عنوان مقاله [English]
One of the important branches of synoptic climatology is identifying the extreme states of environmental features, such as severe storms and especially heavy rains and floods (Alijani et al., 2008, 107). Flood is the most common environmental damage. Every year, floods kill more than 2000 people and unfortunately affect 75 million of the world's population (Mohammadi and Masoudian, 2008, 70). Flood is considered as one of the most important natural disasters in our country, and what makes this natural disaster a disaster is the lack of awareness to deal with its consequences and prevent the adverse effects of natural events on the pillars of economic and environmental well-being. Therefore, since heavy and torrential rains are dangerous and damaging environmental phenomena that occur in most places, especially in areas with little rain, and cause a lot of damage, identifying the synoptic conditions that cause these rains can help in predicting the time of occurrence and implementing the necessary preparations.
In this study, two categories of ground and upper atmosphere data are used as follows:
A) Use of daily rainfall data from October to December for 22 stations located in Kurdistan, Kermanshah, Hamedan, Lorestan and Ilam provinces, which were obtained from the National Meteorological Organization during the statistical period of 1991 to 2019.
B) Using high-level data. Includes reprocessed data for components of geopotential height, sea level pressure, zonal wind, meridional wind, specific humidity, and omega obtained from the US Environmental Prediction Database. The spatial resolution of this data is 2.5 by 2.5 arc degrees. TRMM satellite data with a spatial resolution of 1 x 1 degree was also used to analyze rainfall and runoff. To perform synoptic analysis, the environmental to circulation method has been used; In this way, first, the days of heavy rains were identified with the 98% percentile method, and then their synoptic dimensions in the space between 10 to 70 degrees north latitude and 0 to 80 degrees east longitude were investigated. By considering and applying the two conditions of rainfall threshold greater than 98 percent and covering more than 50 percent on the daily rainfall data of the autumn season (September 23 to December 21) during the period from 1991 to 2019 at selected stations in the western provinces of Iran, 20 days super heavy rain was extracted. The amount of precipitation on the mentioned days ranges from 22 to 81 mm; which shows the lower intensity of heavy rains in the autumn season compared to the winter season of the region. In this research, using the obtained data, maps of sea level pressure, combination of geopotential height and Omega, humidity advection, jet stream, precipitation rate, runoff and profile have been drawn and analyzed using Grads software. In the next step, in order to classify the patterns of sea level pressure on the days of heavy rain, the s-mode cluster analysis method was carried out using Euclidean distance and Ward technique.
Results and Discussion
Figure 2 shows the clustering tree of sea level pressure data during the days of heavy and pervasive rainfall in the western half of Iran. In this research, s-state cluster analysis method with Ward integration technique was used to classify pressure patterns during days of heavy rain. The obtained patterns were also checked by visual method and trial and error in order to select the best classes with the least intra-group difference and the most extra-group differences and changes. Finally, after the investigations, 4 pressure patterns were identified. But in this research, due to the similarity of patterns and the large number of maps, 3 final patterns were selected for analysis. The identified patterns are the first pattern of multi-core low pressures (Sudanese, Mediterranean and Saudi) with 8 days, the second pattern of Sudanese low pressure with 8 days and the third pattern of North European-Atlas, Mediterranean and Sudanese low pressures with 4 days (Table 1). In terms of time, the months of November, December and October have the highest number of rainy days. The slope of the heavy rainfall trend in all stations except Ilam shows a slight increasing trend. Only Ilam station has experienced a weak negative trend of heavy rainfall. This behavior of the trend of heavy rain actually indicates the increase of flood rains in this region of the country in the autumn season. Based on the frequency of days of heavy rain, the selected stations of Sanandaj and Hamedan are equal to each other with 26 days of heavy rain, Kermanshah and Khorramabad stations with 24 days and Ilam with 15 days. It can be seen that there is a slight difference between the stations in terms of the occurrence of this extreme and risky phenomenon, which can be an indication of the effects of circulation patterns on a synoptic scale that have created these conditions.
The statistical analysis of the days of heavy rain in the western part of Iran in the autumn season showed that the months of November, December and October have the most frequent days of heavy rain. Most of the stations in the region experience a weak positive trend of heavy rainfall on an annual scale, which indicates an increase in the occurrence of such extreme events in the autumn season of the last few years. Based on the 98% percentile threshold, the frequency of rainy days in the selected stations ranges from 15 to 26 days, and there is a slight difference
between the stations in this regard. For the synoptic section, after classifying the sea level pressure patterns by cluster analysis method, a visual inspection of the pressure maps from a few days before the peak of rainfall was done to ensure the type of pressure systems and their main origin. Finally, after the investigations, 3 final patterns of super heavy rainfall in the west of the country during the autumn season were identified. The results of the analysis of sea level pressure patterns showed that in the first pattern, multi-core low pressure centers are located over the Middle East. In the second pattern, we see extensive changes in the pressure values in the Northern Hemisphere, in such a way that the Siberian high pressure with an east-west movement covers all of Russia, East and Central Europe, and North-East Africa, It has not allowed the penetration of low pressure systems from the Mediterranean and Europe, and in this case, only the Sudanese low pressure entered the studied area after passing through the Red Sea and Arabia from the south-west and south of Iran. But in the third pattern, the pressure changes in the Northern Hemisphere are the opposite of the other patterns, so that since the days before the peak of rainfall, the closed low pressure center from the north of the Atlantic Ocean has started to advance eastward and with the extension of its tongues on the southern latitudes, created the secondary Mediterranean low pressure in the west of this sea, and on the day of peak rainfall, it is located in the east of the sea, or it merges with the low pressure tongue of the North Atlantic, and it covers the western region of the country in a unified manner with the closed low pressure of Sudan and Arabia.