Document not found! Please try again

Forecasting ambient air pollutants by time series models in ... - TUMS

2 downloads 0 Views 289KB Size Report
for the ambient air quality data of Delhi City. Environ Monit Assess, 157, pp. 105–112. Wagemakers, E. and Farrell, S., 2004. AIC model selection using Akaike ...
‫ﻣﺠﻠﻪ داﻧﺸﻜﺪه ﺑﻬﺪاﺷﺖ و اﻧﺴﺘﻴﺘﻮ ﺗﺤﻘﻴﻘﺎت ﺑﻬﺪاﺷﺘﻲ‬ ‫دوره ‪ ،11‬ﺷﻤﺎره ‪ ، 2‬ﺑﻬﺎر ‪ ،1392‬ﺻﻔﺤﻪ‪75-86‬‬

‫ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري‪ :‬داﻧﺸﺠﻮي دوره ﻛﺎرﺷﻨﺎﺳﻲ ارﺷﺪ‪ ،‬ﮔﺮوه ﻣﻬﻨﺪﺳﻲ ﺑﻬﺪاﺷﺖ ﻣﺤﻴﻂ‪ ،‬داﻧﺸﻜﺪه ﺑﻬﺪاﺷﺖ‪ ،‬داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﻜﻲ ﻛﺮﻣﺎن‪ ،‬ﻛﺮﻣﺎن‪ ،‬اﻳﺮان‬ ‫ﻧﺮﮔﺲ ﺧﺎﻧﺠﺎﻧﻲ‪ :‬اﺳﺘﺎدﻳﺎر‪ ،‬ﮔﺮوه اﭘﻴﺪﻣﻴﻮﻟﻮژي و ﻋﻀﻮ ﭘﻴﻮﺳﺘﻪ ﮔﺮوه ﻣﻬﻨﺪﺳﻲ ﺑﻬﺪاﺷﺖ ﻣﺤﻴﻂ‪ ،‬داﻧﺸﻜﺪه ﺑﻬﺪاﺷﺖ‪ ،‬داﻧﺸﮕﺎه ﻋﻠـﻮم ﭘﺰﺷـﻜﻲ ﻛﺮﻣـﺎن‪ ،‬ﻛﺮﻣـﺎن‪ ،‬اﻳـﺮان‪-‬‬ ‫ﻧﻮﻳﺴﻨﺪة راﺑﻂ‪[email protected]:‬‬ ‫ﻻﻟﻪ راﻧﻨﺪه ﻛﻼﻧﻜﺶ‪ :‬داﻧﺸﺠﻮي دوره ﻛﺎرﺷﻨﺎﺳﻲ ارﺷﺪ‪ ،‬ﮔﺮوه ﻣﻬﻨﺪﺳﻲ ﺑﻬﺪاﺷﺖ ﻣﺤﻴﻂ‪ ،‬داﻧﺸﻜﺪه ﺑﻬﺪاﺷﺖ‪ ،‬داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﻜﻲ ﻛﺮﻣﺎن‪ ،‬ﻛﺮﻣﺎن‪ ،‬اﻳﺮان‬ ‫رﺿﺎ ﭘﻮرﻣﻮﺳﻲ‪ :‬ﻣﺮﺑﻲ‪ ،‬ﺑﺨﺶ آﻣﺎر‪ ،‬داﻧﺸﻜﺪه رﻳﺎﺿﻲ و ﻛﺎﻣﭙﻴﻮﺗﺮ‪ ،‬داﻧﺸﮕﺎه ﺷﻬﻴﺪ ﺑﺎ ﻫﻨﺮ‪ ،‬ﻛﺮﻣﺎن‪ ،‬اﻳﺮان‬ ‫ﺗﺎرﻳﺦ درﻳﺎﻓﺖ‪1391/1/14 :‬‬

‫ﺗﺎرﻳﺦ ﭘﺬﻳﺮش‪1391/11/25 :‬‬

‫ﭼﻜﻴﺪه‬ ‫زﻣﻴﻨﻪ و ﻫﺪف ‪ :‬آﻟﻮدﮔﻲ ﻫﻮا ﻳﻜﻲ از ﻋﻤﺪه ﺗﺮﻳﻦ ﻣﺸﻜﻼت ﺷﻬﺮ ﻫﺎي ﺑﺰرگ در ﻛﺸﻮر ﻫﺎي در ﺣﺎل ﺗﻮﺳﻌﻪ اﺳﺖ و ﻣﻲ ﺗﻮاﻧﺪ ﻋﻮارض ﻣﻨﻔﻲ‬ ‫ﺑﺴﻴﺎري ﺑﺮ ﺳﻼﻣﺖ اﻧﺴﺎﻧﻬﺎ داﺷﺘﻪ ﺑﺎﺷﺪ‪ .‬ﻟﺬا ﻣﻄﺎﻟﻌﻪ ﺗﻐﻴﻴﺮات اﻳﻦ آﻻﻳﻨﺪه ﻫﺎ ﻣﻲ ﺗﻮاﻧﺪ ﻛﻤﻚ ﻣﻬﻤﻲ ﺑﺮاي ﺑﺮﻧﺎﻣﻪ رﻳﺰي و ﻣﻘﺎﺑﻠﻪ ﺑﺎ آﻧﻬﺎ ﺑﺎﺷﺪ‪ .‬ﻫﺪف از‬ ‫اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ ﺗﻐﻴﻴﺮات آﻻﻳﻨﺪه ﻫﺎ در ﻫﻮاي ﺷﻬﺮ ﻛﺮﻣﺎن ﺑﻮد‪.‬‬ ‫روش ﻛﺎر‪ :‬در اﻳﻦ ﻣﻄﺎﻟﻌﻪ اﻛﻮﻟﻮژﻳﻚ اﻃﻼﻋﺎت ﻫﻔﺖ آﻻﻳﻨﺪه ي ﻣﻬﻢ ﺷﻬﺮ ﻛﺮﻣﺎن ﺷﺎﻣﻞ ‪ NO, CO, NO2, NOx, PM10, SO2, O3‬از‬ ‫اﺑﺘﺪاي ﺳﺎل ‪ 1385‬ﺗﺎ آﺧﺮ ﺷﻬﺮﻳﻮر ‪ 1389‬از ﺳﺎزﻣﺎن ﺣﻔﺎﻇﺖ ﻣﺤﻴﻂ زﻳﺴﺖ ﻛﺮﻣﺎن اﺳﺘﻌﻼم ﺷﺪ‪ .‬ﺳﭙﺲ اﻃﻼﻋﺎت ﺑﺼﻮرت ﻣﺘﻮﺳﻂ در ﻣﺎه ﻣﺤﺎﺳﺒﻪ‬ ‫و ﺑﻪ ﻛﻤﻚ روﺷﻬﺎي آﻣﺎري‪ ،‬اﻟﮕﻮﻫﺎي ﺳﺮي زﻣﺎﻧﻲ ﺗﻚ ﻣﺘﻐﻴﺮه ﺑﺮاي ﻫﺮ آﻻﻳﻨﺪه ﺑﺮازش و ﻣﻘﺎدﻳﺮ آن ﭘﻴﺶ ﺑﻴﻨﻲ ﺷﺪ‪.‬‬ ‫ﻧﺘﺎﻳﺞ‪ :‬آﻻﻳﻨﺪه ﻫﺎ در ﻫﻮاي ﻛﺮﻣﺎن روﻧﺪ ﺗﻘﺮﻳﺒﺎً ﺛﺎﺑﺘﻲ داﺷﺘﻨﺪ‪ ،‬ﺑﻪ ﺟﺰ ﻣﻮﻧﻮاﻛﺴﻴﺪ ﻛﺮﺑﻦ ﻛﻪ ﺑﻪ ﻃﻮر ﻣﻌﻨﻲ داري در ﺣﺎل ﻛﺎﻫﺶ و ﮔﺮد و ﻏﺒﺎر ﻛﻪ روﻧﺪ‬ ‫اﻓﺰاﻳﺸﻲ داﺷﺖ‪ .‬ﻫﻤﻪ آﻻﻳﻨﺪه ﻫﺎ اﻟﮕﻮي ﻓﺼﻠﻲ داﺷﺘﻨﺪ‪ .‬اﻟﮕﻮ ﻫﺎي ﺳﺮي زﻣﺎﻧﻲ ﺑﺎ روﻧﺪ ﻓﺼﻠﻲ ‪ 6 ،12 ،12 ،12 ،8 ،3 ،12‬ﻣﺎه ﺑﻪ ﺗﺮﺗﻴﺐ ﺑﺮاي‬ ‫آﻻﻳﻨﺪه ﻫﺎي ‪ NO, CO, NO2, NOx, PM10, SO2, O3‬ﺑﺮازش ﺷﺪ‪.‬‬ ‫ﻧﺘﻴﺠﻪ ﮔﻴﺮي‪ :‬ﻣﻴﺰان ﺗﻮﻟﻴﺪ ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ در ﻛﺮﻣﺎن رو ﺑﻪ ﻛﺎﻫﺶ اﺳﺖ و ﻳﻚ ﻋﻠﺖ اﺣﺘﻤﺎﻟﻲ آن ﻃﺮح ﺟﻤﻊ آوري ﺧﻮد رو ﻫﺎي ﻓﺮﺳﻮده ﻣﻲ‬ ‫ﺗﻮاﻧﺪ ﺑﺎﺷﺪ‪ .‬اﻣﺎ ﻣﻴﺰان ﮔﺮد و ﻏﺒﺎر رو ﺑﻪ اﻓﺰاﻳﺶ و در ﺑﻌﻀﻲ ﻓﺼﻮل ﺳﺎل در ﺣﺪ ﻏﻴﺮ ﺑﻬﺪاﺷﺘﻲ اﺳﺖ وﻟﺬا ﺑﺎﻳﺪ ﺗﺪاﺑﻴﺮ ﻻزم ﺑﺮاي ﻣﻘﺎﺑﻠﻪ ﺑﺎ آن ﺑﻜﺎر‬ ‫ﮔﺮﻓﺘﻪ ﺷﻮد‪.‬‬ ‫واژﮔﺎن ﻛﻠﻴﺪي‪ :‬آﻟﻮدﮔﻲ ﻫﻮا‪ ،‬ذرات ﻣﻌﻠﻖ‪ ،‬ﻛﺮﻣﺎن‪ ،‬ﺳﺮي زﻣﺎﻧﻲ‪ ،‬دي اﻛﺴﻴﺪ ﮔﻮﮔﺮد‪ ،‬ازن‪ ،‬اﻛﺴﻴﺪﻫﺎي ﻧﻴﺘﺮوژن‪ ،‬ﻣﻮﻧﻮ اﻛﺴﻴﺪ ﻛﺮﺑﻦ‪.‬‬

‫ﻣﻘﺪﻣﻪ‬ ‫آﻟﻮدﮔﻲ ﻫﻮا ﻳﻜﻲ از ﻣﺸﻜﻼت ﺷﻬﺮﻫﺎي ﺑﺰرگ و‬

‫ﻧﻴﺘﺮوژن و ‪ PM10‬و ﻣﺮگ و ﻣﻴﺮ ﺗﻨﻔﺴﻲ و ﻗﻠﺒﻲ‪ ،‬ﻋﻔﻮﻧﺖ ﻫﺎي‬

‫ﺻﻨﻌﺘﻲ ﺟﻬﺎن اﺳﺖ و ﺑﻪ ﻃﻮر ﺟﺪي ﺳﻼﻣﺖ ﺳﺎﻛﻨﺎن اﻳﻦ‬

‫رﻳﻮي و ﺑﻴﻤﺎريﻫﺎي ﺣﺎد ﻧﻴﺎزﻣﻨﺪ ﺑﺴﺘﺮي ﺑﻴﻤﺎرﺳﺘﺎﻧﻲ ‪،‬ارﺗﺒﺎط‬

‫ﺷﻬﺮﻫﺎ را ﺗﻬﺪﻳﺪ ﻣﻴﻜﻨﺪ )‪ .(Masjedi et al. 2001‬ﺑﻬﺘﺮﻳﻦ‬

‫وﺟﻮد دارد ) ‪Liang et al. 2009; Hosseinpour et al.‬‬

‫ﻣﺜﺎل آن ﻣﻪ ﻟﻨﺪن در ﺳﺎل ‪ 1952‬ﺑﻮد ﻛﻪ ﺟﺎن ﺗﻌﺪاد زﻳﺎدي از‬

‫‪ .(2005; Ghorbani and Younesian 2010‬ﺑﻪ ﻋﻨﻮان‬

‫اﻓﺮاد را ﮔﺮﻓﺖ )‪ .(Anderson 2009‬ﻣﻄﺎﻟﻌﺎت ﺑﺴﻴﺎري ﻧﺸﺎن‬

‫ﻣﺜﺎل ﻣﻄﺎﻟﻌﻪ اي در ﻫﻨﺪوﺳﺘﺎن ﻧﺸﺎن داد ﻛﻪ ﺑﻪ ازا اﻓﺰاﻳﺶ ﻫﺮ‬

‫داده ﻛﻪ ﺑﻴﻦ دي ﺳﻮﻟﻔﻴﺪ ﻛﺮﺑﻦ‪ ،‬ﻣﻮﻧﻮ اﻛﺴﻴﺪ ﻛﺮﺑﻦ‪ ،‬دي اﻛﺴﻴﺪ‬

‫‪3‬‬

‫‪ 10µg/m‬ذرات ﻣﻌﻠﻖ ‪ PM10‬ﻣﺮگ و ﻣﻴﺮ ﻛﻠﻲ ‪ 0/15‬درﺻﺪ‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ ﻛﺮﻣﺎن ﺑﺎ ﻣﺪل ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ‬

