تحليل فضايي مخاطره‌ي توفان‌هاي تندري بهاره‌ي ايران

The Spatial Analysis of Hazard of Spring Thunderstorms in Iran

درپژوهش حاضر، داده هاي ماهانه ي فراواني وقوع توفان هاي تندري 25 ايستگاه سينوپتيك كشور در طي بازه ي زماني 51 ساله از 1960-2010 با استفاده از روش هاي تحليل خوشه اي سلسله مراتبي وارد شد و روش زمين آمار كريجينگ به منظور طبقه بندي و شناسايي مكان هاي اصلي رخداد مخاطره ي توفان هاي تندري فصل بهار در ايران تحليل شد. پس از محاسبه فراواني وقوع توفان هاي تندري فصل بهار و بررسي ويژگي هاي آماري مربوط به آن، تابع توزيع احتمالاتي مناسب با داده هاي توفان هاي تندري برازش داده شدو سپس، با استفاده از تحليل خوشه اي به تقسيم بندي مناطق گوناگون به گروه هاي همگون و با استفاده از روش كريجينگ به پهنه بندي توفان هاي تندري اقدام گرديد. پس از بررسي فراسنج هاي آمار توصيفي توفان هاي تندري بهاره، مشخص شد كه توزيع احتمال فراواني وقوع داده هاي توفان هاي تندري بهاره ي ايران، مانند بيشتر متغيرهاي تصادفي گسسته از توزيع احتمال ويبول سه پارامتري تبعيت مي كند. براساس نتايج حاصل از تحليل خوشه اي مناطق گوناگون كشور به پنج ناحيه همگن مجزا مشتمل بر نواحي شمالي، مياني، شمال شرقي؛ نواحي مركزي و شرقي؛ شمال غرب؛ غرب و نيمه ي جنوبي با روند مشابه خوشه بندي گرديد. پس از انجام پهنه بندي مشخص شد كه كانون هاي اصلي رخداد اين پديده بيشتر در نواحي شمال غرب و غرب كشور متمركز شده است.

Thunderstorm is one of the most severe atmospheric disturbances in the world and also in Iran, which is characterized by rapid upward movements, abundant moisture, and climatic instability. Since this phenomenon is usually accompanied with hail, lightning, heavy rain, flood and severe winds, it can cause irreparable damage to the environment. Investigation of spring thunderstorms has a great significance regarding the irreparable damages can cause by them and also because of the higher frequency of this phenomenon in the spring and the necessity for preparedness and disaster mitigation actions. To identify the locations of the major thunderstorm risk areas, the entire country with an area of 1648195 square kilometers, which is located between the 25deg 40deg north latitude and 44deg 63deg east longitude is considered. Spatial distribution of the occurrence of hazardous spring thunderstorms was analyzed using a series of monthly thunderstorm frequency data obtained from 25 synoptic stations over a 51yearlong period (19602010). Wards hierarchical clustering and Kriging methods were used for statistical analysis. Initially, total number of thunderstorms in April, May and June were considered as the frequency of occurrence of thunderstorm in different stations in the spring. Measure of central tendency and dispersion which consists of the sum, minimum, maximum, range and coefficient of variation, standard deviation, and skewness were used to clarify the changes of thunderstorms and to determine the spatial and temporal climatic distribution of spring thunderstorms. An appropriate probability distribution function was chosen to determine the distributions of the data. Due to the large volume of data and the uneven distribution of stations, cluster analysis and kriging methods were used to classify different regions into homogeneous groups for zoning and spatial analysis of spring thunderstorms, respectively. The statistical characteristics of spring thunderstorms were reviewed and fitted with a 3parameter Weibull distribution. Regions considered for this study were classified in four separate clusters according to the simultaneity of thunderstorms in the spring. After zoning, it was found that the highest rates of thunderstorm took place in the northwest and west of country. The northeast of Iran has the second highest number of thunderstorm occurrence. The least number of thunderstorm event had happened in the central and southern half of the country. According to the descriptive statistics parameters, maximum number of thunderstorms occurred in May.. Based on the results of the cluster analysis, there is a similar trend in the central and eastern regions, the rest of the country was clustered into five distinct homogeneous regions, including the northwestern, western, southern, northern, central northern and northeastern regions. Zoning results indicate that the highest number of the occurrence of this phenomenon in the country is concentrated in the northwestern and western regions. Higher frequency of occurrence of thunderstorms in the northwestern and western regions may be attributed to local topographic conditions like high mountains, orientation of the terrain, solar radiation on slopes and existence instability conditions, hillside convection, the presence of water resources and specific climatic conditions in these areas. In addition, as a result of a continuous surface obtained by the method of interpolation with the least amount of systematic error and also the use of correlation functions for recognizing the spatial structure of the data and estimating the model error when using the Kriging method, the weights are chosen in order to have a more optimized interpolation function. Also the cluster analysis may significantly reduce the volume of operation without affecting the results and will help in finding a real band due to more appropriate classification of different geographic areas with greater spatial homogeneity and minimal variance within the group. Based on the results of the spatial analysis, it is clear that Kriging and Ward cluster analysis methods are appropriate for thunderstorm zoning and classification of different regions according to occurrence of thunderstorm, respectively.