ارزيابي روند تخريب اراضي جنگلي با استفاده از الگوريتم شبكه عصبي مصنوعي (مورد مطالعه: جنگل هاي فندقلو شهرستان نمين)

Evaluation Of Deforestation Process Using Artificial Neural Networks Algorithm (Case Study: Namin County Hazelnut Forests)

ﭼﮑﯿﺪه ﺷﻬﺮﺳﺘﺎن ﻧﻤﯿﻦ در اﺳﺘﺎن اردﺑﯿﻞ داراي اراﺿﯽ وﺳﯿﻌﯽ از ﮔﻮﻧﻪﻫﺎي ﺟﻨﮕﻠﯽ ﻓﻨﺪﻗﻠﻮ اﺳﺖ. ﺑﻨﺎﺑﺮاﯾﻦ ﻫﺪف از اﯾﻦ ﺗﺤﻘﯿﻖ، ارزﯾﺎﺑﯽ روﻧﺪ ﺗﺨﺮﯾﺐ اراﺿـﯽ ﺟﻨﮕﻞﻫﺎي ﻓﻨﺪﻗﻠﻮ در ﺷـﻬﺮﺳﺘﺎن ﻧﻤﯿﻦ ﺑﺎ اﺳﺘﻔﺎده از ﺗﺼﻮﯾﺮ ﻣﺎﻫﻮارهاي LANDSAT ﺳﻨﺠﻨﺪه+ETM در ﺳﺎل 1386 و ﺗﺼﻮﯾﺮ ﺳﻨﺠﻨﺪه OLI ﻣﺮﺑﻮط ﺑﻪ ﺳﺎل 1396 اﺳﺖ. ﭘﺲ از اﻧﺠﺎم ﭘﯿﺶﭘﺮدازشﻫﺎي ﻟﺎزم، ﺑﻪ ﻣﻨﻈﻮر ﺑﺎﻟﺎ ﺑﺮدن ﻗﺪرت ﺗﻔﮑﯿﮏ ﻣﮑﺎﻧﯽ ﺗﺼـﺎوﯾﺮ از روشIHS ﺟﻬﺖ ادﻏﺎم ﺗﺼـﻮﯾﺮ ﭘﺎﻧﮑﺮوﻣﺎﺗﯿﮏ و ﭼﻨﺪﻃﯿﻔﯽ ﻫﺮ دو ﺳـﻨﺠﻨﺪه اﺳـﺘﻔﺎده ﺷﺪ. ﻃﺒﻘﻪﺑﻨﺪي ﺗﺼـﺎوﯾﺮ ﺑﻪ ﺻﻮرت ﻧﻈﺎرت ﺷﺪه و از روي ﺗﺼﺎوﯾﺮ ادﻏﺎم ﺷﺪه ﮐﻪ داراي ﻗﺪرت ﺗﻔﮑﯿﮏ ﻣﮑﺎﻧﯽ 15 ﻣﺘﺮ ﺑﻮده، ﺑﺎ اﻟﮕﻮرﯾﺘﻢ ﺷﺒﮑﻪ ﻋﺼـﺒﯽ ﻣﺼـﻨﻮﻋﯽ اﻧﺠﺎم ﺷـﺪ. ﻧﺘﺎﯾﺞ ﻧﺸﺎن داد ﮐﻪ از ﮐﻞ ﻣﺴﺎﺣﺖ ﮐﺎرﺑﺮيﻫﺎي ﻃﺒﻘﻪﺑﻨﺪي ﺷﺪه در ﺳﺎل 1379، ﻣﻌﺎدل 9990/06 ﻫﮑﺘﺎر ﺑﻪ اراﺿـﯽ ﺟﻨﮕﻠﯽ ﺗﻌﻠﻖ داﺷـﺘﻪ اﺳـﺖ ﮐﻪ در ﻣﻘﺎﯾﺴـﻪ ﺑﺎ ﺳـﺎل 1386 اﯾﻦ ﻣﯿﺰان ﺑﻪ ﻣﻌﺎدل 9531/74 ﻫﮑﺘﺎر ﮐﺎﻫﺶ داﺷﺘﻪ اﺳـﺖ. ﻫﻢﭼﻨﯿﻦ وﺳـﻌﺖ اراﺿﯽ اﯾﻦ ﺟﻨﮕﻞﻫﺎ ﮐﻪ از ﺳﺎل 1386 ﺗﺎ ﺳﺎل 1396 ﺑﯿﺶ از7700 ﻫﮑﺘﺎر ﮐﺎﻫﺶ ﯾﺎﻓﺘﻪ اﺳﺖ. ﺑﻨﺎﺑﺮاﯾﻦ ﺑﺮاي ﺣﻔﻆ ﻋﺮﺻﻪﻫﺎي ﻃﺒﯿﻌﯽ، ﺗﺜﺒﯿﺖ و ﻗﺎﻧﻮﻧﯽ ﮐﺮدن ﮐﺎرﺑﺮي اراﺿﯽ در دﺳﺘﻮر ﮐﺎر ﻣﺘﺨﺼﺼﺎن و ﻣﺴﺌﻮﻟﺎن ﮐﺸﻮر ﻗﺮار ﮔﯿﺮد.

Namin county in Ardebil province has a large lands of hazelnut forest species, that it is one of the rich habitats of Hyrcanian forests. Therefore, the purpose of this study is to evaluate the land degradation process of hazelnut forests in Namin County using Landsat, ETM+ images in 2007 and OLI images in 2017. Pre-processing was performed to enhance the spatial resolution of images using the IHS method to fusion both of the panchromatic and multispectral images of sensors. The images were classified into supervised and fused images with a spatial resolution of 15 m and performed with artificial neural network algorithm. The results showed that out of the total classified study area in the year 2000, 9999.06 ha belonged to forest land, which decreased to 9531.74 ha compared to 2007. Also the area of these forests decreased from 7,700 hectares from 2007 to 1396. Therefore, to preserve natural landscapes, land consolidation and legalization should be on the agenda of national experts and officials.