Evaluation of Multivariate Rainfall Disaggregation Performance Using MuDRain Model (Case Study: North East of Hormozgan Province

High-resolution spatial and temporal precipitation data are essential for water engineering studies, hydrological modeling, and flood risk assessment, especially in tropical regions with complex rainfall patterns. Due to the lack of data, rainfall disaggregation is an important tool. In this study, the performance of multivariate rainfall disaggregation using MuDRain model and the effect of hourly correlation among stations on simulation accuracy in Hormozgan province were investigated. Comparison between observed and simulated hourly time series showed that the model evaluates the amount of daily precipitation accurately, but in most cases it simulated extreme amounts of precipitation less than the actual amounts. Furthermore, the enough number of heavy rainfall events has not been generated. Comparison of the results of selected dates with the highest rainfall showed that the Correlation Coefficient (R) and Nash–Sutcliffe (NSE) ranged from 0.1898 to 0.9319 and 0.0319 to 0.7251 respectively. Comparison of the hourly correlation impact showed that the accuracy of the model in simulating hourly precipitation was higher for time series having higher mean hourly correlation and the coefficients of R and NSE were 0.7816 and 0.5856, respectively, while these coefficients for time series with lower hourly correlation were 0.5155 and 0.2655 respectively. Generally, this model can be used with more confidence for areas with very high hourly correlations, in this case, the spatial correlation of the stations becomes an advantage, because utilizing the available hourly rainfall data in adjacent stations, it is possible to create series of realistic hourly rainfall at a desired station.