Improving the Performance of Global Rainfall Forecasting Systems in Different Climate Areas of Iran Using Quantile Mapping Method

Precipitation is one of the main components of flood, drought and water resources warning studies, hence, its quantitative prediction is of the great importance. The increasing development of computing and satellite technologies and remote sensing in recent years has led to the development of several meteorological forecasting models, of which the TIGGE database with a large number of powerful forecasting models, is the most important. The aim of this study was to evaluate the performance of all available numerical models in the database to predict daily precipitation in 38 synoptic stations located in different climates of Iran. In addition, removing biases from raw datasets using Quantile Mapping (QM) method is another objective of this study. Results showed that in humid, semi-humid, Mediterranean and Arid climate zones (mostly includes the southwest, northwest and northeast parts of Iran), most of the prediction models are highly correlated with ground observations, while in semi-arid and extra-arid regions the correlation coefficient (CC) between the forecasted and observed datasets is very low. For example, the CC and RMSE values obtained from ECMWF and METEO centers in most parts of the country are higher than 0.6 and lower than 4 mm/day, respectively, while the performance of CMA and CPTEC models is not remarkable and leads to the weak results. Also, evaluation of the corrected precipitation values by QM method indicates that there is a significant improvement in the performance of most prediction systems. Findings in extra-arid, arid, and Mediterranean zones demonstrate an increase in CC value, averagely about 20%. Moreover, the results depicted that by removing biases from the raw datasets, the performance of numerical weather prediction (NWP) models in estimating the low and high precipitation events is improved and this issue further increases the applicability of precipitation forecasting systems in flood warning systems and water resources management.