‫‪ / 76‬ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري و ﻫﻤﻜﺎران‬

‫اﻓﺰاﻳﺶ ‪ 0/84‬در ﺻﺪ در ﻛﻞ ﻣﺮگ و ﻣﻴﺮ اﺗﻔﺎق اﻓﺘﺎده اﺳﺖ‪.‬‬

‫ﻛﺮد ﻛﻪ رﻫﻨﻤﻮد ﺳﺎﻟﻴﺎﻧﻪ و ‪ 24‬ﺳﺎﻋﺘﻪ را ﺑﺮاي ‪ PM10‬ﺑﻪ ﺗﺮﺗﻴﺐ‬

‫)‪ (Rajarathnam et al. 2011‬ﻣﻄﺎﻟﻌﺎﺗﻲ از آﻣﺮﻳﻜﺎي‬

‫ﻛﻤﺘﺮ از ‪ 20 µg /m3‬و ‪ 50 µg /m3‬ﭘﻴﺸﻨﻬﺎد داد و ﻫﻤﻴﻦ‬

‫ﻫﺮ‬

‫ﻃﻮر رﻫﻨﻤﻮد ‪ 24‬ﺳﺎﻋﺘﻪ دي اﻛﺴﻴﺪ ﺳﻮﻟﻔﻮر ﺑﻪ ‪20 µg /m3‬‬

‫ﻫﻢ‬

‫ﺷﻤﺎﻟﻲ‬

‫ﻧﺸﺎن‬

‫داده‬

‫اﺳﺖ‬

‫ﻛﻪ‬

‫اﻓﺰاﻳﺶ‬

‫‪ 10 µg/m2‬از ذرات ﻣﻌﻠﻖ ‪ PM10‬ﻫﻤﺮاه ﺑﺎ اﻓﺰاﻳﺶ ﻣﺮگ و‬ ‫ﻣﻴﺮ ﺣﺪود ‪ 0/3‬ﺗﺎ ‪ 0/5‬در ﺻﺪ ﺑﻮده اﺳﺖ ) ‪Samet et al.‬‬ ‫‪.(2000‬‬

‫ﻛﺎﻫﺶ ﻳﺎﻓﺘﻪ ﺑﻮد )‪.(Anderson 2009‬‬ ‫ﻣﻄﺎﻟﻌﺎت ﺳﺮي زﻣﺎﻧﻲ ﺑﺮاي ﺑﺮرﺳﻲ آﻟﻮدﮔﻲ ﻫﻮاي‬ ‫ﺑﺴﻴﺎري از ﺷﻬﺮ ﻫﺎ ﺑﻜﺎر رﻓﺘﻪ اﺳﺖ ) ;‪Zhang et al. 2011‬‬

‫ﻫﻮا ﻧﻴﺰ ﻣﺎﻧﻨﺪ ﺳﺎﻳﺮ ﻣﻨﺎﺑﻊ ﻣﺤﻴﻂ زﻳﺴﺖ داراي ﻇﺮﻓﻴﺖ‬

‫‪ .(López-Villarrubia et al. 2010‬ﻣﺤﻘﻘﺎن از ﺳﺮي‬

‫ﻣﺤﺪود اﺳﺖ و ﺗﺤﻤﻞ ﺗﺨﻠﻴﻪ ﻣﻮاد زاﻳﺪ و ﺳﻤﻲ ﻣﺨﺘﻠﻒ را در‬

‫ﻫﺎي زﻣﺎﻧﻲ ﺗﻚ ﻣﺘﻐﻴﺮه ﺑﺮاي ﭘﻴﺶ ﺑﻴﻨﻲ ﺑﺮ اﺳﺎس آﻣﺎر ﻣﻮﺟﻮد‬

‫ﺣﺪي ﻛﻪ اﻣﺮوزه ﺑﺸﺮ ﺑﻪ آن ﺗﺤﻤﻴﻞ ﻛﺮده اﺳﺖ ﻧﺪارد‪ .‬ﺗﻐﻴﻴﺮ در‬

‫از ﻣﻘﺎدﻳﺮ آﻻﻳﻨﺪه ﻫﺎ در ﮔﺬﺷﺘﻪ اﺳﺘﻔﺎده ﻛﺮده اﻧﺪ و ﻃﺒﻖ اﻳﻦ‬

‫وﻳﮋﮔﻲ ﻫﺎي ﻓﻴﺰﻳﻜﻲ و ﺷﻴﻤﻴﺎﻳﻲ ﻋﻨﺎﺻﺮ ﺗﺸﻜﻴﻞ دﻫﻨﺪه ﻫﻮا‪،‬‬

‫ﻣﺪل ﻫﺎ ﺗﻮاﻧﺴﺘﻪ اﻧﺪ ﭘﻴﺶ ﺑﻴﻨﻲ ﻫﺎﻳﻲ از ﻣﻘﺎدﻳﺮ آﻻﻳﻨﺪه ﻫﺎ اﻧﺠﺎم‬

‫آﻟﻮدﮔﻲ ﻫﻮا اﻃﻼق ﻣﻲ ﺷﻮد ﻛﻪ ﻣﻲ ﺗﻮاﻧﺪ ﻣﻨﺒﻊ ﻃﺒﻴﻌﻲ و ﻳﺎ‬

‫دﻫﻨﺪ )‪.(Sharma et al. 2009‬‬

‫ﻣﺼﻨﻮﻋﻲ داﺷﺘﻪ ﺑﺎﺷﺪ ) ‪Nasrollhi and Ghaffari‬‬ ‫‪.(Goolak 2010‬‬

‫اﻣﺎ ﻣﻄﺎﻟﻌﺎت ﻣﻌﺪودي از آﺳﻴﺎ در اﻳﻦ ﻣﻮرد وﺟﻮد دارد‬ ‫و ﺑﻪ ﺧﺎﻃﺮ ﻛﻤﺒﻮد ﻣﻄﺎﻟﻌﺎت از آﺳﻴﺎ ﺑﺴﻴﺎري از ﻛﺸﻮر ﻫﺎ ﺻﺮﻓﺎ‬

‫ﺳﺎزﻣﺎن ﺑﻬﺪاﺷﺖ ﺟﻬﺎﻧﻲ ﺑﻪ ﺻﻮرت دوره اي ﺷﻮاﻫﺪي‬

‫از دﺳﺘﻮر اﻟﻌﻤﻞ ﺳﺎزﻣﺎن ﺟﻬﺎﻧﻲ ﺑﻬﺪاﺷﺖ ﺗﺒﻌﻴﺖ ﻣﻲ ﻛﻨﻨﺪ‪ .‬در‬

‫از آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا و ﺳﻼﻣﺖ را ﻣﺮور ﻣﻴﻜﻨﺪ و رﻫﻨﻤﻮد ﻫﺎﻳﻲ‬

‫ﺻﻮرﺗﻲ ﻛﻪ ﻛﺸﻮر ﻫﺎي آﺳﻴﺎﻳﻲ از ﻧﻈﺮ ﻧﻮع آﻟﻮدﮔﻲ و‬

‫را ﭘﻴﺸﻨﻬﺎد ﻣﻴﺪﻫﺪ ﻛﻪ ﺟﻬﺖ ﺣﻔﺎﻇﺖ ﺳﻼﻣﺖ ﻋﻤﻮﻣﻲ ﻣﻮرد‬

‫ﺧﺼﻮﺻﻴﺎت ﺟﻤﻌﻴﺘﻲ ﻣﺘﻔﺎوت ﻫﺴﺘﻨﺪ و ﻣﻘﺎدﻳﺮ ﺑﺎﻻﻳﻲ از‬

‫ﻣﻼﺣﻈﻪ ﻗﺮار ﮔﻴﺮد‪ .‬در ﺳﺎل ‪ 1970‬ﺳﺎزﻣﺎن ﺑﻬﺪاﺷﺖ ﺟﻬﺎﻧﻲ‬

‫آﻻﻳﻨﺪه ﻫﺎ در ﻛﺸﻮر ﻫﺎي آﺳﻴﺎﻳﻲ ﮔﺰارش ﺷﺪه اﺳﺖ‬

‫ﻧﺘﻴﺠﻪ ﮔﺮﻓﺘﻪ ﺑﻮد ﻛﻪ ﺳﻄﺢ ﻣﻮاﺟﻬﻪ ﺑﻠﻨﺪ ﻣﺪت ﺑﺎ ذرات ﻣﻌﻠﻖ ﻛﻪ‬

‫)‪.(Rajarathnam et al. 2011‬‬

‫ﻓﺮاﺗﺮ از آن ﺑﺮاي ﺳﻼﻣﺖ اﻧﺴﺎن ﻣﻀﺮ ﻫﺴﺘﻨﺪ در ﺣﺪود‬

‫ﻣﻄﺎﻟﻌﺎت آﻟﻮدﮔﻲ ﻫﻮاي اﻧﺠﺎم ﺷﺪه در ﻛﺸﻮر ﻣﺎ ﻧﻴﺰ‬

‫‪ 150 µg/m3‬اﺳﺖ و ﺑﻌﺪ از اﺿﺎﻓﻪ ﻛﺮدن ﻳﻚ ﻓﺎﻛﺘﻮر‬

‫ﻣﺤﺪود ﺑﻮده ) ‪Khosravi Dehkordi and Modarres‬‬

‫ﺳﻼﻣﺖ ﻣﻌﺎدل ‪ 2‬ﭘﻴﺸﻨﻬﺎد ﻛﺮد ﻛﻪ اﻳﻦ آﺳﺘﺎﻧﻪ ﺑﺎﻳﺪ ﺣﺪود‬

‫‪(2007; Nasrollahi and Ghaffari Goolak 2010‬‬

‫‪ 60µg/m3‬ﺗﺎ‪ 90 µg/m3‬ﺑﺎﺷﺪ )‪.(Anderson 2009‬‬

‫و اﻛﺜﺮا ﺑﻪ ﻣﻄﺎﻟﻌﻪ اﺛﺮ آﻟﻮدﮔﻲ ﻫﻮا ﺑﺮ ﺳﻼﻣﺖ اﻧﺴﺎن ﭘﺮداﺧﺘﻪ‬

‫در ﺳﺎل ‪ 1987‬ﺑﺎ اﺳﺘﻔﺎده از اﻳﻦ اﺻﻮل ﺳﺎزﻣﺎن‬

‫اﻧﺪ‪ .‬در اﻳﻦ ﻣﻄﺎﻟﻌﻪ آﻟﻮدﮔﻲ ﻫﻮاي ﺷﻬﺮ ﻛﺮﻣﺎن ﺑﺮاي اوﻟﻴﻦ ﺑﺎر‬

‫ﺑﻬﺪاﺷﺖ ﺟﻬﺎﻧﻲ اوﻟﻴﻦ رﻫﻨﻤﻮدش را ﺑﺮاي ذرات ﻫﻮاي آزاد‪،‬‬

‫ﻣﺪل ﺳﺎزي و ﻣﻮرد ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ ﻗﺮار ﻣﻲ ﮔﻴﺮد‪ .‬ﻛﺮﻣﺎن‬

