quantile regression

Global warming cause changes in temperature variables and consequently increase in the occurrence of extreme events like as floods and droughts, result from extreme weather in most cases; Therefore, it is important to study and investigate the future changes of extreme values of meteorological and hydrological parameters, including temperature, more than the average or median value. Therefore, the purpose of this research is to investigate the effects of climate change on the trend of seasonal changes in extreme temperatures (very high and very low values) in coastal stations of Mazandaran province based on CMIP6 models and quantile regression method.

In this research, in order to investigate the trend of extreme values of temperature in Mazandaran province for the future periods, the output of the sixth generation of climate change models (Coupled Model Intercomparison Project phase 6, CMIP6) was used. For this purpose, minimum and maximum temperature of 4 main synoptic stations of the province, including Babolsar, Qarakhil, Ramsar and Nowshahr, and also the output of the NorESM2-MM climate model from the CMIP6 in near future (2026-2055) and far future (2071-2100) at three optimistic (SSP126), moderate (SSP245) and pessimistic (SSP585) scenarios was used. In order to downscaling climate model data at the studied meteorological stations, various methods in the CMhyd downscaling software were used, and the Temperature data output for the method that had higher accuracy, was selected to analysis of trend. In the next step, the quantile regression method was used to investigate the seasonal trend of temperature extreme values and the results were analyzed.

The results of downscaling using different methods have shown that the Variance Scaling method has the best performance among the available methods in CMhyd. In general, the results show a positive temperature anomaly (annual temperature increase compared to the base period in SSP126=1.3, SSP245=2.56, SSP585=3.2 oC) in Mazandaran province in all months of the year until the end of the 21st century. The intensity of the anomalies is higher in the warm months than in the cold ones. Under the optimistic scenario, the extreme values of the temperature variables in the spring and autumn seasons will decrease by a maximum of 1 degree per decade in the distant future. However, the average minimum temperature is increasing in all scenarios (including the optimistic scenario) and the decrease is only predicted for extreme values. But under the average and pessimistic scenarios, there will be an increase in the temperature extreme values in all seasons, so that its intensity will be greater for the pessimistic scenario and in the spring season (0.15 oC per year).

According to the results, it concluded that the significant increase in extreme temperature during the day and night, especially in the warm months of the year, will increase evapotranspiration, and along with the decrease in rainfall in the hot months, it will cause a decrease in water resources in different parts of Mazandaran province and pressure on the groundwater. Therefore, it is very important to formulate and implement appropriate management programs according to the needs of each region, in order to adapt to extreme temperatures and their adverse consequences.

Climate changes caused by the progress and industrialization of human societies have caused changes in the intensity and frequency of heavy precipitation and floods, which have caused irreparable damages. In order to reduce these damages, it is necessary to identify the changes in the threshold values ​​of precipitation and factors affecting it each region. Quantile regression methods are able to examine the trends not only in the median, but also in different ranges of the data series. Therefore, the purpose of this research is to investigate the seasonal trend of different amounts of precipitation and also to investigate the relationship between the climatic parameters of minimum temperature, maximum temperature, minimum relative humidity, maximum relative humidity and wind speed on different amounts of precipitation in different climates of Iran.

In the first step, the daily time series of climate data including precipitation, minimum and maximum temperature, minimum and maximum relative humidity and wind speed for a period of 45 years in different seasons for 5 synoptic stations of Babolsar, Shiraz, Bandar Abbas, Khorram Abad and Torbat Heydarieh were formed. In the selection of study stations, we tried to select stations with different climates and with appropriate statistical period. Then, to investigate the trend of seasonal changes of different amounts of precipitation in different quantiles (quantiles 0.01, 0.05, 0.1, 0.25, 0.5, 0.75, 0.9, 0.95 and 0.99) was analyzed using the quantile regression method in all study stations. In the next step, the relationship between the climatic parameters on different amounts of precipitation (low to very high amounts of precipitation) in different seasons was investigated for each of the stations using the quantile regression method. Then the results were analyzed.

The results of examining the changes trend of daily precipitation are showed that the high amounts of precipitation in the spring season in Bandar Abbas, Shiraz, and Torbat Heydarieh stations were reduced significantly, but very high amounts of precipitation (0.95 and 0.99 troughs) in the station Babolsar and Khorramabad have increased. Also, very high daily precipitation amounts in summer have decreased in Bandar Abbas station but increased in Torbat Heydarieh and Khorram Abad stations, significantly. While in the winter season, different amounts of precipitation in all seasons have a decreasing trend and there was only a significant positive slope in very high amounts of precipitation (slope of 0.99) in Babolsar station. In the investigation of the parameters affecting the extreme precipitation, the results showed that the amount of impact on the occurrence of heavy rainfall was relatively higher than low to median rainfall. The parameters of minimum temperature, minimum humidity, maximum humidity and wind speed have a positive effect and the maximum temperature parameter has a negative effect on heavy rainfall in different seasons and stations.The highest positive effect coefficients were in spring for wind speed in Babolsar (1.8), in summer for wind speed in Babolsar (3.8), minimum and maximum temperature in Bandar Abbas (-4.03 and 1.53), in Autumn season for maximum humidity and wind speed in Babolsar (2.57 and 2.99) and wind speed in Khorram Abad (1.54) and in winter, for wind speed in Babolsar and Bandar Abbas (1.94 and 6. 2), and minimum temperature in Torbat Heydarieh (0.96). Also, the highest negative effect coefficients of maximum temperature were in autumn and winter seasons (-0.88 and -0.72) in Babolsar and autumn season (-0.63) in Shiraz.

The significant changes in increasing and decreasing precipitation are mostly related to the amounts of heavy precipitation, which are different in different seasons and climates. Also, the precipitation of the stations near the north and south coasts have been influenced by climatic parameters to a greater extent. In general, it can be said that flood precipitations are influenced by climatic parameters such as wind speed, humidity and temperature in order of importance and this effect is different depending on the location and time and the influence of different factors. Therefore, it is necessary to apply accurate planning for the correct use of received rainfall and optimal management in the target area using the results of such studies.

Climatic events such as floods, storms and droughts are often caused by extreme weather. Therefore, it is important to study the trend of different ranges of climatic data rather than just average. Most trend detection studies are based on the analysis of changes in mean data and do not provide information on how changes occur in different ranges of the used data range. Therefore, to investigate the trend of changes in different ranges of the time series of climatic data, quantile regression method was proposed. The quantile regression has the ability to examine the changes trend in different quantiles of the data series. Therefore, in this study, we was analyzed the changes trend in the different quantiles of temperature minimum and maximum, precipitation and maximum daily wind speed data time series in Babolsar synoptic station for a period of 62 years (1959- 2020) seasonally and annually. The results showed that the minimum and maximum temperature and wind speed increased significantly in all seasons. The intensity of these increasing trend was higher in the extreme lower quantiles of the daily minimum temperature and extreme upper quantiles of daily maximum temperature in winter. But the wind speed data, in the upper quantiles especially extreme upper quantiles have increased to a much greater intensity than the lower quantiles, especially in autumn. The precipitation did not change significantly in the 62-year period. On an annual scale, the maximum increase for the minimum and maximum temperatures was in the extreme lower quantiles 4.28 and 2.85 degrees centigrade, respectively, and for the wind speed, it was in the extreme upper quantiles 8.24 meters per second.