Estimating Evapotranspiration in the Caribbean Region using an Operational Water and Energy Balance Algorithm

In 2009, collaboration was established between the University Puerto Rico-Mayaguez (UPRM) Campus and the University of Alabama-Huntsville (UAH) Campus. UAH began providing 1-km resolution hourly and daily solar insolation, derived from National Oceanic and Atmospheric Administration (NOAA) Geostationary Environmental (GOES) satellite data. Using these GOES insolation data, UPRM developed an operational water and energy balance algorithm (driven principally by solar radiation) that provides daily gridded hydro-agro-climate variables for Puerto Rico, and to a limited extent for the U.S. Virgin Islands, Hispaniola, Jamaica and Cuba. The algorithm provides estimates of reference evapotranspiration (ET) using three methods (Hargreaves-Samani, Penman-Monteith and Priestley Taylor) and actual ET, based on a land surface energy balance.

The algorithm was useful for providing regular updates on soil and water-related variables during the severe drought of 2015 in Puerto Rico. Updates included agricultural rainfall deficit (rainfall minus reference ET), ratio of actual to reference ET, soil moisture and soil saturation, crop stress factor and Bowen ratio.

A recent improvement to the algorithm involved the near-real-time calibration of the climate variables that are used for input in the model. Temperatures are initially derived from lapse rate models developed for Puerto Rico. Wind speed is obtained from NOAA's National Digital Forecast Database (NDFD). The near-real-time calibration has resulted in a marked improvement in the modeled air temperature, humidity and wind speed data, consequently leading to an improvement in the ET estimation over the island.