Modelling Topo-Climatology and Zoning Frost Statistical Indices in Kurdistan Province

Frost is one of the most important phenomena in climatology, which is caused by changes in temperature over time. The sudden occurrence of this phenomenon at the beginning and end of the cold period can be very dangerous for the agricultural sector. Therefore, the awareness of the frost time - occurrence has long been considered by researchers (Thom and Shaw, 1958; Rosenberg and Myers, 1962; Schmidlin, 1986; Watkins, 1991; Waylen, 1988). In order to manage the reduction of the effects of this destructive climate phenomenon on the agricultural sector and the exploitation of large regional environmental capabilities, it is necessary to notice seriously the detailed study of this phenomenon and its characteristics at the land level. And this will be costly and time-consuming. Therefore, with the purpose of preventing the last two factors and at the same time achieving managerial goals, it seems necessary to accurately zoning and recognizing homogeneity and non-homogeneity between different areas in a large area.  

In the first step, daily minimum temperature data were adjusted based on Julius day, and the averages of the five indicators including the day of the onset of frost, the day of the end of frost, the annual number of days of frost, the length of the frost season, and the length of the growing season were extracted. In the second step, the five indicators were modeled separately with three land-climate factors, namely altitude, longitude, and latitude of the stations, using multivariate regression models. To measure the accuracy of the obtained models, four basic assumptions were examined (). Using the regression models obtained for all parts of the province, the statistical indicators of the frosts were calculated and generalized to the points without stations. Finally, using the kiriging method, each of the five frost indicators of the province was zoned.  

The correlation coefficient of three variables, altitude, length, and latitude with different frost indices was obtained by simultaneously entering these three variables into the regression model. And four basic assumptions for measuring the accuracy of the obtained models were confirmed. The earliest occurrence of the first day of frost arises between September 21 and October 27, and in the mountains of northwestern Kurdistan, especially the Chehel Cheshmeh. The latest occurrence of the first day of frost also happens in the eastern lowlands of the province between October 17 and November 23. The earliest occurrence of the last day of frost arises between March 22 and 30 in the lowlands of southeastern and southwestern Kurdistan, and the latest happens between May 24 and June 1 in the high peaks of the west and northwest of the province, such as Chehel Cheshmeh Heights at an altitude of about 3173 meters, Ketresh Mountain with a height of 2592 meters, and Vazneh Mountain with a height of 2697 meters. The highest frequency of frost is in the mountains of the region with more than 196 days and the lowest frequency is in the eastern borders of the province with less than 72 days. The northwest mountains with 235 to 248 days and the eastern and southeastern regions of Kurdistan with 123 to 137 days, respectively, have the longest and shortest length of the frosted season. The longest growing season belongs to the eastern part of the province. The average growing season in this area is between 214 and 227 days. However, within this area, small sections that are lower in height have a longer growth period. On the other hand, the shortest growth period is in the western and northwestern mountains, averaging 116 to 129 days.  

The results show that the three factors of altitude, latitude, and longitude can determine between 72 and 95% of the changes in different frost indicators. These three factors explain the 95, 90, 88, 80, and 72 percent changes in the length of the growth period, the occurrence of the first day of frost, the length of the frosted period, the frequency of frost, and the last day of frost, respectively. The Coefficient of determination is 95% for the first day of frost and 72% for the last day of frost. It seems that other factors besides the three mentioned factors play a role in changing the date of the last day of frost. Therefore, based on the studies of Noohi et al. in 2007, Noohi et al. 2009, and Alijani et al. in 2014, it can be inferred that the end frosts of the cold period can be more than the type of the advection frost. In other words, the synoptic factors can play a more important role in the occurrence of the last days of frost and its variability. But the spatial arrangement of different frost indices in Kurdistan province indicates a western to the eastern arrangement in the values of different frost indices. This means that with more movement from west to east, the number of frost days as well as the length of the frosted period decreases, and as a result, the growing season increases. In accordance with these changes, the occurrence of the first day and the last day of frost also arose with many delays between the eastern and western parts of the province. A comparison of the maps obtained from this algorithm showed that this method can provide more accurate details of the frost indicators compared to the zoning that used only stationary data (Mianabadi et al., 2009 and Ziaee et al. 2006).