Introduction and Goal Drought is a complex natural hazard that can cause substantial socio-economic and environmental impacts. Traditional methods of drought assessment often rely on ground-based precipitation measurements, which can be sparse and unevenly distributed, especially in underdeveloped regions. The Standardized Precipitation Index (SPI) is widely used for its simplicity and effectiveness in quantifying precipitation deficits over various timescales. However, the reliability of SPI calculations heavily depends on the availability and accuracy of precipitation data. Satellite precipitation products (SPPs) offer an alternative by providing comprehensive coverage and consistent data over large areas, which is particularly useful in regions with limited ground-based observations. Materials and Methods This study evaluated the application of four monthly SPPs—TRMM 3B43 (TRMM), GPM-IMERG v6 (GPM), PERSIANN-CDR (PERSIANN), and ERA5—for monitoring the 12-month SPI across 10 synoptic stations individually and in an aggregated form across the entire Khorasan Razavi province in northeastern Iran. The selected stations include Quchan, Gonabad, Kashmar, Mashhad, Neyshabour, Sarakhs, Sabzevar, Golmakan, Torbat-e Jam and Torbat-e Heydarieh. The study period spans from 2000 to 2020, covering various climatic variations and drought events. The SPI was calculated using satellite-derived precipitation data and the long-term ground-based monthly precipitation records from the synoptic stations. To assess the performance of the satellite data, several statistical metrics were computed, including the correlation coefficient (CC), root mean square error (RMSE), relative bias (RBIAS), Nash-Sutcliffe efficiency (NSE), and estimation probability (POD). These metrics provide a robust framework for evaluating the accuracy and reliability of the satellite products in replicating ground-based precipitation measurements. Results and Discussion The performance evaluation of the satellite products revealed varied results across different stations. The TRMM product demonstrated superior performance at Torbat-e Heydarieh, Quchan, Kashmar, Mashhad, and Neyshabour. It showed high correlation coefficients with ground-based data and low RMSE values, indicating its reliability in these regions. The PERSIANN product demonstrated superior performance at Golmakan. The GPM product excelled in Sabzevar, where it closely matched the precipitation patterns observed from the ground stations. On the other hand, the ERA5 product was found to be most effective in Torbat-e Jam, Gonabad, and Sarakhs, demonstrating a high degree of accuracy in these areas. The relative bias and mean error metrics also showed that the satellite products generally provided unbiased and accurate estimates of precipitation, with minimal systematic deviations from the ground-based data. The NSE values, which measure the predictive power of the models, were above 0.65 for the top-performing products, suggesting a high level of efficiency in drought monitoring. The statistical analyses from point-to-point station data yielded better results compared to the aggregated data at the provincial level. In the aggregated analysis, the highest correlation coefficient (CC=0.8), the highest Nash-Sutcliffe efficiency (NSC=0.55), the lowest root mean square error (RMSE=0.61), and the highest probability of detection (POD=0.70) were obtained from the TRMM satellite data. It seems that individual stations may provide different data compared to the entire province due to specific local conditions such as the presence of local water sources, specific vegetation cover, or human influences. When data are aggregated at the provincial level, the impact of these local anomalies increases, and the results obtained cannot accurately represent the actual climatic conditions of the province due to reduced precision and increased influence of local anomalies. Therefore, it seems that the validation of these satellite products for determining the 12-month SPI drought index should be conducted regionally. Conclusion and Suggestions In this study, the accuracy and applicability of long-term precipitation products TRMM, GPM, PERSIANN-CDR, and ERA5 for monitoring drought on a 12-month scale were comprehensively evaluated based on data from 10 synoptic stations for the years 2000-2020, both individually and aggregated at the provincial level in Khorasan Razavi. Five statistical criteria and classifications were used to evaluate and compare the TRMM, GPM, PERSIANN-CDR, and ERA5 products with synoptic station data in Khorasan Razavi. The evaluation showed that the satellite-derived precipitation products performed well in the studied areas on a 12-month SPI scale. Statistical analyses indicated that individual station data provided better results compared to aggregated data. Specifically, the TRMM product was accurate for Torbat-e Heydarieh, Quchan, Mashhad, Neyshabour, and Kashmar stations; the GPM product for Sabzevar station; the ERA5 product for Torbat-e Jam, Sarakhs, and Gonabad stations; and the PERSIANN product for Golmakan station, making them suitable for water resource management and drought monitoring. Nonetheless, given the appropriate accuracy of the satellite data, these products are considered valuable long-term datasets for studying droughts in study areas in Khorasan Razavi province. Satellite products use various data sources, and each of these products may be influenced by different environmental, geographical, and climatic factors. Therefore, it is not unexpected that their performance may vary across different stations, usually due to local characteristics and inherent differences in the algorithms used by each product. The application of satellite products on a larger scale requires the integration of data analysis methods, local calibration, and uncertainty modeling to achieve an appropriate confidence level for these products. |