Estimating Reference Evapotranspiration Using Reanalysis Data from the Google Earth Engine: A Case Study From an Arid Hydrological Region in Mexico

ABSTRACT Reference evapotranspiration (ETo) and crop coefficients (Kc) can be used for the estimation of crop water requirements. Estimating the ETo using the FAO‐56 Penman–Monteith (FAO‐56 PM) method requires various climatic variables that are not commonly available in developing countries. The emergence of long‐term assimilated reanalysis products offers an alternative for consistent climate data. However, reanalysis data (RD) must be validated before being applied regionally. This document presents the evaluation of five reanalysis datasets (CFSv2, GLDAS, NLDAS, RTMA, and the NASA‐POWER [NP]) to estimate ETo in an arid hydrological region of northern Mexico using daily information observed at 26 meteorological stations. Additionally, four different cases are examined, in which the FAO‐56 PM and Hargreaves and Samani (HS) methods are used with reanalysis and measured data. According to the results, the RTMA data show the best performance for temperature, CFSv2 for wind speed, and the NP for solar radiation. Although NLDAS contains hourly information, it has low performance. The five datasets tend to overestimate ETo. The FAO‐56 PM method combined with RD outperforms the HS method. In the study area, when observed data are not available, ETo can be estimated via the FAO‐56 PM method and RD from different sets.