Evaluating The Accuracy of Gridded Climate Datasets for Precipitation, Surface Air Temperature, and Sea Surface Temperature in Central Java, Indonesia

<span lang="EN-AU">Studies of climate information that rely on accurate and reliable data are essential in hydrometeorological monitoring, early warning, and climate change impacts in areas with varied topography and limited observation data, such as Central Java, Indonesia. This study aims to assess the accuracy of gridded satellite and reanalysis on three main variables. Precipitation was analyzed utilizing CHIRPS, ERA5 Precipitation, and GSMaP products; surface air temperature (SAT) was assessed with ERA5-Land, FLDAS, and AIRS; and sea surface temperature (SST) was evaluated using OSTIA, RAMSSA, and GAMSSA. Observational data from six BMKG stations and iQuam functioned as the reference standard. The datasets were extracted using bilinear interpolation and evaluated using a bias, mean absolute error (MAE), mean absolute percentage error (MAPE), symmetric mean absolute percentage error (SMAPE) for precipitation, and root mean square error (RMSE). The evaluation showed that CHIRPS performed better estimation with the lowest RMSE and SMAPE (17.20 mm/day; 111.42 mm/month; 96.97% daily; 54.09% monthly) compared to ERA5-Precipitation and GSMaP. ERA5-Land in SAT showed better accuracy in MAE and MAPE of 1.2°C and <10% at most locations. For SST evaluation, OSTIA demonstrated the highest agreement with iQuam, showing RMSE of 0.246°C and MAPE of 0.552% in the Southern Sea, while GAMSSA recorded the highest errors across all zones. This study presents a variety of gridded dataset performances based on scale and time to illustrate the importance of validation against observational data. These results can guide researchers in processing the right dataset collection in climate applications in tropical ocean areas.</span>