Comparison of computed and measured evaporation in the Colorado River basin in Texas

Computational evaporation techniques have been available for decades. The techniques routinely are used in agricultural management, but are not commonly used in water supply management. With increasing availability of weather observations online, implementing evaporation techniques in water supply management is now a simple and low-cost alternative to field measurement. Here we demonstrate automated computed evaporation for reservoirs in the lower Colorado River basin in Texas. We compare two computational methods, Van Bavel hourly method and Penman-Monteith daily, to the historical pan evaporation method for Lake Travis in 2015 and 2016. Both methods reproduce gross pan evaporation with negligible differences and within the measurement error of pan evaporation, but it is unclear which of the methods is closer to the true value. This year, the techniques have correctly captured record evaporation rates in the basin. The Lower Colorado River Authority (LCRA) is beta testing basin-wide daily computed evaporation. LCRA also is considering computed evaporation as an alternative to automated pans and pan replacements. Because of the low cost of implementation, evaporation can be easily computed for reservoirs not previously measured with pans. Later in 2016, we will compare the techniques to newer eddy-covariance methods.