‫دي اﻛﺴﻴﺪ ﺳﻮﻟﻔﻮر‪ ،‬ﻧﻴﺘﺮوژن دي اﻛﺴﺎﻳﺪ و ازن ﻣﻨﺘﺸﺮ ﻛﺮد‪.‬‬

‫از ﺷﻬﺮﻫﺎي ﻛﻮﻳﺮي ﺟﻨﻮب ﺷﺮق اﻳﺮان ﺑﻮده‪ ،‬ﺟﻤﻌﻴﺖ آن در‬

‫رﻫﻨﻤﻮد ﻫﺎ ﺑﺮاي ذرات ﻣﻌﻠﻖ و دي اﻛﺴﻴﺪ ﺳﻮﻟﻔﻮر ﻣﺸﺎﺑﻪ ﻫﻢ‬

‫ﺳﺮﺷﻤﺎري ‪ 677650 ،1385‬ﻧﻔﺮ ﺑﻮده اﺳﺖ و ﻋﻼوه ﺑﺮ آﻟﻮدﮔﻲ‬

‫ﺑﻮد ﻛﻪ ﺑﺮاي ﺗﻤﺎس ﻫﺎي ﻃﻮﻻﻧﻲ ﻣﺪت )ﺳﺎﻟﻴﺎﻧﻪ( ﺗﻘﺮﻳﺒﺎ‬

‫ﻫﻮا در ﻓﺼﻮﻟﻲ ﺑﺎ ﻃﻮﻓﺎﻧﻬﺎي ﺷﻦ و اﻓﺰاﻳﺶ ﮔﺮد و ﻏﺒﺎر ﻫﻮا‬

‫‪ 50 µg /m3‬ﭘﻴﺸﻨﻬﺎد ﺷﺪه ﺑﻮد‪ ،‬در ﺣﺎﻟﻲ ﻛﻪ رﻫﻨﻤﻮد ‪ 24‬ﺳﺎﻋﺘﻪ‬

‫ﻧﻴﺰ روﺑﺮو ﻣﻲ ﺷﻮد‪ .‬آب و ﻫﻮاي آن ﮔﺮم و ﺧﺸﻚ اﺳﺖ‪.‬‬

‫ﺑﺮاي ﻫﺮ دو ﺗﻘﺮﻳﺒﺎ ‪ 125 µg /m3‬اﺳﺖ‪ .‬رﻫﻨﻤﻮد ﻫﺎي ‪24‬‬

‫ﺑﻪ ﺧﺎﻃﺮ اﻫﻤﻴﺖ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا ﭘﻴﺶ ﺑﻴﻨﻲ و ﻣﻄﺎﻟﻌﻪ‬

‫ﺳﺎﻋﺘﻪ ﺑﺮاي ذرات ﺗﻨﻔﺴﻲ ‪ 70 µg /m3 PM10‬ﺑﻮد‪ .‬در ﻃﻲ‬

‫ﺗﻐﻴﻴﺮات آﻧﻬﺎ از اﻫﻤﻴﺖ ﺑﺴﻴﺎري ﺑﺮﺧﻮردار اﺳﺖ‪ .‬ﻣﺪل ﺳﺎزي‬

‫ﻛﻤﺘﺮ از ‪ 20‬ﺳﺎل اﺧﻴﺮ ﺳﺎزﻣﺎن ﺑﻬﺪاﺷﺖ ﺟﻬﺎﻧﻲ ﻳﻚ رﻫﻨﻤﻮد‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫اﻓﺰاﻳﺶ داﺷﺘﻪ و ﺑﻪ ازاء اﻓﺰاﻳﺶ ﻫﺮ ‪10 µg/m3‬در ‪NO2‬‬

‫ﺟﺪاﮔﺎﻧﻪ اي را ﺑﺮاي ذرات ﻣﻌﻠﻖ و دي اﻛﺴﻴﺪ ﺳﻮﻟﻔﻮر ﻣﻨﺘﺸﺮ‬

‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ ‪77 /....‬‬

‫آﻟﻮدﮔﻲ ﻫﻮا ﺑﻪ ﺷﻨﺎﺳﺎﻳﻲ ﺗﻐﻴﻴﺮات زﻣﺎﻧﻲ اﻳﻦ آﻻﻳﻨﺪه ﻫﺎ ﺑﺮاي‬

‫ﻧﻤﻮدار ﻫﺎي )‪Autocorrelation Function (ACF‬‬

‫ﺑﺮﻧﺎﻣﻪ رﻳﺰي ﺑﻬﺘﺮ و ﻛﻨﺘﺮل آﻻﻳﻨﺪه ﻫﺎ ﻛﻤﻚ ﺧﻮاﻫﺪ ﻛﺮد‪.‬‬

‫و )‪Partial Autocorrelation Function (PACF‬‬

‫روش ﻛﺎر‬

‫‪.(1976‬‬

‫در اﻳﻦ ﭘﮋوﻫﺶ داده ﻫﺎي آﻟﻮدﮔﻲ ﻫﻮا از اﺑﺘﺪاي ﺳﺎل‬

‫ﺑﺮ اﺳﺎس ‪ ACF‬و ‪ PACF‬ﻣﻘﺎدﻳﺮ ‪ p‬و ‪ q‬ﺣﺪس زده‬

‫‪ 1385‬ﺗﺎ آﺧﺮ ﺷﻬﺮﻳﻮر ‪ 1389‬از ﺳﺎزﻣﺎن ﺣﻔﺎﻇﺖ ﻣﺤﻴﻂ زﻳﺴﺖ‬

‫ﻣﻲ ﺷﻮد و اﻟﮕﻮي ﻓﺼﻠﻲ را ﻫﻢ ﺑﺎﻳﺪ ﻣﺸﺨﺺ ﻛﺮد‪ .‬ﺑﺎﻗﻴﻤﺎﻧﺪه‬

‫ﻛﺮﻣﺎن درﻳﺎﻓﺖ ﮔﺮدﻳﺪ‪ .‬ﺳﺎزﻣﺎن ﺣﻔﺎﻇﺖ ﻣﺤﻴﻂ زﻳﺴﺖ ﻛﺮﻣﺎن‬

‫ﻫﺎي ﻣﺪل و داده ﻫﺎي واﻗﻌﻲ ﺑﺎﻳﺪ اﻳﻦ ﺧﺼﻮﺻﻴﺎت را داﺷﺘﻪ‬

‫از اﺑﺘﺪاي ‪ 1385‬ﺷﺮوع ﺑﻪ اﻧﺪازه ﮔﻴﺮي آﻻﻳﻨﺪه ﻫﺎ ﻧﻤﻮده اﺳﺖ‪.‬‬

‫ﺑﺎﺷﻨﺪ ﻛﻪ ﻣﻴﺎﻧﮕﻴﻦ آﻧﻬﺎ ﺻﻔﺮ ﺑﺎﺷﺪ‪ ،‬وارﻳﺎﻧﺲ ﺛﺎﺑﺖ ﺑﺎﺷﺪ‪ ،‬ﺑﺮاي‬

‫اﻳﻦ ﺳﺎزﻣﺎن ‪ 7‬آﻻﻳﻨﺪه ‪NO2, CO, NO2, NOx, PM10,‬‬

‫ﻫﻤﻪ ﺗﺎﺧﻴﺮ ﻫﺎ ﻫﻤﺒﺴﺘﮕﻲ وﺟﻮد ﻧﺪاﺷﺘﻪ ﺑﺎﺷﺪ و ﺗﻮزﻳﻌﺸﺎن‬

‫‪ SO2, O3‬را ﺑﻪ ﻃﻮر روزاﻧﻪ ﺗﻮﺳﻂ ﺗﻨﻬﺎ اﻳﺴﺘﮕﺎه ﺛﺎﺑﺖ داﺧﻞ‬

‫ﻧﺮﻣﺎل ﺑﺎﺷﺪ‪ .‬ﻟﺬا ﻣﺪل ﺑﺎﻳﺪ آﻧﻘﺪر ﺗﻐﻴﻴﺮ ﻛﻨﺪ ﺗﺎ ﺑﻪ ﻣﺪل ﻣﻄﻠﻮب‬

‫ﺷﻬﺮي ﺧﻮد واﻗﻊ در ﻣﻴﺪان ﺷﻬﺪا اﻧﺪازه ﮔﻴﺮي ﻣﻴﻨﻤﺎﻳﺪ‪.‬‬

‫ﺑﺮﺳﻴﻢ )‪ .(Lumbreras et al. 2009‬در ﺻﻮرﺗﻲ ﻛﻪ اﻟﮕﻮي‬

‫داده ﻫﺎي آﻟﻮدﮔﻲ ﻫﻮا اﺑﺘﺪا ﺑﺼﻮرت ﻣﺘﻮﺳﻂ ﻣﺎﻫﺎﻧﻪ در‬

‫ﻓﺼﻠﻲ‬

‫ﻧ ﻴﺰ‬

‫ﻣﻮﺟﻮد‬

‫ﺑﺎﺷﺪ‪،‬‬

‫ﻣﺪل‬

‫ﺑﺼﻮرت‬

‫آﻣﺪﻧﺪ‪ .‬ﺳﭙﺲ ﺑﺎ ﻧﺮم اﻓﺰار ‪ MiniTab‬و ﺑﺎ اﺳﺘﻔﺎده از ﻣﺪﻟﻬﺎي‬

‫‪ SARIMA(p,d,q)(P,D,Q)T‬در ﻣﻲ آﻳﺪ ﻛﻪ در آن ‪T‬‬

‫ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ ﺗﻚ ﻣﺘﻐﻴﺮه ﺳﻌﻲ ﺷﺪ ﻛﻪ ﺑﺮاي ﻫﺮ آﻻﻳﻨﺪه در‬

