پهنه‌بندي تأثير عوامل مديريتي و بيوفيزيكي بر وقوع آتش‌سوزي در جنگل‌ها و مراتع استان چهارمحال‌و‌بختياري با رگرسيون وزن‌دار جغرافيايي

Modelling and mapping effects of managerial and biophysical factors on occurrence of forest and rangelands fire in Chahrmahal-va-Bakhtiari province using geographically weighted regression

ﯾﮑﯽ از ﻣﻬﻢﺗﺮﯾﻦ ﻋﻮاﻣﻞ ﺗﻬﺪﯾﺪ ﮐﻨﻨﺪة ﺟﻨﮕﻞﻫﺎ و ﻣﺮاﺗﻊ، آﺗﺶﺳـــﻮزيﻫﺎي ﻋﻤﺪي و ﻏﯿﺮﻋﻤﺪي اﺳـــﺖ ﮐﻪ در ﭼﻨﺪ ﺳـــﺎل اﺧﯿﺮ ﻣﻨﺠﺮ ﺑﻪ ﺗﺨﺮﯾﺐ ﺑﺨﺶ وﺳــﯿﻌﯽ از ﻣﻨﺎﺑﻊ ﻃﺒﯿﻌﯽ ﺷــﺪه اﺳــﺖ. ﻣﻄﺎﻟﻌﻪ و اراﺋﮥ ﺗﺪاﺑﯿﺮ ﻣﺪﯾﺮﯾﺘﯽ ﻣﯽﺗﻮاﻧﺪ ﻧﻘﺶ ﮐﻨﺘﺮل ﮐﻨﻨﺪهاي ﺑﺮاي ﻣﻘﺎﺑﻠﻪ ﺑﺎ اﯾﻦ ﺑﺤﺮان اﯾﻔﺎ ﮐﻨﺪ. ﻣﻄﺎﻟﻌﮥ ﺣﺎﺿﺮ، ﺟﻬﺖ ﺷﻨﺎﺳﺎﯾﯽ و ارزﯾﺎﺑﯽ ﻣﻬﻢﺗﺮﯾﻦ ﻓﺎﮐﺘﻮرﻫﺎي ﺗﺎًﺛﯿﺮﮔﺬار ﺑﺮ ﻣﻬﺎر آﺗﺶﺳﻮزي و ﺷﻨﺎﺳﺎﯾﯽ ﻣﻨﺎﻃﻖ ﺑﺎ ﺿﺮﯾﺐ ﺧﻄﺮ آﺗﺶﺳﻮزي ﺑﺎﻻ ﻃﺮاﺣﯽ ﮔﺮدﯾﺪ. ﺑﺮاي اﯾﻦ ﻣﻨﻈﻮر، راﺑﻄﮥ ﺑﯿﻦ3 ﮔﺮوه ﻋﻮاﻣﻞ اﻧﺴﺎﻧﯽ )اﻧﮕﯿﺰﺷﯽ(، ﻋﻮاﻣﻞ ﺑﯿﻮﻓﯿﺰﯾﮑﯽ، ﻋﻮاﻣﻞ ﺗﺠﻬﯿﺰات و اﻣﮑﺎﻧﺎت ﻣﺸــﺘﻤﻞ ﺑﺮ 26 ﻣﺘﻐﯿﺮ )ﻣﺘﻐﯿﺮﻫﺎي ﻣﺴــﺘﻘﻞ( ﺑﺮ ﺗﻌﺪاد وﻗﻮع آﺗﺶﺳــﻮزيﻫﺎي رخ داده )ﻣﺘﻐﯿﺮ واﺑﺴــﺘﻪ( در دﻫﺴــﺘﺎنﻫﺎي اﺳــﺘﺎن ﭼﻬﺎرﻣﺤﺎل و ﺑﺨﺘﯿﺎري ﻣﻮرد ﺑﺮرﺳـــﯽ ﻗﺮار ﮔﺮﻓﺖ. از روش رﮔﺮﺳـــﯿﻮن وزندار ﻓﻀـــﺎﯾﯽ (GW) ﺑﺮاي ﺗﻬﯿﮥ ﻧﻘﺸـــﮥ ﭘﻬﻨﻪﺑﻨﺪي ﺗﻌﺪاد آﺗﺶﺳﻮزيﻫﺎي رخ داده، ﻃﯽ ﺳﺎلﻫﺎي 1386 ﺗﺎ 1392و ﺗﻌﯿﯿﻦ ارﺗﺒﺎط آن ﺑﺎ ﻣﺘﻐﯿﺮﻫﺎي ﻣﺴﺘﻘﻞ اﺳﺘﻔﺎده ﮔﺮدﯾﺪ. ﻧﺘﺎﯾﺞ ﻧﺸﺎن داد از ﺑﯿﻦ 26 ﻣﺘﻐﯿﺮ ﻣﻮرد ﺑﺮرﺳـــﯽ، ﺗ ﻌﺪاد 6 ﻣﺘﻐﯿﺮ ﺷـــﺎﻣﻞ ﻣﯿﺰان ﺣﻘﻮق درﯾﺎﻓﺘﯽ ﻣﺎﻫ ﯿﺎﻧﻪ (0/61-=R و 8/08=VIF)، ﺗ ﻌﺪاد ﻧﯿﺮوي ﻣ ﺤﺎﻓﻆ (0/56-=R و 10/81=VIF)، ﺗﻌﺪاد ﭘﺎ ﺳﮕﺎه ﺣﻔﺎﻇﺘﯽ (0/54-=R و 2/2=VIF)، ﻣﺘﻮ ﺳﻂ ﺳﻦ اﻓﺮاد ﻣﺤﺎﻓﻆ (0/53=R و 9/71 =VIF)، ﻣﺘﻮﺳــﻂ ﺷــﯿﺐ ﻋﺮﺻــﻪﻫﺎي ﻃﺒﯿﻌﯽ (0/50=R و 8/99 =VIF)، ﺗﻌﺪاد ﭘﺮﺳــﻨﻞ اﻓﺘﺨﺎري (0/42-=R و 15/11 =VIF) ﺑﻪ ﺗﺮﺗﯿﺐ داراي ﺑﯿﺸﺘﺮﯾﻦ ﺗﺄﺛﯿﺮﮔﺬاري ﺑﺮ روي ﺗﻌﺪاد آﺗﺶﺳﻮزيﻫﺎي رخ داده ﺑﻮدﻧﺪ. در ﻧﻬﺎﯾﺖ ﻃﺒﻖ ﻧﻘﺸﮥ ﭘﻬﻨﻪﺑﻨﺪي ﺗﺮﺳﯿﻢ ﺷﺪة دﻫﺴﺘﺎنﻫﺎي وردﻧﺠﺎن، ﻣﯿﺰدج ﻋﻠﯿﺎ، ﭘﺸــﺖﮐﻮه اردل و ارﻣﻨﺪ داراي ﺑﯿﺸــﺘﺮﯾﻦ ﺧﻄﺮﭘﺬﯾﺮي و دﻫﺴــﺘﺎن ﻣﯿﺎﻧﮑﻮه ﻣﻮﮔﻮﯾﯽ و ﻣﻨﺞ داراي ﮐﻤﺘﺮﯾﻦ ﺧﻄﺮﭘﺬﯾﺮي از ﻧﻈﺮ ﺗﻌﺪاد ﺣﺮﯾﻖﻫﺎي رخ داده ﺑﻮدﻧﺪ.

