Abstract:
Climate change is found to be the most important global issue in the 21st century, so to monitor its trend is of great importance. Atmospheric General Circulation Models because of their large scale computational grid are not able to predict climatic parameters on a point scale, so small scale methods should be adapted. Among downscaling methods, statistical methods are used as they are easy to run. Two famous models, ClimGen and SDSM, were studied for daily total precipitation and temperature data in Qazvin station. For this purpose, three steps of models calibration, verification and simulation, in Qazvin station were performed and model performances in terms of similarities in produced data with those using parameters such as root mean square error (RMSE), coefficient of determination (R2) and Nash coefficient (NSE) were assessed. The results in climatic range showed that Climgen outperform in rainfall data generation while SDSM outperforms in simulating average temperatures. However, both models have high potential to simulate temperature and precipitation.
Machine summary:
"The results in climatic range showed that Climgen outperform in rainfall data generation while SDSM outperforms in simulating average temperatures.
Bazrafshan et al (2009) studied two models including ClimGen and LARS-WG for total precipitation, minimum and maximum temperatures of air and solar radiation in fifteen climatic regions of Iran.
(2014) evaluated performance of three models CLIMGEN, LARS-WG and WeatherMan in fine-scale prediction and in weather stations scale for climate variables including maximum temperature, minimum temperature, precipitation and solar radiation for the years 2000-2009 in three areas Gorgan, Mashhad and Gonbad.
This study was conducted to evaluate potentials of these two methods for fine-scale forecasting and in weather stations for climatic variables and models performances were evaluated in simulation of data in a given statistical period.
In the case of the SDSM, after the preparation and quality control of observational data, statistical downscaling was performed using NCEP, model calibration and validation was performed (respectively, 1961-1988 and 1989- 2001) and eventually climate scenarios were simulated using observed predictors.
The results for the observed and modeled data by two models of SDSM and Climgen in the future period show a decrease in the average annual precipitation and an increase in average annual temperature relative to the base period.
Comparison between the average observed and synthesized monthly temperature data by two models of Climgen and SDSM in the calibration phase (1961-1988) Fig. 5.
Future changes of annual rainfall and average temperature according to baseline period (1961-2001) based on two models of CLIMGEN and SDSM (View the image of this page) 4."