Identification of variations trend of thunder storm Occurrence frequency in Iran

This study mainly aims to investigate trends of the number of days with thunder storms throughout Iran. To accomplish so, the observed data for present weather codes of thunderstorms (codes 17, 29, and 99-91) at the main synoptic hours in 50 synoptic stations during the 35years period (1979-2013) are used. After extraction of the present weather codes of thunderstorms, especial quality control procedures are applied to data; therefore, noisy data were excluded from the time series data and the time series values related to the annual frequency of thunderstorm occurrences of each station are obtained. Then, to determine the overall condition of the thunderstorm in the studied stations, a distribution diagram is drawn to demonstrate the days with thunderstorms. T-test and Kendall test are employed to investigate the whole process and its significance. Also, the Pettitt and Mann-Kendall (MK) tests are conducted to find mutations in the frequency of thunderstorm occurrences and, later, the significance is acquired using the Mann-Whitney test (for single mutation) and the Kruskal-Wallis test (multi-mutation). The Iran’s annual thunderstorm distribution diagram are concentrated in the northern and northwestern parts of the country.The results indicate an increasing trend in the number of days with thunderstorm in 39 stations and a decrease in 11 stations accordingly. Besides, the linear least squares method is employed to calculate the trend values; then, the total amount of the annual trend is demonstrated in terms of percentage of the average annual thunderstorm.The results show that the maximum amount belongs to Kashan and Yazd’s stations, both located in the margin of Iran’s central desert. In other words, the maximum positive variability comparing to the long-term average occurred in these two cities. To monitor the first occurrence of mutations along the longitude and latitude, the first mutation occurrence is mapping using the Kriging method in areas with significant mutations. The results indicate that the variations are greater along latitude rather than longitude.