Accuracy of Downscaling methods, display of minimum and maximum temperature parameters(Case study: Birjand and Rasht synoptic stations)

The consequences of climate change have led the international community to study more broadly that changes in natural resources, ecosystems, and populations will be affected by future climate change. Recent studies show that the global climate cycle will intensify. Climate change can have destructive effects on various sources such as water, forest, agriculture, etc. The first effects of climate change on atmospheric elements, especially temperature and precipitation, so it is essential to study the trend of climate change. The large scale model used in the present study is the CanESM2 model. Also, for exponential Downscaling in this research, 3 models of LARS -WG, SDSM and qmap have been used. The scenarios studied in the present study are two climatic scenarios RCP 4.5 and RCP 8.5. In this study, the performance of the introduced models as well as the temperature changes of the next period 2025-2056 in two synoptic stations of Rasht and Birjand have been investigated. The SDSM model combines linear regression and meteorological random generator, because the humidity variables and large-scale silicon pattern of the atmosphere are used linearly for local-scale meteorological generating parameters at single stations. The LARS-WG model is one of the most popular random weather data generator models and is used to generate daily series of rainfall values and minimum and maximum temperatures and sunny hours in a station under the conditions of basic and future climate conditions. This method is based on using random weather generators, which are offered based on the time series pattern and Fourier series. The dynamic bias correction method (BCSD) was first used to estimate the long-term components of hydrology and is now widely used in monthly climatic studies. By performing preprocessing operations on the minimum and maximum temperature data in the two stations analyzed, the results of the box diagram method showed that the studied data lacked outdated data. In addition, the results of data trend using the Mann-Kendall test for two parameters of minimum and maximum temperature in Birjand and Rasht stations show an increasing trend. According to these results, the SDSM model has a very high performance for predicting both stations’ minimum and maximum temperature parameters. The results of the LARS-WG model also show that the correlation between predicted data and daily observational data per month in both stations is perfect. The LARS model has a good performance in general, especially in predicting maximum temperature data, also the best performance of the LARS model According to the maximum temperature data in Birjand synoptic station. The worst performance is related to the minimum temperature data in this station. In addition, the results of the qmap model show that, in general, the best performance of qmap model is related to the simulation of the minimum temperature parameter for the next period in the Rasht station. The worst performance associated with the simulation of the minimum temperature parameter in the Birjand station. According to the results of the R2 index, this model has a good performance. Still, according to the NSE index results; this model's performance is not suitable for simulating the minimum temperature parameter for the next period. Comparison of annual maximum and minimum temperature changes in two synoptic stations of Birjand and Rasht shows that the parameters of minimum and maximum temperature in both stations will increase in the next period of 2056-2025 compared to the base period (1974-2005). Also, the temperature changes under the RCP 8.5 scenario are more than the rcp4.5 scenarios. In addition, the fluctuations of the minimum temperature parameter in the future period of Rasht station are more than Birjand station. In addition to these changes, the average maximum and minimum monthly temperatures in the period 2056-2025 compared to the base period (1974-2005) show that these changes are incremental for Birjand station and in Rasht station, except for the SDSM model in February, March, April, October, November and December and qmap in April, the changes are incremental. Changes in the average monthly minimum temperature of the next period in both stations are also an increase. However, in the SDSM model in April for Birjand station and in October, November, and December for Rasht station these changes decrease. However, according to the obtained results, in the period 2056-2025, the warming process is taking place in the study area.