پهنه‌بندي تعداد روز طوفاني در كشور ايران با استفاده از روش‌هاي زمين‌آماري و رياضي

Mapping of Dusty Stormy Days in Iran Using Geostatistical and Mathematical Methods

زﻣﯿﻨﻪ و ﻫﺪف: ﺗﻌﺪاد روز ﻃﻮﻓﺎﻧﯽ ﺑﻪ وﺳﯿﻠﻪ ﻋﻮاﻣﻞ ﮔﻮﻧﺎﮔﻮﻧﯽ ﻣﺎﻧﻨﺪ ﺳﺮﻋﺖ وزش ﺑﺎد، ﻣﯿﺰان ﺑﺎرش، رﻃﻮﺑﺖ ﺧﺎك و ﻏﯿﺮه ﻗﺮار دارد. ﺑﺮرﺳﯽ اﯾﻦ ﺷﺎﺧﺺ در ﮐﺸﻮر ﻣﯽﺗﻮاﻧﺪ در ﺑﺮﻧﺎﻣﻪرﯾﺰيﻫﺎي ﮔﻮﻧﺎﮔﻮن ﻣﻮرد ﺗﻮﺟﻪ ﻗﺮار ﮔﯿﺮد. ﻫﺪف از اﻧﺠﺎم اﯾﻦ ﺗﺤﻘﯿﻖ ﭘﻬﻨﻪﺑﻨﺪي ﺗﻌﺪاد روزﻫﺎي ﻃﻮﻓﺎﻧﯽ در ﮐﺸﻮر و اﻧﺘﺨﺎب ﺑﻬﺘﺮﯾﻦ ﻣﺪل ﺑﺮ اﺳﺎس دادهﻫﺎي 150 اﯾﺴﺘﮕﺎه ﻫﻮاﺷﻨﺎﺳﯽ در دوره آﻣﺎري 25 ﺳﺎﻟﻪ 1986-2010ﻣﯿﻼدي اﺳﺖ. روش ﺑﺮرﺳﯽ: ﺑﻌﺪ از ﺗﻌﯿﯿﻦ اﯾﺴﺘﮕﺎهﻫﺎ ﺑﺎ ﭘﺮاﮐﻨﺪﮔﯽ ﻣﻨﺎﺳﺐ، دادهﻫﺎي ﺗﻌﺪاد روز ﻃﻮﻓﺎﻧﯽ در ﮐﺸﻮر اﯾﺮان ﻣﺮﺑﻮط ﺑﻪ دورهي زﻣﺎﻧﯽ 2010-1986 در ﺳﺎل 1395 ﺟﻤﻊآوري ﺷﺪ. ﺑﻪ ﻣﻨﻈﻮر ﺑﺮرﺳﯽ ﻧﺮﻣﺎل ﺑﻮدن دادهﻫﺎ از روش ﮐﻠﻤﻮﮔﺮوف- اﺳﻤﯿﺮﻧﻒ اﺳﺘﻔﺎده ﮔﺮدﯾﺪ. ﺟﻬﺖ ﻧﺸﺎن دادن ﻫﻤﺒﺴﺘﮕﯽ ﻣﮑﺎﻧﯽ ﺑﯿﻦ دادهﻫﺎي ﺗﻌﺪاد روز ﻃﻮﻓﺎﻧﯽ از وارﯾﻮﮔﺮام اﺳﺘﻔﺎده ﺷﺪ. واﯾﺮوﮔﺮام ﮔﻮﺳﯿﻦ ﺑﺎ ﻣﯿﺰان ﻫﻤﺒﺴﺘﮕﯽ 0/96، ﺑﻬﺘﺮﯾﻦ ﻫﻤﺒﺴﺘﮕﯽ ﺑﯿﻦ دادهﻫﺎ را ﻣﺪلﺳﺎزي ﻧﻤﻮد و ﺑﺮاي درونﯾﺎﺑﯽ از روش ﮐﺮﯾﺠﯿﻨﮓ اﺳﺘﻔﺎده ﮔﺮدﯾﺪ. ﺳﭙﺲ ﺑﺎ اﺳﺘﻔﺎده از روشﻫﺎي ﻣﺨﺘﻠﻒ زﻣﯿﻦآﻣﺎري و رﯾﺎﺿﯽ، ﻧﻘﺸﻪ ﺗﻌﺪاد روزﻫﺎي ﻃﻮﻓﺎﻧﯽ در ﮐﺸﻮر ﺗﺮﺳﯿﻢ ﺷﺪ. ﺑﻪ اﯾﻦ ﻣﻨﻈﻮر از روشﻫﺎي درونﯾﺎﺑﯽ رﯾﺎﺿﯽ ﺷﺎﻣﻞ روش ﻋﮑﺲ ﻓﺎﺻﻠﻪ )IDW(، درونﯾﺎب ﭼﻨﺪ ﺟﻤﻠﻪاي ﺟﻬﺎﻧﯽ )GPI(، ﺗﺎﺑﻊ ﺷﻌﺎﻋﯽ )RBF(، درونﯾﺎب ﻣﻮﺿﻌﯽ )LPI( و روش زﻣﯿﻦآﻣﺎري ﮐﺮﯾﺠﯿﻨﮓ اﺳﺘﻔﺎده ﺷﺪ. ﺟﻬﺖ اﻧﺘﺨﺎب ﺑﻬﺘﺮﯾﻦ روش درونﯾﺎﺑﯽ از ﺷﺎﺧﺺﻫﺎي آﻣﺎري رﯾﺸﻪ ﻣﯿﺎﻧﮕﯿﻦ ﻣﺮﺑﻌﺎت )RMS( و ﻣﯿﺰان ﻫﻤﺒﺴﺘﮕﯽ دادهﻫﺎي ﻣﺸﺎﻫﺪاﺗﯽ و ﭘﯿﺶﺑﯿﻨﯽ-ﺷﺪه اﺳﺘﻔﺎده ﮔﺮدﯾﺪ. ﯾﺎﻓﺘﻪ ﻫﺎ: ﻧﺘﺎﯾﺞ ﻧﺸﺎن داد ﮐﻪ ﭘﯿﺶﺑﯿﻨﯽ روش ﮐﺮﯾﺠﯿﻨﮓ ﺷﺎﺧﺺ )Kriging Indicator( داراي ﺑﯿﺸﺘﺮﯾﻦ ﻣﯿﺰان ﻫﻤﺒﺴﺘﮕﯽ ﺑﺎ دادهﻫﺎي ﻣﺸﺎﻫﺪاﺗﯽ اﺳﺖ )R2=0.74(. ﻫﻤﭙﻨﯿﻦ ﻣﻨﺎﻃﻖ ﺟﻨﻮب ﺷﺮﻗﯽ و ﺟﻨﻮب ﻏﺮﺑﯽ ﮐﺸﻮر داراي ﺑﯿﺸﺘﺮﯾﻦ ﻣﯿﺰان ﺗﻌﺪاد روز ﻃﻮﻓﺎﻧﯽ در ﮐﺸﻮر ﻣﯽ-ﺑﺎﺷﻨﺪ. ﺑﺤﺚ و ﻧﺘﯿﺠﻪ ﮔﯿﺮي: ﺑﺎﻻ ﺑﻮدن اﯾﻦ ﺷﺎﺧﺺ در ﺟﻨﻮب ﺷﺮق ﻧﺎﺷﯽ از ﺧﺸﮏ ﺷﺪن درﯾﺎﭼﻪ ﻫﺎﻣﻮن و وﺟﻮد ﺑﺎدﻫﺎي 120 روزهي ﺳﯿﺴﺘﺎن و در ﺟﻨﻮب ﻏﺮب ﮐﺸﻮر ﻧﺎﺷﯽ از ﮔﺮد و ﻏﺒﺎر ورودي از ﮐﺸﻮرﻫﺎي ﻋﺮﺑﯽ اﺳﺖ. ﻧﻮار ﺷﻤﺎﻟﯽ ﮐﺸﻮر ﻧﯿﺰ داراي ﮐﻤﺘﺮﯾﻦ ﻣﯿﺰان ﺷﺎﺧﺺ ﺗﻌﺪاد روز ﻃﻮﻓﺎﻧﯽ در ﮐﺸﻮر اﺳﺖ.