‫ﻣﻌﺮف اﻟﮕﻮي ﻓﺼﻠﻲ اﺳﺖ‪ .‬ﺑﺮاي ﺳﺎﺧﺘﻦ ﻣﺪل ﻫﺎي ‪Box-‬‬

‫ﻫﻮاي ﻛﺮﻣﺎن ﺑﻬﺘﺮﻳﻦ ﻣﺪل ﭘﻴﺸﻨﻬﺎد ﺷﻮد ﺗﺎ از روي آن ﺑﺘﻮان‬

‫‪ Jenkins‬ﻣﺮاﺣﻠﻲ ﺑﺎﻳﺪ ﻃﻲ ﺷﻮد ﻛﻪ ﺷﺎﻣﻞ‪ :‬ﺷﻨﺎﺳﺎﻳﻲ‪ ،‬ﺗﺨﻤﻴﻦ‪،‬‬

‫ﻋﻼوه ﺑﺮ ﺗﺤﻠﻴﻞ داده ﻫﺎ‪ ،‬ﻣﻘﺎدﻳﺮ آﺗﻲ را ﻧﻴﺰ ﭘﻴﺶ ﺑﻴﻨﻲ ﻧﻤﻮد‪.‬‬

‫ﭼﻚ ﻛﺮدن و در ﻧﻬﺎﻳﺖ ﭘﻴﺶ ﺑﻴﻨﻲ اﺳﺖ‪ .‬اﻳﻦ ﻣﺮاﺣﻞ آﻧﻘﺪر‬

‫ﺳﺮي زﻣﺎﻧﻲ ﻣﺠﻤﻮﻋﻪ ﻣﺸﺎﻫﺪاﺗﻲ اﺳﺖ ﻛﻪ ﺑﺮ ﺣﺴﺐ زﻣﺎن ﻳﺎ‬

‫ﺗﻜﺮار ﻣﻲ ﺷﻮد ﺗﺎ ﻧﻬﺎﻳﺘﺎ ﺑﺘﻮان ﺑﻪ ﺑﻬﺘﺮﻳﻦ ﻣﺪل رﺳﻴﺪ ) ‪Duenas‬‬

‫ﻣﺮﺗﺒﻪ ﺗﻜﺮار ﻣﺮﺗﺐ ﺷﺪه اﻧﺪ‪ .‬ﺧﺎﻧﻮاده اﺻﻠﻲ اﻟﮕﻮﻫﺎي ﺳﺮي‬

‫‪ .(et al. 2005‬ﻣﺪﻟﻬﺎي آرﻳﻤﺎ ﻧﻪ ﺗﻨﻬﺎ ﭘﻴﺶ ﺑﻴﻨﻲ ﻣﻲ ﻛﻨﻨﺪ ﺑﻠﻜﻪ‬

‫زﻣﺎﻧﻲ ﻳﻌﻨﻲ اﻟﮕﻮي ‪Auto Regressive Integrated‬‬

‫ﺣﺘﻲ داﻣﻨﻪ اﻃﻤﻴﻨﺎن ﻫﻢ ﺑﺮاي ﭘﻴﺶ ﺑﻴﻨﻲ اراﺋﻪ ﻣﻲ دﻫﻨﺪ‪.‬‬

‫)‪ ،Moving Average (ARIMA‬ﻛﻪ ﺑﺎ ﻣﺮﺗﺒﻪ ﻫﺎي ‪p‬‬

‫)‪.(Lumbreras et al. 2009‬‬

‫)اﺗﻮرﮔﺮﺳﻴﻮن(‪) d ،‬در ﺟﻪ ﺗﻤﺎﻳﺰ( و ‪) q‬ﺗﺮﺗﻴﺐ ﻣﻴﺎﻧﮕﻴﻨﻬﺎي‬

‫ﺑﺮاي ﺗﻌﻴﻴﻦ ﻣﻨﺎﺳﺐ ﺑﻮدن )‪ ،( Goodness of fit‬ﻣﺪل‬

‫ﻣﺘﺤﺮك( ﺗﻌﻴﻴﻦ ﻣﻲ ﺷﻮد‪ ،‬در اﻟﮕﻮ ﺳﺎزي از وﺿﻌﻴﺖ ﻫﺎي‬

‫ﺳﺮي زﻣﺎﻧﻲ از ‪(AIC) Akaike Information‬‬

‫دﻧﻴﺎي واﻗﻌﻲ از اﻫﻤﻴﺖ زﻳﺎدي ﺑﺮﺧﻮردار اﺳﺖ‪ .‬اﻳﻦ ﺧﺎﻧﻮاده‬

‫‪ Criterion‬ﻳﺎ ﺿﺮﻳﺐ آﻛﺎﻳﻲ اﺳﺘﻔﺎده ﻣﻲ ﺷﻮد‪ .‬ﺿﺮﻳﺐ آﻛﺎﻳﻲ‬

‫ﻛﻪ ﺑﻪ اﻟﮕﻮي ﺑﺎﻛﺲ‪-‬ﺟﻨﻜﻴﻨﺲ ﻧﻴﺰ ﻣﻌﺮوف اﺳﺖ‪ ،‬ﺗﻮﺳﻂ‬

‫ﻳﻚ ﻣﻌﻴﺎر ﻣﺴﺘﻘﻞ و ﻋﻴﻨﻲ ﺑﺮاي اﻧﺘﺨﺎب ﻣﻨﺎﺳﺐ ﺗﺮﻳﻦ ﻣﺪل‬

‫ﺳﻴﺴﺘﻢ ﻫﺎي آﻣﺎري ﻧﻈﻴﺮ ‪ MiniTab‬ﺑﻪ ﺳﻬﻮﻟﺖ ﻗﺎﺑﻞ ﺑﺮرﺳﻲ‬

‫اﺳﺖ‪ .‬در اﻳﻦ روش ﻣﻨﺎﺳﺐ ﺑﻮدن ﻣﺪل ﺑﺮ اﺳﺎس ﻣﺠﻤﻮع‬

‫اﺳﺖ‪ .‬در ﺑﺴﻴﺎري از زﻣﻴﻨﻪ ﻫﺎﻳﻲ ﻛﻪ ﺳﺮي زﻣﺎﻧﻲ ﻣﻮرد اﺳﺘﻔﺎده‬

‫ﻣﺮﺑﻌﺎت ﺑﺎﻗﻴﻤﺎﻧﺪه اﺳﺖ؛ و ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﺗﻌﺪاد ﭘﺎراﻣﺘﺮ ﻫﺎي‬

‫ﻗﺮار ﻣﻲ ﮔﻴﺮد‪ ،‬ﮔﺮاﻳﺸﻬﺎي دوره اي ﻣﻌﻤﻮل اﺳﺖ‪ .‬در ﭼﻨﻴﻦ‬

‫ﻣﺴﺘﻘﻞ ﻛﺎﻫﺶ ﻣﻲ ﻳﺎﺑﺪ‪ (Ingrisch et al. 2009) .‬اﮔﺮ ﭼﻪ‬

‫ﻣﻮاردي ﺟﻬﺖ ﺗﺤﻠﻴﻞ داده ﻫﺎ از ﻣﺪل ﻫﺎي ﻓﺼﻠﻲ ﺿﺮﺑﻲ‬

‫وﻗﺘﻲ ﻛﻪ ﺗﻌﺪاد ﭘﺎراﻣﺘﺮ ﻫﺎي ﻣﺴﺘﻘﻞ اﻓﺰاﻳﺶ ﻣﻲ ﻳﺎﺑﺪ ﻣﺪل‬

‫)‪ ARIMA(p,d,q)(P,D,Q‬اﺳﺘﻔﺎده ﻣﻲ ﺷﻮد ﻛﻪ در آن‬

‫اﻧﻌﻄﺎف ﭘﺬﻳﺮ ﺗﺮ ﻣﻲ ﺷﻮد‪ ،‬اﻣﺎ از آﻧﺠﺎ ﻛﻪ ﻫﻤﻮاره ﻣﻨﻄﻘﻲ ﺗﺮ‬

‫‪ p,d,q‬ﻣﺮﺗﺒﻪ ﻫﺎي ﻏﻴﺮ ﻓﺼﻠﻲ و ‪ P,D,Q‬ﻣﺮﺗﺒﻪ ﻫﺎي ﻓﺼﻠﻲ‬

‫اﺳﺖ ﻛﻪ ﻣﺪل ﺳﺎده ﺗﺮ و ﻧﻪ ﭘﻴﭽﻴﺪه ﺗﺮ اﻧﺘﺨﺎب ﺷﻮد‪ ،‬اﻳﻦ‬

‫اﺳﺖ‪ .‬در ﺗﺤﻠﻴﻞ ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ ﺷﻨﺎﺳﺎﻳﻲ اﻟﮕﻮ ﻳﻌﻨﻲ ﺗﻌﻴﻴﻦ‬

‫ﻣﻮﺿﻮع ﻋﻤﻮﻣﺎ ﻣﻮرد ﻗﺒﻮل اﺳﺖ ﻛﻪ ﺑﻬﺘﺮﻳﻦ ﻣﺪل ﻣﺪﻟﻲ اﺳﺖ‬

‫ﻣﻘﺎدﻳﺮ ﻣﺮﺗﺒﻪ ﻫﺎي ﻓﺼﻠﻲ و ﻏﻴﺮ ﻓﺼﻠﻲ ﺑﺮ اﺳﺎس رﻓﺘﺎر‬

‫ﻛﻪ ﻫﻢ ﺑﻪ ﻣﻴﺰان ﻛﺎﻓﻲ ﭘﻴﺶ ﺑﻴﻨﻲ ﻛﻨﺪ و ﻫﻢ از ﺗﻌﺪاد ﻛﻤﺘﺮ‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫ﻣﻲ ﺑﺎﺷﺪ ) ‪Hamilton 1994; Box and Jenkins‬‬

‫‪ / 78‬ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري و ﻫﻤﻜﺎران‬

‫ﭘﺎراﻣﺘﺮ اﺳﺘﻔﺎده ﻛﺮده ﺑﺎﺷﺪ‪ .‬ﻳﻜﻲ از راﻫﻬﺎي رﺳﻴﺪن ﺑﻪ ﭼﻨﻴﻦ‬ ‫ﻣﺪﻟﻲ اﺳﺘﻔﺎده از ‪ AIC‬اﺳﺖ‪Wagemakers and ) .‬‬

‫‪ ARIMA‬ﭘﻴﺸﻨﻬﺎد ﺷﺪ ﻛﻪ ﺗﻤﺎم ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ‬ ‫‪ 0/05=α‬ﻣﻌﻨﻲ دار ﺑﻮدﻧﺪ‪) .‬ﺷﻜﻞ ‪.(1 b‬‬

‫از ﻣﻘﺎدﻳﺮ ‪ ACF, PACF‬در ﺳﻄﺢ ﻣﻌﻨﻲ داري ‪ 0/05‬ﻣﺪل‬

‫و ﺷﻴﺐ ﺧﻂ ﻧﺰدﻳﻚ ﺑﻪ ﺻﻔﺮ و ﻏﻴﺮ ﻣﻌﻨﻲ دار داﺷﺖ‪ .‬ﺗﺠﺰﻳﻪ‬

‫ﻫﺎي ﻣﻨﺎﺳﺐ اوﻟﻴﻪ ﺷﻨﺎﺳﺎﻳﻲ ﺷﺪ و ﭘﺲ از ﺳﺎﺧﺘﻦ ﻣﺪل‪ ،‬ﺑﺎ‬

‫داده ﻫﺎ ﻣﺪل ‪ ARIMA (0،1،1) (1،1،1) 6‬را ﭘﻴﺸﻨﻬﺎد ﻛﺮد‬

‫ﺿﺮﻳﺐ آﻛﺎﻳﻲ ﺑﻬﺘﺮﻳﻦ ﻣﺪل اﻧﺘﺨﺎب ﺷﺪ‪.‬‬

‫ﻛﻪ ﺗﻤﺎم ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ دار ﺑﻮدﻧﺪ‪.‬‬ ‫)ﺷﻜﻞ ‪.(1 c‬‬ ‫روﻧﺪ ‪ NO2‬ﻫﻢ در ﻃﻮل ﻣﺪت ﺗﻘﺮﻳﺒﺎ ﺛﺎﺑﺖ ﺑﻮد و ﺷﻴﺐ‬

‫ﻧﺘﺎﻳﺞ‬ ‫ﻣﻴﺎﻧﮕﻴﻦ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﻛﺮﻣﺎن در دوره ﻣﻮرد‬

‫ﺧﻂ ﻧﺰدﻳﻚ ﺑﻪ ﺻﻔﺮ و ﻏﻴﺮ ﻣﻌﻨﻲ دار داﺷﺖ‪ .‬ﺑﺎ ﺗﺠﺰﻳﻪ ﻣﺪل‬

‫ﺑﺮرﺳﻲ در ﺟﺪول ‪ 1‬آﻣﺪه اﺳﺖ‪ .‬در اﻳﻦ ﻣﻄﺎﻟﻌﻪ اﻧﺪازه ﮔﻴﺮي ‪7‬‬

‫ﻣﺸﺨﺺ ﮔﺮدﻳﺪ ﻛﻪ اﻟﮕﻮي ﻓﺼﻠﻲ ‪ 12‬ﻣﺎﻫﻪ اي در اﻳﻦ داده ﻫﺎ‬

‫آﻻﻳﻨﺪه ﻫﻮاي ﻛﺮﻣﺎن ﺑﺎ روش ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ ﻣﻮرد ﺑﺮرﺳﻲ‬

‫وﺟﻮد دارد و ﻣﺪل‪ ARIMA (0،0،0) (0،1،1)12‬ﭘﻴﺸﻨﻬﺎد‬

‫ﻗﺮار ﮔﺮﻓﺖ‪ .‬اﻳﻦ آﻻﻳﻨﺪه ﻫﺎ ﺷﺎﻣﻞ ‪CO, NO, NO2, NOx,‬‬

‫ﺷﺪ ﻛﻪ ﺗﻤﺎم ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ دار ﺑﻮدﻧﺪ‪.‬‬

‫‪ O3, SO2‬و ﮔﺮد و ﻏﺒﺎر )‪ (PM10‬ﺑﻮدﻧﺪ‪ .‬ﻣﺪﻟﻬﺎي ﭘﻴﺶ ﺑﻴﻨﻲ‬

‫)ﺷﻜﻞ ‪.(1 d‬‬

‫آرﻳﻤﺎ ﺑﺮاي ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﻛﺮﻣﺎن ﺑﺮاي ﻫﺮ‬

‫روﻧﺪ ‪ NOx‬ﻫﻢ در ﻃﻮل اﻳﻦ ﻣﺪت ﺛﺎﺑﺖ ﺑﻮد‪ .‬ﺑﺎ ﺗﺤﻠﻴﻞ داده ﻫﺎ‬

‫ﻛﺪام ﺟﺪاﮔﺎﻧﻪ ﺳﺎﺧﺘﻪ ﺷﺪ‪.‬‬

‫ﻣﺸﺨﺺ ﮔﺮدﻳﺪ ﻛﻪ اﻟﮕﻮي ﻓﺼﻠﻲ ‪ 12‬ﻣﺎﻫﻪ اي در اﻳﻦ داده ﻫﺎ‬

‫روﻧﺪ ﮔﺎز ﻣﻮﻧﻮاﻛﺴﻴﺪ ﻛﺮﺑﻦ در ﻃﻮل اﻳﻦ ‪ 4‬ﺳﺎل و ﻧﻴﻢ‬

‫وﺟﻮد دارد‪ .‬ﻧﻬﺎﻳﺘﺎ ﻣﺪل‪(0،0،0) (0،0،1)12‬‬

‫‪ARIMA‬‬

‫ﺑﺎ ﻛﺎﻫﺶ ﻫﻤﺮاه ﺑﻮد‪ .‬اﻳﻦ ﻛﺎﻫﺶ در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ دار‬

