Evaluation of NASA POWER and ERA5-Land for estimating tropical precipitation and temperature extremes

Abstract

Long-term gridded climate data, NASA Prediction of Worldwide Energy Resources (NASA POWER) and the fifth generation of European ReAnalysis Land Component (ERA5-Land), are important alternatives to gauges, but their effectiveness in capturing tropical precipitation and temperature extremes has not been thoroughly studied. Therefore, this study aims to evaluate the performance of NASA POWER and ERA5-Land over the Kelantan River Basin from 1981 to 2020 with 19 gauges in the aspects of climatology characteristics, overall assessment, precipitation detection capability, probability distribution function (PDF), extreme indices calculation, and the 2014–2015 flood. The important findings are as follows: (1) NASA POWER and ERA5-Land are able to reasonably capture the climatology patterns of precipitation, maximum, and minimum temperatures, with the latter performing slightly better; (2) Both products have a better correlation with gauges over coastal regions than inland mountainous regions, but they tend to underestimate precipitation and maximum temperature and overestimate minimum temperature; (3) With high POD values and moderate FAR and CSI values, they are able to detect precipitation days more effectively than non-precipitation days; (4) The PDF assessment reveals that both products overestimate moderate precipitation (1–20 mm/day) and underestimate no/tiny (0–1 mm/day), heavy (20–50 mm/day), and violent (greater than50 mm/day) precipitation; (5) With the exception of CWD and R10mm, both products tend to significantly underestimate most of the precipitation extreme indices, while a better correlation with gauges can be found in the calculations of CDD, R10mm, R20mm, TXx, and TNn; (6) Both products have a higher correlation and a better precipitation detection ability during the 2014–2015 flood, but they significantly underestimate the amount of precipitation. The findings show that NASA POWER and ERA5-Land offer valuable climate information for locations without gauges, but it is recommended to apply bias correction prior to applying them in this tropical basin.