Assessing NASA POWER and TerraClimate datasets for climate analysis and aridity characterization in Saïss basin, Morocco

Effective management of climate events requires a solid understanding of past and present conditions to mitigate climate change impacts, especially in regions with limited weather station coverage where access to accurate hydrometeorological data is crucial.This study evaluates the performance of two global climate databases, NASA POWER (PN) and TerraClimate (TC), in the Saïss basin (Morocco), by comparing their estimates with ground observations from Meknes, Chelihate, and Fez.The analysis focuses on precipitation, maximum and minimum temperatures, and extreme climate indices.Results reveal strong correlations between satellite and station data, with performance varying by location.In Meknes, PN provides higher correlation for precipitation (r = 0.9833), while TC shows lower bias (-7.76 mm/month) and a better ratio (0.91).At Chelihate, PN outperforms TC in correlation (r = 0.9773) and bias, whereas TC achieves a more favorable ratio.In Fez, PN shows excellent correlation (r = 0.9935) but higher bias, with TC yielding a better ratio.For temperature, PN achieves higher correlations at all stations, while TC provides lower biases and more balanced ratios.The De Martonne aridity index indicates a predominance of semi-arid climate, significant interannual variability, and a trend toward increasing aridification over the last two decades.Overall, PN and TC provide valuable climate information in data-scarce regions, but biases, particularly in extreme precipitation, highlight the need for local validation before application in water management, agriculture, and climate impact assessments.