Compare the performance of statistical model and dynamic model to simulate extreme rainfall simulation

Abstract:
Water resources has been experienced serious tensions in recent years due to climate change. Purpose of this study is investigation of dynamical and statistical downscaling in order to downscaling extreme precipitation in catchment of Gorganrood river over 13 and 14 may 1992 that Leading to record the extreme discharge in this region. Two model namely SDSM and Regcm4 for his purpose were used. The results showed SDSM has very low ability to simulate extreme precipitation so that the mean absolute error (MAE) was about 20 mm in the years 1983-2012 with very low coefficients of determination of 0.18 to 0.002 whereas Regcm4 model has recorded a very high coefficient of determination and mean absolute error of the model was about 67 mm. Although this model could not well simulate of Lazoreh station precipitation, because the model taking into account the overall dynamic patterns to simulate. Weather maps analysis indicates that Low pressure patterns dominant in the western half and parts of South and Central Iran in this two-days. Direction South and West Stream was observed from Adan sea and the Persian Gulf as well as the Black Sea and the Mediterranean Sea. Geopotential Maps show low-height patterns be in level 500 Geopotential high and n level 850 Geopotential high in this region.
Language:
Persian
Published:
Iranian Journal of Eco Hydrology, Volume:4 Issue: 2, 2017
Pages:
301 to 313
https://www.magiran.com/p1686893