ﺑﺮرﺳﯽ ﮐﺎرآﯾﯽ دادهﻫﺎي ﻣﺎﻫﻮارهايGeoEye-1 ﺑﺮاي ﺗﻬﯿﻪ ﻧﻘﺸﻪ ﻣﺮاﺣﻞ ﺗﺤﻮﻟﯽ در ﺟﻨﮕﻞﻫﺎي ﺷﻤﺎل اﯾﺮان

Investigation of the efficiency of GeoEye-1 satellite data for mapping the development stages in the forests of Northern Iran

اﻣﺮوزه در ﻋﻠﻢ ﺳﻨﺠﺶ از دور ﺑﺎ وﺟﻮد ﺳﻨﺠﻨﺪهﻫﺎي ﺑﺎ ﺗﻮان ﺗﻔﮑﯿﮏ ﻣﮑﺎﻧﯽ و ﻃﯿﻔﯽ ﺑﺎﻻ و ﺑﻪﮐﺎرﮔﯿﺮي اﻧﻮاع ﺷﺎﺧﺺﻫﺎي ﮔﯿﺎﻫﯽ، اﻣﮑﺎن ﺷﻨﺎﺳﺎﯾﯽ و ﺗﻔﮑﯿﮏ دﻗﯿﻖﺗﺮ ﺳﺎﺧﺘﺎر ﭘﺪﯾﺪهﻫﺎي ﻃﺒﯿﻌﯽ ﺑﺎ اﺳﺘﻔﺎده از دادهﻫﺎي ﻣﺎﻫﻮارهاي ﻓﺮاﻫﻢ ﺷﺪه اﺳﺖ. ﺑﺮ اﯾﻦ اﺳﺎس ﮐﺴﺐ اﻃﻼﻋﺎت از ﺳﺎﺧﺘﺎر ﺟﻨﮕﻞﻫﺎ ﺑﺎ اﺳﺘﻔﺎده از دادهﻫﺎي ﻣﺎﻫﻮارهاي ﺑﺮاي ﮐﻤﮏ ﺑﻪ ﻣﺪﯾﺮﯾﺖ ﻣﻨﺎﺑﻊ ﺟﻨﮕﻠﯽ ﺗﺠﺪﯾﺪﭘﺬﯾﺮ ﻣﻮرد ﺗﻮﺟﻪ ﻣﺪﯾﺮان و ﻣﺤﻘﻘﯿﻦ ﻗﺮار دارد. در اﯾﻦ ﺗﺤﻘﯿﻖ ﺑﺎ ﺑﻬﺮهﮔﯿﺮي از دادهﻫﺎي ﻣﺎﻫﻮاره 1-GeoEye و اﻟﮕﻮرﯾﺘﻢﻫﺎي ﻃﺒﻘﻪﺑﻨﺪي ﻣﺨﺘﻠﻒ در روشﻫﺎي ﭘﯿﮑﺴﻞﭘﺎﯾﻪ، ﺷﯽﭘﺎﯾﻪ و ﻣﺪلﭘﺎﯾﻪ ﻧﺴﺒﺖ ﺑﻪ ﺗﻬﯿﻪ ﻧﻘﺸﻪﻫﺎي ﻣﺮاﺣﻞ ﺗﺤﻮﻟﯽ ﺟﻨﮕﻞﻫﺎي ﻃﺒﯿﻌﯽ در ﺷﻤﺎل ﮐﺸﻮر اﻗﺪام ﺷﺪ. ﺻﺤﺖ ﻧﻘﺸﻪﻫﺎي ﺧﺮوﺟﯽ ﺑﺎ ﭘﺎراﻣﺘﺮﻫﺎي ﺻﺤﺖ ﮐﻠﯽ و ﺿﺮﯾﺐ ﮐﺎﭘﺎ ارزﯾﺎﺑﯽ ﮔﺮدﯾﺪ. ﻧﺘﺎﯾﺞ ﻧﺸﺎن داد ﮐﻪ ﺣﺪاﮐﺜﺮ دﻗﺖ )78 درﺻﺪ( در روش ﺷﯽﭘﺎﯾﻪ ﻣﺘﮑﯽ ﺑﻪ ﻗﻄﻌﻪﺑﻨﺪي ﻧﺎﻇﺮ ﺑﻪ ﺷﺎﺧﺺ NDVI و اﻟﮕﻮرﯾﺘﻢ ﺣﺪاﮐﺜﺮ اﺣﺘﻤﺎل ﻣﯽ ﺑﺎﺷﺪ. ﻫﻤﭽﻨﯿﻦ اﻟﮕﻮرﯾﺘﻢﻫﺎي ﺗﺎﺑﻊ اﻟﮕﻮﻫﺎي ﺳﺎﺧﺘﺎري در ﺗﻮزﯾﻊ ﭘﯿﮑﺴﻞﻫﺎ، دﻗﺖ ﺑﺎﻻﺗﺮي را اراﯾﻪ دادﻧﺪ. ﺑﻪﻃﻮر ﮐﻠﯽ ﻧﺘﺎﯾﺞ ﺑﯿﺎﻧﮕﺮ ﭘﺘﺎﻧﺴﯿﻞ ﺑﺎﻻي دادهﻫﺎي 1-GeoEye در ﺗﻬﯿﻪ ﻧﻘﺸﻪﻫﺎي ﻣﺮاﺣﻞ ﺗﺤﻮﻟﯽ ﺟﻨﮕﻞ ﺑﻮده اﺳﺖ.

Today, remote sensing with high spatial and spectral resolutions along with the use of a variety of plant indices has provided the possibility of more accurate analysis and classification of satellite data in the identification and structural resolution of natural phenomena. Accordingly obtain information from the structure of forests using satellite data to help the management of forest renewable resources is of interest to managers and researchers. In this research, maps of the development stages of natural forests in the north of the country were prepared using GeoEye-1 satellite data and different classification algorithms in pixel-based, object-based, and model-based methods. The ultimate accuracy of the classification maps was calculated with the overall accuracy parameters and the Kappa coefficient. The results of this study showed that maximum accuracy (78%) in the object-based method was estimated on the basis of the segmentation of the NDVI and the maximum likelihood algorithm. Also, the algorithms that follow the structural patterns for the classification of the pixel distribution provided a higher accuracy. Overall, the results showed the high potential of GeoEye-1 data in mapping forest development stages.