Uncertainty Analysis of Downscaled Precipitation Using LARS-WG Statistical Model in Shahrekord Station, Iran

摘要

Studies related to the effects of large scale changes in regional climate using Global Circulation Models (GCM) under the influence of Green House (GH) gas emissions are very important in the evaluation of the effects/impacts of GH gases on the natural resources of a region. Once the outputs from the GCM are downscaled (reduced) to smaller scale (i.e. regional watersheds) the simulation results allow the investigators to use it for hydrologic and ecologic studies. In this paper, the results from an application of GCM to north Karoon River sub-watershed of Behesht-abad using the Shahrekord synoptic station with 30 years of climatologic records and the ECHO-GCM using the emission scenario A-l are presented. In this work the model was downscaled and the results were used via a stochastic simulation model LARS-WG to simulate a base period of 30 years and forecast future trends for an additional 30 years (2010-390) simultaneously. As an integral part of this investigation the uncertainty of the GCM simulations using the Bootstrap method with confidence interval of 95% was also evaluated. The long term monthly and annual precipitation for the base period and the simulated results from the subsequent 30 years into the future were compared. The modeling results indicate that the accuracy of predicted precipitations is within an acceptable range and that the simulation results can be used for future water supply planning. Based on the preliminary results from the model for the month of July an increase in precipitation of about 0.3mm is expected in future years while for all other months the trends are decreasing precipitation. For example, for the month of September, the long term trends are a reduction of 0.14 mm and for December 44mm reduction, which is the maximum predicted monthly precipitation decrease. Overall, on an annual basis it is expected that precipitation will decrease by 37% on a long term basis; that is, the total precipitation for the region represented by this station will decrease from 334 mm to 208 mm. This information has great ramification for the water resources management of the region and future water supply.