Comparison of Simulated Cotton Evapotranspiration with Eddy Covariance Measurements.

There is a new effort underway in the modeling community aimed at improving the ability of crop models to accurately simulate the water use of agricultural crops. Currently, evapotranspiration (ET) is estimated in crop models using methods such as the Penman-Monteith Equation, Priestley-Taylor Equation, and soil water balance residual approach, to name a few. Rigorous testing of ET simulations by crop models is not widely done because actual measurements of field-scale ET are not widely available. Since crop growth is highly dependent on water in semi-arid and arid regions, accurate simulation of ET is critical for good crop growth simulations. Cotton2K and CSM-CROPGRO-Cotton are the two most widely used simulation programs for modeling cotton. The cotton version of the simpler model GRAMI in simulating yields within our agricultural region is also used in this study. Daily measurements of crop ET are needed for testing these models. The eddy covariance (EC) method is the best method for making actual field-scale measurements of ET under commercial crop production conditions. We have collected continuous daily measurements of ET over the entire growing season for irrigated and dryland cotton fields in the Texas High Plains region near Plainview, TX. We have two years of ET data available for irrigated cotton (2013 and 2014) and one year of data for dryland cotton (2014). We are collecting similar data for the 2015 growing season. We will use these data sets as the basis for rigorously testing the ET simulations of cotton models under commercial crop production conditions. In this presentation, we will present results from comparison on the performance of the models in simulating ET.