Evaluation of Precipitation and Temperature: An Analysis of In-Situ Observations Versus Gridded Data within the Great Salt Lake Basin

This study evaluates the consistency between in-situ measurements and gridded datasets for precipitation and temperature within the Great Salt Lake Basin, highlighting the significant implications for hydrological modelling and climate analysis. We analysed five widely recognized gridded datasets: GRIDMET, DAYMET, PRISM, NLDAS-2, and CONUS404, utilizing statistical metrics such as the Pearson Correlation Coefficient, Root Mean Square Error (RMSE), and Kling-Gupta Efficiency to assess their accuracy and reliability against ground truth data from 30 meteorological stations. Our findings indicate that the PRISM dataset outperformed others, demonstrating the lowest median RMSE values for both precipitation (approximately 1.9 mm/day) and temperature (approximately 0.9°C), which is attributed to its advanced interpolation methods that effectively incorporate orographic adjustments. In contrast, NLDAS-2 and CONUS404, despite their finer temporal resolutions, showed greater error variability and lower performance metrics, which may limit their utility for detailed hydrological applications. Through the use of visual analytical tools such as heatmaps and boxplots, we were able to vividly illustrate the performance disparities across the datasets, thereby providing a clear comparative analysis that underscores the strengths and weaknesses of each dataset. The study emphasizes the need for careful selection of gridded datasets based on specific regional characteristics to improve the accuracy and reliability of hydro climatological studies and supports better-informed decisions in climate-related adaptations and policy-making. The insights gained from this analysis aim to guide researchers and practitioners in selecting the most appropriate datasets that align with the unique climatic and topographical conditions of the Great Salt Lake Basin, enhancing the efficacy of environmental forecasting and resource management strategies.