‫ﭘﻴﺸﻨﻬﺎد ﺷﺪ ﻛﻪ ﺗﻤﺎم ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ‬

‫ﺑﻮده و ﺑﺼﻮرت‬

‫دار ﺑﻮدﻧﺪ‪ .‬ﺑﻪ ﻧﻈﺮ ﻣﻲ رﺳﺪ ﻛﻪ در ﻛﺮﻣﺎن ﻣﻘﺪار ‪ NOx‬در‬

‫‪ y =2/056 – 0/02 ×t‬ﺑﻮد‪ .‬ﺑﺎ ﺗﺠﺰﻳﻪ ﻣﺪل ﻣﺸﺨﺺ ﮔﺮدﻳﺪ ﻛﻪ‬

‫ﺣﻮاﻟﻲ ﻣﺎﻫﻬﺎي ‪ 12‬و ‪) 11‬ﻧﻮاﻣﺒﺮ و دﺳﺎﻣﺒﺮ( از ﺑﻘﻴﻪ ﺑﻴﺸﺘﺮ‬

‫دوره ﺗﻨﺎوب ‪ 12‬ﻣﺎﻫﻪ اي در اﻳﻦ اﻃﻼﻋﺎت وﺟﻮد دارد و‬

‫اﺳﺖ‪) .‬ﺷﻜﻞ ‪.(1 e‬‬

‫اﻓﺰاﻳﺶ ﻣﻮﻧﻮ اﻛﺴﻴﺪ ﻛﺮﺑﻦ در ﻛﺮﻣﺎن ﻫﻤﺰﻣﺎن ﺑﺎ ﻣﺎﻫﻬﺎي‬

‫روﻧﺪ ‪ O3‬در ﻃﻮل اﻳﻦ ﻣﺪت ﺗﻘﺮﻳﺒﺎ ﺛﺎﺑﺖ ﺑﻮد و ﺷﻴﺐ ﺧﻂ‬

‫زﻣﺴﺘﺎن ﺑﻮده اﺳﺖ و ﻧﻬﺎﻳﺘﺎ ﺑﺮ اﺳﺎس ﻣﻘﺎدﻳﺮ ‪ ACF‬و ‪PACF‬‬

‫ﻧﺰدﻳﻚ ﺑﻪ ﺻﻔﺮ و ﻏﻴﺮ ﻣﻌﻨﻲ دار ﺑﻮد‪ .‬ﺳﭙﺲ ﺑﺎ ﺗﺤﻠﻴﻞ داده ﻫﺎ‬

‫داده ﻫﺎ اﻟﮕﻮي ﻓﺼﻠﻲ ﺿﺮﺑﻲ ‪ARIMA (0،1،1) (1،0،2)12‬‬

‫ﻣﺸﺨﺺ ﮔﺮدﻳﺪ ﻛﻪ اﻟﮕﻮي ﻓﺼﻠﻲ ‪ 12‬ﻣﺎﻫﻪ اي در اﻳﻦ داده ﻫﺎ‬

‫ﭘﻴﺸﻨﻬﺎد ﺷﺪ ﻛﻪ ﺗﻤﺎم ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ‬

‫وﺟﻮد دارد‪ .‬درﻧﻬﺎﻳﺖ ﻣﺪل ‪ARIMA (1،1،1) (1،0،1)12‬‬

‫دار ﺑﻮدﻧﺪ و ﻧﻤﻮدار ‪ ACF‬و ‪ PACF‬ﺑﺎﻗﻴﻤﺎﻧﺪه ﻫﺎ ﻫﻢ ﻣﻨﺎﺳﺐ‬

‫ﭘﻴﺸﻨﻬﺎد ﺷﺪ‪ .‬در اﻟﮕﻮي ﻓﺼﻠﻲ ﺗﻘﺮﻳﺒﺎ ‪ 12‬ﻣﺎﻫﻪ دﻳﺪه ﺷﺪه‪ ،‬ﺑﻪ‬

‫ﺑﻮدن اﻳﻦ ﻣﺪل را ﺗﺎﻳﻴﺪ ﻛﺮد )ﺷﻜﻞ ‪.(1 a‬‬

‫ﻧﻈﺮ ﻣﻲ رﺳﺪ ﻛﻪ ﻣﻴﺰان ازن در ﻫﻮاي ﻛﺮﻣﺎن در ﺣﻮاﻟﻲ ﻣﺎه ‪ 6‬و‬

‫روﻧﺪ ﮔﺮد و ﻏﺒﺎر در ﻃﻮل اﻳﻦ ‪ 4‬ﺳﺎل و ﻧﻴﻢ اﻓﺰاﻳﺸﻲ‬

‫‪ 7‬ﻳﻌﻨﻲ ژوﺋﻦ و ژوﺋﻴﻪ )ﻣﺎﻫﻬﺎي ﮔﺮم( اﻓﺰاﻳﺶ دارد ﻛﻪ ﺗﻤﺎم‬

‫ﺑﺼﻮرت ‪ y =117/1 + 0/313×t‬داﺷﺖ‪ .‬ﺑﺎ ﺗﺠﺰﻳﻪ ﻣﺪل‬

‫ﭘﺎراﻣﺘﺮﻫﺎي آن در ﺳﻄﺢ ‪ 0/05=α‬ﻣﻌﻨﻲ دار ﺑﻮدﻧﺪ‪) .‬ﺷﻜﻞ ‪(1 f‬‬

‫ﻣﺸﺨﺺ ﮔﺮدﻳﺪ ﻛﻪ اﻟﮕﻮي ﻓﺼﻠﻲ ﺣﺪوداً ‪ 8‬ﻣﺎﻫﻪ اي در اﻳﻦ‬

‫روﻧﺪ ‪ SO2‬در ﻃﻮل اﻳﻦ ‪ 4‬ﺳﺎل و ﻧﻴﻢ ﺗﻘﺮﻳﺒﺎ ﺛﺎﺑﺖ ﺑﻮد و ﺷﻴﺐ‬

‫اﻃﻼﻋﺎت وﺟﻮد دارد‪ ،‬داده ﻫﺎي ﺧﺎم ﻧﺸﺎن داد ﻛﻪ ﻣﻴﺰان ﮔﺮد و‬

‫ﺧﻂ ﻧﺰدﻳﻚ ﺑﻪ ﺻﻔﺮ و ﻏﻴﺮ ﻣﻌﻨﻲ دار ﺑﻮد‪ .‬در اﻳﻦ داده ﻫﺎ‬

‫ﻏﺒﺎر در ﻣﺎه ﭘﻨﺠﻢ )ﻣﻪ( و دﻫﻢ )اﻛﺘﺒﺮ( ﻳﻌﻨﻲ اواﺳﻂ ﺑﻬﺎر و ﭘﺎﻳﻴﺰ‬

‫وارﻳﺎﻧﺲ ﻫﺎ ﺗﻐﻴﻴﺮات ﻣﻌﻨﻲ داري داﺷﺖ و ﺑﺎ ‪Box-Cox‬‬

‫ﺑﻴﺸﺘﺮ از ﺑﻘﻴﻪ ﻣﺎﻫﻬﺎ ﺑﻮده اﺳﺖ‪ .‬ﻧﻬﺎﻳﺘﺎ ﻣﺪل ‪(0،1،1) (0،1،1)8‬‬

‫‪ Transformation‬ﺑﺼﻮرت ﺛﺎﺑﺖ در آﻣﺪ‪ .‬ﺳﭙﺲ اﻟﮕﻮي‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫‪ (Farrell 2004‬در اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﺑﺮاي رﺳﻴﺪن ﺑﻪ ﻣﺪل ﻧﻬﺎﻳﻲ‬

‫روﻧﺪ ‪ NO‬در ﻃﻮل اﻳﻦ ‪ 4‬ﺳﺎل و ﻧﻴﻢ ﺗﻘﺮﻳﺒﺎ ﺛﺎﺑﺖ ﺑﻮد‬

‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ ‪79 /....‬‬

‫ﻓﺼﻠﻲ ﺗﻘﺮﻳﺒﺎً ‪3‬ﻣﺎﻫﻪاي ‪ ARIMA (0،1،0) (1،0،1)3‬ﭘﻴﺸﻨﻬﺎد‬ ‫ﺷﺪ )ﺷﻜﻞ ‪.(1 g‬‬

‫ازن در اﺳﺘﺮاﺗﻮﺳﻔﺮ ﻳﻚ ﻣﺤﺎﻓﻆ اﺳﺖ اﻣﺎ در ﺗﺮوﭘﻮﺳﻔﺮ‬ ‫ﻳﻚ آﻟﻮده ﻛﻨﻨﺪه اﺳﺖ و از واﻛﻨﺶ ﺷﻴﻤﻴﺎﻳﻲ ﺑﻴﻦ ‪ NO2‬و‬ ‫اﻋﻼم ﺳﺎزﻣﺎن ﺑﻬﺪاﺷﺖ ﺟﻬﺎﻧﻲ در ‪ 2006‬ﻣﻘﺪار ‪100 µg/m³‬‬

‫ﺑﺤﺚ‬ ‫ﻣﺪﻟﻬﺎي ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ در ﺑﺴﻴﺎري از ﻣﻄﺎﻟﻌﺎت ﻋﻠﻮم‬

‫ازن ﺑﺮاي ﺣﺪاﻛﺜﺮ ﻣﻴﺎﻧﮕﻴﻦ روزاﻧﻪ ‪ 8‬ﺳﺎﻋﺘﻪ ﻣﺠﺎز ﻣﻲ ﺑﺎﺷﺪ‪ .‬از‬

‫ﻣﺤﻴﻄﻲ اﺳﺘﻔﺎده ﺷﺪه و ﻧﺘﺎﻳﺞ ﻣﻄﻠﻮﺑﻲ داﺷﺘﻪ اﺳﺖ ) ‪Kumar‬‬

‫اﻳﻦ ﻣﻘﺪار ﺑﻴﺸﺘﺮ ﺑﺴﺘﻪ ﺑﻪ ﺣﺴﺎﺳﻴﺖ ﻓﺮد ﻣﻲ ﺗﻮاﻧﺪ ﻋﻮارﺿﻲ‬

‫‪ .(and De Ridde 2010‬ﻣﻄﺎﻟﻌﺎت ﻧﺸﺎن داده ﻛﻪ ﻫﻴﭻ ﻣﺪل‬

‫اﻳﺠﺎد ﻛﻨﺪ‪ .‬ﺑﻨﺎﻳﺮاﻳﻦ ﻛﻨﺘﺮل ازن ﻫﻢ ﻣﺜﻞ ﺳﺎﻳﺮ آﻻﻳﻨﺪه ﻫﺎ‬

‫ﺧﻄﻲ ﺳﺎده اي ﺑﺮاي آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا وﺟﻮد ﻧﺪارد و ﻣﻌﻤﻮﻻ‬

‫ﻻزم اﺳﺖ)‪Francisco-‬‬

‫ﺑﺮاي ﻣﺪﻟﺴﺎزي اﻳﻦ آﻻﻳﻨﺪه ﻫﺎ از ﻣﺪﻟﻬﺎي ﭘﻴﭽﻴﺪه ﺗﺮ و از ﺟﻤﻠﻪ‬

‫‪.(Fernandez 2011‬‬

‫‪Quintela-del-Rio and‬‬

‫ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ اﺳﺘﻔﺎده ﻣﻲ ﺷﻮد ) ‪Kumar and De‬‬

