Correcting Thornthwaite potential evapotranspiration using a global grid of local coefficients to support temperature-based estimations of reference evapotranspiration and aridity indices

Abstract. Thornthwaite's formula is globally an optimum candidate for large-scale applications of potential evapotranspiration and aridity assessment at different climates landscapes since it has lower data requirements compared to other methods especially from the ASCE-standardized reference (formerly FAO-56), which most data-demanding method commonly used as benchmark method. The aim study develop a global database local coefficients correcting monthly Thornthwaite (Ep) using (Er). validity will be verified by testing hypothesis that correction coefficient, integrates mean effect wind speed, humidity, solar radiation, can improve performance original formula. was developed gridded temperature, rainfall, Er period 1950–2000 30 arcsec resolution (∼ 1 km Equator) freely available climate geodatabases. were produced partial weighted averages Er/Ep ratios setting ratios' weight according magnitude excluding colder months with values or Ep < 45 mm per month because their ratio becomes highly unstable low temperatures. validation made raw 525 stations Europe; California, USA; Australia including up 2020. procedure showed corrected Eps led reduction RMSE 37.2 30.0 m−1 step estimations 388.8 174.8 yr−1 annual formulas also evaluated use in UNEP (United Nations Environment Program) indices respective estimated Er. analysis stations, results increased accuracy detecting identical classes 63 % 76 case classification 93 classification. both extremely improved non-humid classes. factors support index minimum (i.e., temperature) locations where climatic are limited. grids this archived PANGAEA assessed following link: https://doi.org/10.1594/PANGAEA.932638 (Aschonitis et al., 2021).