Background and Objective: The number of stormy days is determined by various factors such as wind speed, rainfall, soil moisture and so on. The study of this index in the country can be considered in various plans. The purpose of this research is mapping of the number of dusty stormy days in Iran and selecting the best model based on the climatic data of 150 meteorological stations for the period of 25 years (1986-2010). Method: Dust stormy days’ data of the studied stations were analyzed using variogram curves to represents their spatial correlation. Gaussian variogram (R2=0.96) shows the highest correlation between the data. Then, map of the number of dust stormy days in Iran were prepared using different geostatistical and mathematical methods. For this purpose, several mathematical interpolation methods including Inverse Distance Method (IDW), Global Polynomial Interpolation (GPI), Radial Basis Function (RBF), Local Polynomial Interpolation (LPI), and geostatistical method of Kriging were used. To select the best interpolation method among several geostatistical and mathematical methods, statistical indicators of Root Mean Square (RMS) and correlation coefficient between observed and predicted data were used. Findings: Results show that the highest correlation between predicted and observed data (R2 = 0.74) was found in kriging indicator method. The southeast and southwest of the country have the highest number of dust storm days. Discussion and Conclusion: High number of dust stormy days in the southeast is resulting from drying of Hammon lakes and blowing of 120-day winds in Sistan plain, and entering of dust from Arabic countries form the direction of southwest. North part of the country has the lowest number of dust storm days.