‫در ﺣﺎﻟﻴﻜﻪ ‪ NO2‬ﻳﻚ ﻣﻨﺒﻊ ﺑﺮاي ﺗﻮﻟﻴﺪ ازن اﺳﺖ‪،‬‬

‫‪ .(Ridder 2010‬وﻗﺘﻲ ﻛﻪ ﻫﺪف ﻣﻄﺎﻟﻌﻪ ﭘﻴﺶ ﺑﻴﻨﻲ اﺳﺖ و‬

‫ﻧﻴﺘﺮوژن اﻛﺴﺎﻳﺪ )‪ (NO‬ازن را ﺗﺨﺮﻳﺐ ﻣﻲ ﻛﻨﺪ و ﺑﺮاي ﻫﻤﻴﻦ‬

‫ﻣﺘﻐﻴﺮ ﻣﻮرد ﺑﺮرﺳﻲ ﺗﺤﺖ ﺗﺎﺛﻴﺮ ﻋﻮاﻣﻞ ﻣﺘﻌﺪد ﻣﻲ ﺑﺎﺷﺪ و رواﺑﻂ‬

‫اﺳﺖ ﻛﻪ ﻣﻘﺎدﻳﺮ ازن در ﺷﻬﺮ ﻫﺎ ﺑﻪ اﻧﺪازه روﺳﺘﺎﻫﺎ ﺑﺎﻻ ﻧﻴﺴﺖ‪،‬‬

‫ﺑﻪ ﺧﻮﺑﻲ ﺷﻨﺎﺧﺘﻪ ﺷﺪه ﻧﻴﺴﺘﻨﺪ‪ ،‬ﻣﺜﻞ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا‪ ،‬روﺷﻬﺎي‬

‫زﻳﺮا ‪ NO‬ﺗﻮﺳﻂ اﺗﻮﻣﺒﻴﻠﻬﺎ ﺗﻮﻟﻴﺪ ﻣﻲ ﺷﻮد‪ .‬اﻟﺒﺘﻪ ﻋﻮاﻣﻞ دﻳﮕﺮ‬

‫ﺳﺮي ﻫﺎي زﻣﺎﻧﻲ ﺗﻚ ﻣﺘﻐﻴﺮه ﺑﻬﺘﺮ از ﻣﺪﻟﻬﺎي دﻳﮕﺮ ﻣﻲ ﺗﻮاﻧﻨﺪ‬

‫از ﺟﻤﻠﻪ ﺗﻌﺪاد ﺳﺎﻋﺘﻬﺎي ﻣﻮاﺟﻬﻪ ﺑﺎ ﻧﻮر ﺧﻮرﺷﻴﺪ‪ ،‬دﻣﺎ ‪ ،‬ﺑﺎد و‬

‫ﻛﺎر ﭘﻴﺶ ﺑﻴﻨﻲ را اﻧﺠﺎم دﻫﻨﺪ) ‪Kumar and De Ridder‬‬

‫ارﺗﻔﺎع وﺟﻮد دارد ﻛﻪ ﺗﻮﻟﻴﺪ ازن را ﺗﺤﺖ ﺗﺎﺛﻴﺮ ﻗﺮار ﻣﻲ دﻫﺪ‬

‫‪ .(2010‬در اﻳﻦ ﻣﺪﻟﻬﺎ ﺑﺮ اﺳﺎس داده ﻫﺎي ﻗﺒﻠﻲ ﻣﺤﺎﺳﺒﺎﺗﻲ‬

‫) ‪Quintela-del-Rio and Francisco-Fernandez‬‬

‫اﻧﺠﺎم ﻣﻲ ﺷﻮد و داده ﻫﺎﻳﻲ ﺑﺎ داﻣﻨﻪ اﻃﻤﻴﻨﺎن ﺑﺮاي آﻳﻨﺪه ﭘﻴﺶ‬

‫‪.(2011‬‬

‫ﺑﻴﻨﻲ ﻣﻲ ﺷﻮد‪ .‬ﻣﺪل ﺳﺮي زﻣﺎﻧﻲ ﺑﺮاي ﺳﺎﻟﻴﺎن زﻳﺎدي ﺑﺎ ﻣﻮﻓﻘﻴﺖ‬

‫در ﻃﺒﻴﻌﺖ ازن از ﺷﻮﻛﻬﺎي اﻟﻜﺘﺮﻳﻜﻲ ﻧﺎﺷﻲ از ﻃﻮﻓﺎنﻫﺎ‬

‫ﺑﺮاي ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا ﺑﻜﺎر رﻓﺘﻪ اﺳﺖ‬

‫در ﻻﻳﻪ ﻫﺎي ﺑﺎﻻي اﺗﻤﺴﻔﺮ ﻳﺎ ﺑﺨﺎﻃﺮ ﺗﺎﺛﻴﺮ اﺷﻌﻪ ﻣﺎورا ﺑﻨﻔﺶ‬

‫)‪.(Lumbreras et al. 2009‬‬

‫روي ﻣﻮﻟﻜﻮﻟﻬﺎي اﻛﺴﻴﮋن ﺗﻮﻟﻴﺪ ﻣﻲ ﺷﻮﻧﺪ‪ .‬ﺗﻮﻟﻴﺪ ﺑﻌﻀﻲ از ﻣﻮاد‬

‫ﻣﻄﺎﻟﻌﺎت ﻧﺸﺎن داده اﺳﺖ ﻛﻪ ازن در روزﻫﺎي ﮔﺮم ﻛﻪ‬

‫ﺷﻴﻤﻴﺎﻳﻲ از ﺟﻤﻠﻪ ﻛﻠﺮوﻓﻠﻮروﻛﺮﺑﻨﻬﺎ ﺗﻮﺳﻂ اﻧﺴﺎن ﺑﺎﻋﺚ‬

‫ﺗﺎﺑﺶ ﻧﻮر آﻓﺘﺎب زﻳﺎد اﺳﺖ و ﺑﺎد ﻛﻢ اﺳﺖ و ﻓﺸﺎر ﺳﻄﺤﻲ‬

‫ﺗﺨﺮﻳﺐ ازن ﻣﻲ ﺷﻮد ‪ .‬ﺑﺮ اﺳﺎس ﻗﺮاردادﻫﺎي ﺑﻴﻦ اﻟﻤﻠﻠﻲ ﺗﻮﻟﻴﺪ‬

‫ﺑﺎﻻﺳﺖ‪ ،‬اﻓﺰاﻳﺶ ﻣﻲ ﻳﺎﺑﺪ )‪ .(Duenas et al. 2005‬ﻣﻌﻤﻮﻻ‬

‫اﻳﻦ ﻣﻮاد ﺑﺎﻳﺪ ﻛﺎﻫﺶ ﻳﺎﺑﺪ ) ‪Quintela-del-Rio and‬‬

‫ﻣﻘﺪار ازن در ﺗﺎﺑﺴﺘﺎن ﺑﻴﺸﺘﺮ اﺳﺖ )‪(Duenas et al. 2005‬‬

‫‪.(Francisco-Fernandez 2011‬‬

‫و در ﻣﻄﺎﻟﻌﻪ ﻣﺎ ﻫﻢ ﻫﻤﻴﻦ ﻧﺘﻴﺠﻪ ﮔﺮﻓﺘﻪ ﺷﺪ‪ .‬ﻣﺪﻟﻬﺎي‬

‫ذرات ﻣﻌﻠﻖ ﻧﻴﺰ ﻳﻜﻲ از آﻻﻳﻨﺪه ﻫﺎي ﻣﻬﻢ ﻫﻮا در ﺷﻬﺮﻫﺎ‬

‫‪ ARIMA‬در ﻣﻄﺎﻟﻌﺎت دﻳﮕﺮ ﺑﺮاي ﺑﺮرﺳﻲ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮا‬

‫ﻣﺤﺴﻮب ﻣﻲ ﺷﻮﻧﺪ )‪ .(Goyal et al. 2006‬ﻣﻄﺎﻟﻌﺎت ﻣﺘﻌﺪد‬

‫از ﺟﻤﻠﻪ ازن ﺑﻜﺎر ﮔﺮﻓﺘﻪ ﺷﺪه اﻧﺪ‪ .‬از ﺟﻤﻠﻪ در ﻣﻄﺎﻟﻌﻪ اي در‬

‫ﻧﺸﺎن داده ﻛﻪ آﻻﻳﻨﺪهﻫﺎي ‪ PM10‬ﻓﺼﻠﻲ اﺳﺖ )‪(Liu 2009‬‬

‫اﺳﭙﺎﻧﻴﺎ ﻣﺪل ‪ ARIMA (1،0،0) (1،0،1)24‬ﺑﺮاي ﭘﻴﺸﮕﻮﺋﻲ‬

‫و در ﻣﻄﺎﻟﻌﻪ ﻣﺎ ﻫﻢ ﺣﺪاﻗﻞ دو ﭘﻴﻚ ﺳﺎﻟﻴﺎﻧﻪ ﺑﺮاي اﻳﻦ ذرات‬

‫ازن ﻣﻨﺎﻃﻖ ﺷﻬﺮي و ﻣﺪل‪ARIMA (2،1،1) (0،1،1)24‬‬

‫ﻣﺸﺎﻫﺪه ﺷﺪ‪ .‬ذرات ﻣﻌﻠﻖ ﻋﻼوه ﺑﺮ ﻛﺎﻫﺶ دﻳﺪ و رﺳﻮب در‬

‫ﺑﺮاي ﭘﻴﺸﮕﻮﺋﻲ ازن ﻣﻨﺎﻃﻖ روﺳﺘﺎﺋﻲ ﺑﻜﺎر ﮔﺮﻓﺘﻪ ﺷﺪ‬

‫ﻣﺤﻞﻫﺎي ﻣﺨﺘﻠﻒ از ﻃﺮﻳﻖ اﺳﺘﻨﺸﺎق ﻣﻲ ﺗﻮاﻧﻨﺪ ﺳﻼﻣﺖ اﻧﺴﺎن‬

‫)‪.(Duenas et al. 2005‬‬

‫را ﺑﻪ ﺧﻄﺮ اﻧﺪازﻧﺪ‪ .‬ﻣﻄﺎﻟﻌﺎت ﻣﺘﻌﺪدي اﺛﺮات ﻣﻀﺮ ﺑﺮ ﺳﻼﻣﺖ‬ ‫اﻧﺴﺎن در اﺛﺮ ذرات ﻣﻌﻠﻖ را ﮔﺰارش ﻛﺮده اﺳﺖ‬

‫) ‪Goyal‬‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫ﻫﻴﺪرو ﻛﺮﺑﻨﻬﺎ و ﻧﻮر ﺧﻮرﺷﻴﺪ اﻳﺠﺎد ﻣﻲ ﺷﻮد‪ .‬ﺑﺮ اﺳﺎس آﺧﺮﻳﻦ‬

‫‪ / 80‬ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري و ﻫﻤﻜﺎران‬

‫) اﺳﺘﺎﻧﺪارد وﺿﻌﻴﺖ ‪ 24‬ﺳﺎﻋﺘﻪ آﻻﻳﻨﺪه ﻫﺎ ﺑﺮ ﻃﺒﻖ اﺳﺘﺎﻧﺪارد‬

‫اﻛﺴﻴﺪ ﻫﺎي ﻧﻴﺘﺮوژن اﺳﺖ‪ .‬واﻛﻨﺸﻬﺎي ﻓﺘﻮ ﺷﻴﻤﻴﺎﻳﻲ ﺑﺎ اﻓﺰاﻳﺶ‬

‫‪ 24‬ﺳﺎﻋﺘﻪ ‪ AQI‬ﺑﺪﻳﻦ ﺗﺮﺗﻴﺐ ﺗﻌﺮﻳﻒ ﻣﻴﮕﺮدد‪ :‬وﺿﻌﻴﺖ‬

‫اﺷﻌﻪ ﺷﺮوع ﻣﻲ ﺷﻮد و ﺑﺎﻋﺚ اﻓﺰاﻳﺶ ‪ O3, HNO3‬و ‪SO4-²‬‬

‫ﺧﻮب ﻫﻮا ﺑﺮاي ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ )‪ ، (0 -3/4 PPM‬ﺑﺮاي‬

‫ﻣﻲ ﺷﻮﻧﺪ‪ ،‬ﺑﺨﺼﻮص ‪ HNO3‬ﻛﻪ ﻣﺤﻠﻮل در آب اﺳﺖ و‬

‫ذرات ﻣﻌﻠﻖ ﻛﻮﭼﻜﺘﺮ از ‪ 10‬ﻣﻴﻜﺮون) ‪،(0-39 µg/m3‬‬

‫ﻣﻲ ﺗﻮاﻧﺪ روي ﺳﻄﻮح ﻧﺸﺴﺖ ﻛﻨﺪ )‪ .(Liu 2009‬در ﻣﻄﺎﻟﻌﻪ‬

‫وﺿﻌﻴﺖ ﻣﺘﻮﺳﻂ ﻫﻮا ﺑﺮاي ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ) ‪-3/5PPM‬‬

