Nowadays, climate and rainfall pattern studies are vital for the construction of hydraulic structures, water resources management and allocation. Intensity-Duration-Frequency (IDF) curves are a valuable tools for calculating flood discharge as well as a basis for designing hydraulic structures and urban infrastructures such as drainage and culvert design. These curves are obtained using rainfall data recorded at different duration. Despite a lot of rain gauges around the world, due to rather sparse spatial distribution, it is not possible to make accurate estimates for remote areas. Iran is no exception to this rule and faces the problem of lack of rain gauge stations in some areas. Satellite precipitation products, as a supplement to ground rainfall stations, provides useful information on rainfall properties in different parts of the world. In this study, the capability of PERSIANN-CDR precipitation data with spatial resolution of 0.25 0 0.25 ° and daily time resolution has been eva‎luated for the development of IDF curves over Iran. The satellite precipitation data was first corrected based on altitude level factor in each cell with the proposed correction model and interpolation on satellite data related to precipitation in order to convert its spatial resolution to dimensions of 0.05 × 0.05°. In this study, different downscaling methods including probability matching methods, probability matching window and Bayesian averaging were applied to extract the precipitation data across Iran. Bayesian averaging method has been introduced as a new method for downscaling and converting of area rainfall to point. eva‎luation and comparison of IDF curves obtained from satellite model data with IDF curves obtained from Bayesian averaging methods and probability matching in most of the studied stations showed acceptable results. In Bayesian averaging method, Isfahan stations with 0.12% and Ahvaz stations with 0.67% showed the lowest and highest amplitude of relative error respectively. In the probability matching method, Tabriz, Isfahan and Qazvin stations showed the lowest amplitude of relative error changes with values of 0.07, 0.06 and 0.08% respectively, while Mashhad station with 0.75% had the highest range of relative error. The results of probability and Bayesian matching methods for the return period of 2 years in Isfahan, Tabriz, Birjand and Hamedan stations showed that the estimated rainfall intensity values are within the 90% confidence interval estimated based on ground rain gauge precipitation.

عليرضا گوهري

سعيد اسلاميان

جهانگير عابدي كوپايي، شمس اله ايوبي