عنوان مقاله [English]
Drought is one of the most severe weather phenomena that has the greatest impact on the world's population and can affect millions of people around the world every year. Various indicators have been defined to calculate the drought index. Among these, the SPI index is widely used due to the simplicity of calculations and the use of available data and the ability to calculate at different time scales. Also, the output of general circulation models (GCM) can be suitable tools for predicting future drought events in each region if the optimal models are selected and a suitable and valid method is used in downscaling. On the other hand, with the studies conducted in Mazandaran province regarding the increasing trend of temperature and decreasing rainfall over the past 30 years, its effect on increasing drought is undeniable. Also, in Mazandaran province, rainfall has an important role in maintaining the livelihood of people, ecosystems and agriculture, forest communities, local hydrological regime, which needs to be studied possible the effects of climate change on meteorological components in the future. Therefore, in this research, initially the precipitation and temperature values at Qarakhil station in Siahrood watershed in Mazandaran province were simulated using three models: CanESM2, CNRM-CM5, CESM1-WACCM under two scenarios: RCP 8.5 and RCP 2.6. Also, the statistical method based on linear regression and SDSM software was used. Finally, with the results obtained from the simulation and using SPI index, monitoring and analysis of drought characteristics were considered for the basic decades (1985-2005) and future (2020-2040) in time scales 6, 12 and 24 months.
Statistics of maximum temperature, minimum temperature and precipitation of 3 models CanESM2, CNRM-CM5, CESM1-WACCM were extracted from AOGCM models in the period 2020-2040 as the future period and in the period 1985-2005 as the base period. In this study, from the three mentioned models, two models, CanESM2 and CNRM-CM5 for temperature and CNRM-CM5 model for suitable precipitation were identified. The problem with the rejected models was the large changes in the monthly temperature of the future period compared to the base period, as well as the very large changes in the base period and the observation period. One of the major problems in using the output of AOGCM models is the large scale of their computational cell relative to the study area. In this study, the statistical method is based on linear regression and SDSM software to scale the rainfall data of selected models. The model establishes statistical relationships between large-scale and local behaviors based on multiple linear regression methods. These connections are made using station observational data and the outputs of general circulation models in the same observation period, and it is assumed that these relationships will be true in the future. In other words, the basic premise in statistical downscaling is that time is independent of these relationships. Then, changes in temperature and precipitation in the period 2020-2040 compared to the period 1985 to 2005 and were examined. Finally, by introducing the rainfall time series produced in the future period and the rainfall time series in the observation period with the standard rainfall drought index (SPI), the drought status of the study area in time scales 6, 12 and 24 Months were examined. The calculation steps and computational formulas of this index are fully presented in the research of Loukas et al. 2004.
Future period data of AOGCM models (CanESM2 and CNRM-CM5) under two scenarios of RCP2.6 and RCP8.5, were examined with baseline data. The results showed that the minimum and maximum annual temperature changes of the models under both scenarios will be in the direction of temperature increase. The highest annual minimum temperature increases of AOGCM models under RCP2.6 scenario was predicted by 2.6 ° C in February and under RCP8.5 scenario by 2.5 ° C in January. On the other hand, the maximum maximum temperature increases under scenario 8 will be 4.7 ° C in January and under scenario 2 will be 3.6 ° C in January and April. It should be noted that the temperature pattern changes under the CanESM2 model will be more intense than the CNRM-CM5 model, although it is not significant. The results of seasonal precipitation of CNRM-CM5RM models under both scenarios show a decrease in precipitation in spring and in autumn and winter we will not have significant changes in precipitation. The CNRM-CM5 model also predicted a 44% decrease in rainfall under the RCP2.6 scenario and an 18% increase in rainfall under the RCP8.5 scenario in the summer. The average annual rainfall of both models and under both scenarios showed a decrease. In the period 1985-2005, it had severe drought on a 6-month scale in 1991, 1985, 1995, on a 12-month scale in 1986, and on a 24-month scale in 1987. In all three-time scales of 6, 12 and 24 months, the drought index in the future period showed a lot of fluctuations compared to the base period under both diffusion scenarios. It is also important to predict the probability of the occurrence of drought limit conditions in the coming period. The results of seasonal precipitation of CNRM-CM5RM models under both scenarios show a decrease in precipitation in spring and in autumn and winter we will not have significant changes in precipitation. The CNRM-CM5 model also predicted a 44% decrease in rainfall under the RCP2.6 scenario and an 18% increase in rainfall under the RCP8.5 scenario in the summer. The average annual rainfall of both models and under both scenarios showed a decrease. As the results showed, the study area will experience climate change and drought in the coming period, which will undoubtedly have a significant impact on the ecosystem of agricultural systems and water resources (quantity and quality of water) in these areas. Therefore, facing these conditions requires extensive studies and planning in the field of the impact of future climate change on the country's natural resources, and the results of climate models of this study can deal with this phenomenon, especially in agriculture, water resources and The environment and adopting appropriate adaptation and mitigation measures to be used.