A Review of APHRODITE Performance in Asia: Current Status and Research Concerns

Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation (APHRODITE) is extensively utilized for climatological analysis, hydrological modelling and climate model bias correction. This review evaluates the performance of APHRODITE in Asia by analyzing 111 studies published from 2011 to 2023. The findings show that APHRODITE generally performs better during the wet season across most evaluated regions and shows higher correlations with monthly precipitation estimates compared to daily precipitation. However, it often underestimated light and extreme precipitation while overestimating moderate precipitation. Additionally, APHRODITE encounters difficulties in estimating precipitation in regions with complex topography, like the Himalayan range, the northern Tibetan Plateau and desert areas. Despite these challenges, APHRODITE outperforms many other precipitation products in hydro-climatic simulations. With the release of APHRODITE-2 and the upcoming APHRODITE-3, further validation studies are needed, particularly for West Asia and island nations like Indonesia, the Philippines and Sri Lanka. It is recommended to use various reference sources, such as commercial gauges and ground-based radar, for validation or to apply the triple collocation approach when high-quality gauges are unavailable. A comprehensive validation framework has been proposed to address the issue of data independence, incorporating common continuous metrics (CC, RMSE, and RBias), categorical metrics (POD, FAR, and CSI), precipitation distribution functions, extreme value analysis, and ranking systems for comparing APHRODITE with other products. Ideally, validation should incorporate hydrological assessments using models such as the Soil and Water Assessment Tool (SWAT) or the Variable Infiltration Capacity (VIC) models, and evaluate with on R², NSE and RBias metrics at daily, monthly and annual scales. Future studies should explore how various comparison approaches, such as point-grid, grid-grid and direct comparisons, could affect validation results. Lastly, improving data sharing practices and installing gauges in less accessible areas is crucial, as the quality of APHRODITE relies on these input gauges.