‫‪ Liu‬و ﺑﻌﻀﻲ ﻣﻄﺎﻟﻌﺎت دﻳﮕﺮ )‪ (Liu 2009‬ﺑﻴﻦ ﻣﻘﺎدﻳﺮ‬

‫‪ ،(6/8‬ﺑﺮاي ذرات ﻣﻌﻠﻖ ﻛﻮﭼﻜﺘﺮ از ‪ 10‬ﻣﻴﻜﺮون)‪µ g/m3‬‬

‫اﻛﺴﻴﺪ ﻫﺎي ﻧﻴﺘﺮوژن و ذرات ﻣﻌﻠﻖ ﻫﻤﺒﺴﺘﮕﻲ ﻣﻌﻨﻲ داري دﻳﺪه‬

‫‪ ،( 99 - 40‬وﺿﻌﻴﺖ ﻏﻴﺮ ﺑﻬﺪاﺷﺘﻲ ﺑﺮاي ﮔﺮوه ﻫﺎي ﺣﺴﺎس‬

‫ﺷﺪ‪ .‬ﺧﻮﺷﺒﺨﺘﺎﻧﻪ در ﻣﺪل ﺳﺮي زﻣﺎﻧﻲ ﻧﻴﺎزي ﺑﻪ ﻓﺮض اﺳﺘﻘﻼل‬

‫در ﻣﻮرد ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ)‪ (10/3-6/9 PPM‬وﺑﺮاي ذرات‬

‫آﻻﻳﻨﺪه ﻫﺎ ﻧﻴﺴﺖ و ﻣﻲ ﺗﻮان از آن ﺑﺮاي ﭘﻴﺸﮕﻮﺋﻲ آﻻﻳﻨﺪه‬

‫ﻣﻌﻠﻖ ﻛﻮﭼﻜﺘﺮ از ‪10‬ﻣﻴﻜﺮون) ‪ ،(100- 199µg/m3‬وﺿﻌﻴﺖ‬

‫ﻫﺎﻳﻲ ﻛﻪ ﻫﻤﺒﺴﺘﮕﻲ دارﻧﺪ ﻧﻴﺰ اﺳﺘﻔﺎده ﻛﺮد )‪.(Liu 2009‬‬

‫وﺧﻴﻢ ﺑﺮاي ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ ﺑﻴﺸﺘﺮ از )‪ (10/4 PPM‬و‬

‫از ﻛﺎﺳﺘﻲ ﻫﺎي اﻳﻦ ﻣﻄﺎﻟﻌﺎت وﺟﻮد روزﻫﺎﻳﻲ ﺑﻮد ﻛﻪ در آن‬

‫ﺑﺎﻻﺧﺮه وﺿﻌﻴﺖ وﺧﻴﻢ ﺑﺮاي ذرات ﻣﻌﻠﻖ ﻛﻮﭼﻜﺘﺮ از ‪10‬‬

‫ﻣﻴﺰان آﻻﻳﻨﺪه ﻫﺎ ﺑﻪ ﻋﻠﺖ ﺧﺮاﺑﻲ دﺳﺘﮕﺎه اﻧﺪازه ﮔﻴﺮي ﻧﺸﺪه ﺑﻮد‬

‫ﻣﻴﻜﺮون ﺑﻴﺸﺘﺮ از)‪ (200 g/mµ 3‬ﻣﻲ ﺑﺎﺷﺪ‪ .‬ﺑﺮاي آﻻﻳﻨﺪه ﻫﺎ‬

‫و اﺳﺘﻔﺎده از ﻣﻴﺎﻧﮕﻴﻦ ﻣﺎﻫﺎﻧﻪ ﺗﺎ ﺣﺪي اﻳﻦ ﻣﺸﻜﻞ را ﻣﺮﺗﻔﻊ ﻛﺮد‪.‬‬

‫وﺿﻌﻴﺖ ذرات ﻣﻌﻠﻖ در ﻛﺮﻣﺎن در ﺑﺴﻴﺎري از ﻣﺎﻫﻬﺎي ﺳﺎل‬ ‫ﺑﺎﻻﺗﺮ از ‪ 100 µg/m3‬و ﻏﻴﺮ ﺑﻬﺪاﺷﺘﻲ اﺳﺖ و ﻣﻄﺎﻟﻌﻪ ﺣﺎﺿﺮ‬

‫ﻧﺘﻴﺠﻪ ﮔﻴﺮي‬

‫ﻧﺸﺎن داد ﻛﻪ ﮔﺮد و ﺧﺎك در ﻫﻮاي ﻛﺮﻣﺎن رو ﺑﻪ اﻓﺰاﻳﺶ اﺳﺖ‪.‬‬

‫ﺧﻮﺷﺒﺨﺘﺎﻧﻪ ﻣﻴﺰان ﺗﻮﻟﻴﺪ ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ در ﻛﺮﻣﺎن رو ﺑﻪ‬

‫در ﻣﻄﺎﻟﻌﻪ ﻣﺎ ﻣﻘﺪار ﻣﻮﻧﻮﻛﺴﻴﺪ ﻛﺮﺑﻦ در زﻣﺴﺘﺎﻧﻬﺎ ﺑﻴﺸﺘﺮ ﺑﻮد‬

‫ﻛﺎﻫﺶ اﺳﺖ‪ ،‬و در ﺣﺎل ﺣﺎﺿﺮ ﻋﻤﺪه ﺗﺮﻳﻦ ﻣﺸﻜﻞ آﻟﻮدﮔﻲ‬

‫ﻛﻪ ﺑﻪ اﺣﺘﻤﺎل زﻳﺎد ﺑﻪ ﺧﺎﻃﺮ اﺳﺘﻔﺎده از وﺳﺎﺋﻞ ﮔﺮم ﻛﻨﻨﺪه در‬

‫ﻫﻮا در ﻛﺮﻣﺎن ذرات ﻣﻌﻠﻖ اﺳﺖ و ﻣﺪل ﺑﺪﺳﺖ آﻣﺪه ﻧﻴﺰ ﻧﺸﺎن‬

‫زﻣﺴﺘﺎن اﺳﺖ‪ .‬ﻣﻄﺎﻟﻌﻪ ‪ Lau‬و ﻫﻤﻜﺎران ﻫﻢ ﻧﺸﺎن داد ﻛﻪ ﻣﻮﻧﻮ‬

‫ﻣﻲ دﻫﺪ ﻛﻪ ﻋﻼوه ﺑﺮ اﻓﺰاﻳﺶ در ﻣﻘﺎﻃﻊ زﻣﺎﻧﻲ ﺧﺎص‪ ،‬روﻧﺪ‬

‫اﻛﺴﻴﺪ ﻛﺮﺑﻦ در ﺗﺎﺑﺴﺘﺎن ﻛﻤﺘﺮ اﺳﺖ و در ﻓﺼﻞ زﻣﺴﺘﺎن و در‬

‫ﻛﻠﻲ آن ﻧﻴﺰ رو ﺑﻪ اﻓﺰاﻳﺶ اﺳﺖ‪ .‬ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ اﺛﺮات ﻣﻨﻔﻲ اﻳﻦ‬

‫ﻣﻜﺎﻧﻬﺎي ﺑﺎ ﺗﺮاﻓﻴﻚ ﺑﺎﻻ ﺑﻴﺸﺘﺮ اﺳﺖ‪ .‬ﻣﻮﻧﻮ اﻛﺴﻴﺪ ﻛﺮﺑﻦ ﻳﻚ‬

‫آﻻﻳﻨﺪه ﻫﺎ ﺑﺮ ﺳﻼﻣﺖ ﻻزم اﺳﺖ ﺑﺮﻧﺎﻣﻪ رﻳﺰي ﻣﻨﺎﺳﺐ ﺑﺮاي‬

‫آﻟﻮده ﻛﻨﻨﺪه اﺻﻠﻲ ﺷﻬﺮ ﻫﺎي ﺑﺰرگ اﺳﺖ و ﻣﻲ ﺗﻮاﻧﺪ ﺑﺮ‬

‫ﻣﻘﺎﺑﻠﻪ ﺑﺎ آن ﺑﺨﺼﻮص در ﻓﺼﻮل ﺧﺎص ﺻﻮرت ﮔﻴﺮد‪ .‬درﺧﺖ‬

‫ﺳﻼﻣﺖ اﻧﺴﺎن ﺗﺎﺛﻴﺮات ﻣﻨﻔﻲ ﻣﺘﻌﺪد داﺷﺘﻪ ﺑﺎﺷﺪ) ‪Lau et al.‬‬

‫ﻛﺎري در اﻃﺮاف ﺷﻬﺮ ﻛﺮﻣﺎن و ﺑﺨﺼﻮص اﺳﺘﻔﺎده از درﺧﺘﺎن‬

‫‪ .(2009‬در ﻣﺪل ﻣﺎ ﻣﻮﻧﻮ اﻛﺴﻴﺪ ﻛﺮﺑﻦ ﻳﻚ ﺳﻴﺮ ﻧﺰوﻟﻲ ﻫﻢ در‬

‫ﻛﻮﻳﺮي ﻛﻪ ﺑﺎ آب و ﻫﻮاي ﻣﻨﻄﻘﻪ ﺳﺎزﮔﺎري دارﻧﺪ ﻣﻲ ﺗﻮاﻧﺪ از‬

‫ﻃﻮل اﻳﻦ ﭼﻨﺪ ﺳﺎل ﻧﺸﺎن داد ﻛﻪ ﻳﻚ ﺗﻮﺟﻴﻪ آن اﺟﺮاي ﻃﺮح‬

‫ﺟﻤﻠﻪ اﻳﻦ اﻗﺪاﻣﺎت ﺑﺎﺷﺪ‪.‬‬

‫ﺗﻌﻮﻳﺾ ﺧﻮدروﻫﺎي ﻓﺮﺳﻮده ﻣﻲ ﺑﺎﺷﺪ‪.‬‬ ‫در ﺑﻌﻀﻲ از ﻣﻄﺎﻟﻌﺎت )‪ (Liu 2009‬ﻣﻘﺪار ازن و‬ ‫اﻛﺴﻴﺪ ﻫﺎي ﻧﻴﺘﺮوژن ﺑﺎ ‪ PM10‬ﻫﻤﺒﺴﺘﮕﻲ داﺷﺘﻨﺪ ﻛﻪ اﺣﺘﻤﺎﻻ‬ ‫ﺑﻪ ﺧﺎﻃﺮ واﻛﻨﺸﻬﺎي ﻓﺘﻮ ﺷﻴﻤﻴﺎﻳﻲ اﺳﺖ ﻛﻪ ﺑﺎ اﻓﺰاﻳﺶ اﺷﻌﻪ‬ ‫ﺷﺮوع ﻣﻲ ﺷﻮد‪ .‬اﻛﺴﻴﺪ ﻫﺎي ﻧﻴﺘﺮوژن ﻣﻲ ﺗﻮاﻧﻨﺪ در ﺷﺮاﻳﻂ‬ ‫ﻣﺮﻃﻮب ﺑﻪ ﺻﻮرت ذرات در آﻳﻨﺪ و ﺑﺎﻋﺚ اﻓﺰاﻳﺶ ‪PM10‬‬ ‫ﺷﻮﻧﺪ )‪ .(Liu 2009‬ﻳﻚ ﻓﺮﺿﻴﻪ در ﻣﻮرد ﻫﻤﺒﺴﺘﮕﻲ ﺑﻴﻦ‬

‫ﺗﺸﻜﺮ و ﻗﺪر داﻧﻲ‬ ‫ﻧﻮﻳﺴﻨﺪﮔﺎن اﻳﻦ ﻣﻘﺎﻟﻪ از ﺳﺎزﻣﺎن ﺣﻔﺎﻇﺖ ﻣﺤﻴﻂ زﻳﺴﺖ ﻛﺮﻣﺎن‬ ‫و ﺑﺨﺼﻮص ﺧﺎﻧﻢ ﻣﻬﻨﺪس ﺑﺎﺑﻜﻲ ﺑﻪ ﺧﺎﻃﺮ ﺣﺴﻦ ﻫﻤﻜﺎري‬ ‫ﺷﺎن ﻛﻤﺎل ﺗﺸﻜﺮ را دارﻧﺪ‪ .‬اﻳﻦ ﻣﻄﺎﻟﻌﻪ ﺗﻮﺳﻂ ﻛﻤﻴﺘﻪ ﺗﺤﻘﻴﻘﺎﺗﻲ‬ ‫ﭘﺰﺷﻜﻲ ﻣﺤﻴﻄﻲ داﻧﺸﻜﺪه ﺑﻬﺪاﺷﺖ ﺗﺼﻮﻳﺐ و ﺗﻮﺳﻂ ﻣﻌﺎوﻧﺖ‬ ‫ﭘﮋوﻫﺸﻲ داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﻜﻲ ﻛﺮﻣﺎن ﺗﺎﻣﻴﻦ اﻋﺘﺒﺎر ﺷﺪ‪.‬‬

