An Intercomparison of Evapotranspiration Estimation Methods for the Godomey Well Field in Benin, West Africa

The Godomey well field supplies groundwater for Cotonou, the largest city in Benin, West Africa. Due to the proximity of the wells to the Atlantic Ocean (5 km north of the ocean) and to Lake Nokoue, a shallow lake with high levels of chloride, the wells are threatened by saltwater intrusion. Ongoing efforts aim to characterize this groundwater system to provide management and sustainability information. As part of this effort, the goal of this study was to determine whether remote sensing of evapotranspiration provides a unique alternative to overcome data limitations of modeling in a developing country.;This study utilized three methods to estimate evapotranspiration (ET) for use in an existing groundwater model for the area. ET methods included: a crop coefficient method that used remotely sensed Normalized Difference Vegetation Index (NDVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS), Complementary Relationship Areal Evapotranspiration (CRAE), and evapotranspiration from the Noah Land Surface Model (LSM) provided in the Global Land Data Assimilation System (GLDAS). The CRAE model estimated the highest ET in the study area for the years 2001-2010. The Noah LSM estimated the lowest ET, while the MODIS NDVI crop coefficient method provided intermediate ET estimates. Although the methods varied in magnitude of ET, timing of seasonal changes in ET was consistent for the three methods and reflected the weather variation in the study area. The MODIS NDVI crop coefficient method is practical for use in developing countries because it requires few data inputs, uses accessible satellite data, has a low cost, and is relatively simple to apply.;The final products of this research include spatially distributed actual ET for the groundwater model study area at monthly intervals from 2001-2010. Preliminary calculations utilizing limited precipitation data suggest that the study area may be a zone of net groundwater discharge, with some recharge occurring near the coast and Lake Nokoue. The ET products created in this study are intended for future use with runoff and precipitation data to determine spatially distributed recharge or discharge for the groundwater model. The MODIS NDVI crop coefficient method provides a unique opportunity for improving hydrologic budgets in developing communities that are data limited.