimates with relatively different levels of accuracies. This could ultimately result in different agricultural water management practices associated with relative levels of uncertainties. To minimize such uncertainties, an inter-comparison of ET models estimate was conducted as part of a USGS effort to use ET estimates in water resources management in the Western US with the aim to develop guidelines and specifications for the use of ET by water managers. This effort was funded by the USGS as part of the Sustain and Manage America’s Resources for Tomorrow program (WaterSMART) program to provide acceptable and agreeable set of methods and ET estimates by a wide range of users and decision makers. The models being compared include the ReSET, METRIC, SEBS, SSEBop, and DisALEXI models. These models were evaluated at two climatically different regions in the US; irrigated and rainfed agricultural fields in the Palo Verde Irrigation District (PVID), California and Mead, Nebraska. Estimates of daily and seasonal actual ET, which were obtained using Landsat data, were compared with ground-based Bowen ratio and eddy covariance. Analysis of water balance is also carried out to highlight the performance of the different models in providing seasonal consumptive water-use.