‫‪Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017‬‬

‫‪ .(et al. 2006‬ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ اﺳﺘﺎﻧﺪارد ‪ 24‬ﺳﺎﻋﺘﻪ ‪AQI‬‬

‫اﻛﺴﻴﺪ ﻫﺎي ﻧﻴﺘﺮوژن ﺑﺎ ‪ PM10‬ﺑﻪ ﺧﺎﻃﺮ واﻛﻨﺸﻬﺎي ﺷﻴﻤﻴﺎﻳﻲ‬

81 /.... ‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ‬





 ‫ ا‬

‫ ا‬

١/٥٢ ١٢٥/٦٩ ٠/٠١٧٣ ٠/٠٢٥٩ ٠/٠٣٧٦ ٠/٠١٥٦٥ ٠/٠٠٩٦

١/٤٨ ١٢٣/٥٧ ٠/٠١٢٢ ٠/٠٢٣٨ ٠/٠٣٣٣ ٠/٠١٤١٥ ٠/٠٠٨٢

٠/٢٩٣١ ٥٤/٣٨ ٠/٠٠٢٦ ٠/٠٠١٥ ٠/٠١٠٢ ٠/٠٠٤٤٥ ٠/٠٠٠٤

٢/٩ ٢٣٤/٠١ ٠/٠٦٨٣ ٠/٠ ٨ ٠/١١٨١ ٠/٠٣١٧ ٠/٠٢٥٣٥

  CO (ppm) PM10 (µg/m³) NO (ppm) NO2 (ppm) NOx (ppm) O3 (ppm) SO2 (ppm)

‫ )واﺣﺪ زﻣﺎن در ﻣﺤﻮر اﻓﻘﻲ‬.‫ ﭘﻴﺶ ﺑﻴﻨﻲ ﺗﻐﻴﻴﺮات آﻻﻳﻨﺪه ﻫﺎ در ﻫﻮاي ﻛﺮﻣﺎن ﻃﺒﻖ اﻟﮕﻮي ﺳﺮي زﻣﺎﻧﻲ ﭘﻴﺸﻨﻬﺎد ﺷﺪه‬:1 ‫ﺷﻜﻞ‬ (.‫ ﻣﻲ ﺑﺎﺷﺪ‬1389 ‫ ﺗﺎ ﺷﻬﺮﻳﻮر‬1385 ‫ﻣﺎه اﺳﺖ و از ﻓﺮوردﻳﻦ‬

Time Series Plot for CO (with forecasts and their 95% confidence limits) 4 3

CO

2 1 0 -1 -2 1

6

12

18

24

30

36 42 Time

48

54

60

66

72

(a) CO

Time Series Plot for Dust (with forecasts and their 95% confidence limits)

300

200 Dust

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

‫ وﺿﻌﻴﺖ آﻻﻳﻨﺪه ﻫﺎي ﺷﻬﺮ ﻛﺮﻣﺎن در دوره ﻣﻮرد ﺑﺮرﺳﻲ‬-1 ‫ﺟﺪول‬

100

0

-100 1

5

10

15

20

25

30 35 Time

40

45

50

55

60

(b) PM10

Time Series Plot for NO (with forecasts and their 95% confidence limits) 0.20 0.15

NO

0.10 0.05 0.00 -0.05 -0.10 1

6

12

18

24

30

36 42 Time

48

54

60

66

72

( c) NO

Time Series Plot for NO2 (with forecasts and their 95% confidence limits) 0.08 0.07 0.06

NO2

0.05 0.04 0.03 0.02 0.01 0.00 -0.01 1

6

12

18

24

30

36 42 Time

48

54

60

66

72

(d) NO2

Time Series Plot for NOx (with forecasts and their 95% confidence limits) 0.125 0.100 0.075 NOx

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

‫ ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري و ﻫﻤﻜﺎران‬/ 82

0.050 0.025 0.000

1

6

12

18

24

30

36 42 Time

48

54

60

66

72

(e) NOx

83 /.... ‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ‬

Time Series Plot for O3 (with forecasts and their 95% confidence limits)

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

0.04

0.03

O3

0.02

0.01

0.00

-0.01 1

6

12

18

24

30

36 42 Time

48

54

60

66

72

(f) O3

Time Series Plot for SO2 (with forecasts and their 95% confidence limits) 0.10

SO2

0.05

0.00

-0.05

-0.10 1

5

10

15

20

25

30 Time

35

40

45

50

55

(g) SO2

References Anderson, H.R., 2009. Air pollution and mortality: A history. Atmospheric Environment, 43, pp. 142-152. Box, GEP. and Jenkins, G.M., 1976. Time series analysis: forecasting and control, San Francisco, Holden Day Pulications. Duenas, C., Fernandez, M.C., Canete, S., Carretero,Liger E, 2005. Stocastic model to forecast ground level ozone concentration at

urban and rural areas. Chemosphere, 61(10), pp. 1379-1389. Ghorbani, M. and Younesian, M., 1389. Research Projects in Air pollution Epidemiology. Iranian Epidemiology Journal. 5, pp. 44-52 [In Persian]. Goyal, P., Chan, A.T. and Jaiswal, N., 2006. Statistical models for the prediction of respirable suspended particulate matter in

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

‫ ﻓﺎﻃﻤﻪ ﻣﻨﺼﻮري و ﻫﻤﻜﺎران‬/ 84

urban cities. Atmospheric Environment, 40, pp. 2068-2077. Hamilton, JD., 1994. Time series analysis, Princeton Publications, USA. Hosseinpour, A.R., Forouzanfar, M.H., Yunesian, M., Asghari, F., Holakouie Naieni, K. and Farhood, D., 2005. Air pollution and hospitalization due to angina pectoris in Tehran, Iran: A time-series study. Environmental Research, 99, pp. 126-131 [In Persian]. Ingrisch, M., Sourbron, S., Reiser, M.F. and Peller, M., 2009. Model selection in dynamic contrast enhanced MRI: The Akaike Information Criterion. In Dössel, O. and Schlegel, WC. (Eds.) IFMBE Proceedings 25/IV Khosravi Dehkordi, A. and Modarres, R., 1386. Time Series analysis of the daily air pollution in Isfahan from the Petrolium Industry. Mohit shenasi. 33, pp. 33-42 [In Persian]. Kumar, U. and De Ridder, K., 2010. GARCH modelling in association with FFT-ARIMA to forecast ozone episodes. Atmospheric Environment, 44, pp. 4252-4265. Lau, J.C., Hung, W.T., Yuen, D.D. and Cheung, C.S., 2009. Long memory characteristics of urban roadside air quality. Transportation Research Part D, 14, pp. 353359. Liang, W., Wei, H. and Kuo, H., 2009. Association between daily mortality from respiratory and cardiovascular diseases and air pollution in Taiwan. Environmental Research, 109, pp. 51-58. Liu, P.G., 2009. Simulation of the daily average PM10 concentrations at Ta-Liao with Box–Jenkins time series models and

multivariate analysis. Atmospheric Environment, 43, pp. 2104-2113. López-Villarrubia, E., Ballester, F., Iñiguez, C. and Peral, N., 2010. Air pollution and mortality in the Canary Islands:a time-series analysis. Environmental Health, 9. Lumbreras, J., Garcia-Martos, C., Mira, J. and Borge, R., 2009. Computation of uncertainty for atmospheric emission projections from key pollutant sources in Spain. Atmospheric Environment, 43, pp. 1557-1564. Masjedi, M.R., Jamaati, H.R., Dokoohki, P., Ahmadzadeh, Z., Alinejad Taheri, S., Bigdeli, M., Agin, K., Ghavam, S.M., Rostiman, A. and Izadi S., 2001. The correlation between air pollution and acute respiratory and cardiac attacks. Pazhoohesh dar pezeshki, 25, pp. 25-33 [In Persian]. Nasrollhi, Z. and Ghaffari Goolak, M., 2010. Air pollution and its effective factors. Faslnameh Pazhoohesh Eghtesadi, 3, pp. 375-395 [In Persian]. Quintela-del-Rio, A. and Francisco-Fernandez, M., 2011. Nonparametric functional data estimation applied to ozone data: Prediction and extreme value analysis. Chemosphere, 82, pp. 800-808. Rajarathnam, U., Sehgal M., Nairy S., Patnayak R.C., Chhabra S.K., Kilnani, K.V., R and Committee., HHR, 2011. Time Series study on air pollution and mortality in Dehli. Res Rep Health Eff Inst, Mar, pp. 47-74. Samet, J.M., Dominici, F., Zeger, S.L., Schwartz, J. and Dockery, D.W., 2000. The national morbidity, mortality and air pollution study. Part 1: Methods and Methodologic Issues. Research Report 94 Cambridge, MA, Health Effects institute.

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

85 /.... ‫ﺑﺮرﺳﻲ و ﭘﻴﺶ ﺑﻴﻨﻲ وﺿﻊ آﻻﻳﻨﺪه ﻫﺎي ﻫﻮاي ﺷﻬﺮ‬

Sharma, P., Chandra, A. and Kaushik, S.C., 2009. Forecasts using Box–Jenkins models for the ambient air quality data of Delhi City. Environ Monit Assess, 157, pp. 105–112. Wagemakers, E. and Farrell, S., 2004. AIC model selection using Akaike weights. Psychonomic Bulletin and Review, 11, pp. 192-196.

Zhang, F., Wang, W., Lv, J., Krafft, T. and Xu, J., 2011. Time-series studies on air pollution and daily outpatient visits for allergic rhinitis in Beijing, China. Science of the Total Environment, 409, pp. 2486–2492.

Downloaded from sjsph.tums.ac.ir at 22:26 IRST on Wednesday February 8th 2017

Scientific Journal of School of Public Health and Institute of Public Health Research Vol. 11, No. 2, Summer 2013

/86

Forecasting air pollutant situation using the time series models in Kerman, Iran Mansouri, F., MS.c. Student, Dept of Environmental Health Engineering, Faculty of Kerman Medical University, Kerman, Iran

Public Health,

Khanjani, N., Ph.D. Assistant Professor, Department of Epidemiology and Department of Environmental Health, Faculty of Public Health, Kerman Medical University, Kerman, Iran - Corresponding author: [email protected]

Rananadeh Kalankesh, L., MS.c. Student, Department of Environmental Health Engineering, Faculty of Public Health, Kerman Medical University, Kerman, Iran Pourmousa, R., MS.c. Lecturer, Department of Statistics, School of Mathematics and Statistics, Shahid Bahonar University, Kerman, Iran

Received: Apr 3, 2012

Accepted: Feb 14, 2013

ABSTRACT Background and Aim: Air pollution, one of the most important problems in big cities in developing countries, can have several negative health effects on humans. Therefore, studying air pollutants can help in developing programs for air pollution control. The aim of this study was to determine and predict the changes of air pollutants in the City of Kerman, Iran. Materials and Methods: In this ecological study, data about seven important air pollutants in Kerman, including carbon monoxide (CO), nitrogen oxides (NO, NO2 and NOx), particulate matter (PM10), sulfur dioxide (SO2) and ozone (O3) , between March 2006 and September 2010 was obtained from the Kerman Province Environmental Protection Agency. Monthly averages were calculated for all the pollutants and, using univariate time series models, predictions were made for changes in each pollutant. Results: All of the air pollutants had a steady trend in Kerman, except CO and PM10, the former having a decreasing, and that latter an increasing, trend. All of the pollutants had a seasonal pattern. Time series models with a 12-, 3-, 8-, 12-, 12-, 12- and 6-month seasonal patterns were fit for O3, SO2, PM10, NOx, NO2, CO and NO, respectively. Conclusion: The production of ambient CO is decreasing in Kerman, a probable reason being discontinuing use of old automobiles. However, PM10 is increasing and in some seasons it is above standard; therefore, strategies should be adopted to control this problem. Key words: Air pollution, particulate matter, PM10, Kerman, Time Series, Sulfur dioxide, Ozone, Nitrogen oxides, Carbon monoxide