One of the important factors threatening forests and rangeland is fires that leads to destruction of a large part of forests and rangelands. Study of this phenomenon and providing management strategies plays a major role to deal with and control such crisis. This study aimed to identify the affecting factors of fire occurrence. The identification of major criteria conducted using a questionnaire as well as gathering information from Natural Resources and Watershed Management Organization in addition to calculation some others. To do so, 3 category of variables including 1- Human factors, 2- Biophysical factors and 3- Instrumental and logistic factors that contained altogether 26 variables are studied. In this research, Geographically Weighted Regression (GWR) method was used for mapping and zonation of the burned areas in the province that were occurred from 2007 to 2013. Results showed that amongst the 26 studied variables, monthly income (R=-0.61 and VIF=8.08) and number of rangeland and forest guardian members (R=-.56 and VIF=10.81), number of guard stations (R=-0.54 and VIF=2.2), guardsmen’s’ average age (R=0.53 and VIF=9.71 ), average of slope (R=0.5 and VIF=8.99) and number of voluntary rangeland and forest guards (R=-0.42 and VIF=15.11) are respectively the most affective variables on the occurred fires in range and forestlands. Finally, based on extracted predicted map Vardenjan, Mizdej-olia, Poshtkouh Ardal, and Aarmand are the most vulnerable counties for fire incident. Whilst, Monj and Mougooei encountered the least vulnerability of number of fire occurrence that is significantly in line with occurred fires.