پديد آورندگان :
ﯾﻮﺳﻔﯽ، ﺣﺴﯿﻦ داﻧﺸﮕﺎه ﺗﻬﺮان - داﻧﺸﮑﺪه ﻋﻠﻮم و ﻓﻨﻮن ﻧﻮﯾﻦ، ايران , ﯾﻮﻧﺴﯽ، ﺣﺠﺖ ﷲ داﻧﺸﮕﺎه ﻟﺮﺳﺘﺎن - داﻧﺸﮑﺪه ﮐﺸﺎورزي و ﻣﻨﺎﺑﻊ ﻃﺒﯿﻌﯽ - ﮔﺮوه ﻣﻬﻨﺪﺳﯽ آب، ايران , داودي ﻣﻘﺪم، داود داﻧﺸﮕﺎه ﻟﺮﺳﺘﺎن - داﻧﺸﮑﺪه ﮐﺸﺎورزي و ﻣﻨﺎﺑﻊ ﻃﺒﯿﻌﯽ، ايران , ارﺷﯿﺎ، آزاده داﻧﺸﮕﺎه ﻟﺮﺳﺘﺎن - داﻧﺸﮑﺪه ﮐﺸﺎورزي و ﻣﻨﺎﺑﻊ ﻃﺒﯿﻌﯽ، ايران , ﺷﻤﺴﯽ، زﻫﺮا داﻧﺸﮕﺎه ﻟﺮﺳﺘﺎن - داﻧﺸﮑﺪه ﮐﺸﺎورزي و ﻣﻨﺎﺑﻊ ﻃﺒﯿﻌﯽ، ايران
كليدواژه :
ﺷﺎﺧﺺ رﻃﻮﺑﺖ ﺗﻮﭘﻮﮔﺮاﻓﯽ , AUC , ﻧﻘﺸﻪ ﺳﯿﻞ , ROC , TSS
چكيده فارسي :
ﺳﯿﻞ ﭘﺪﯾﺪهاي اﺳﺖ ﮐﻪ ﻣﻮﺟﺐ آﺳﯿﺐﻫﺎي زﯾﺴﺖﻣﺤﯿﻄﯽ و اﻗﺘﺼﺎدي-اﺟﺘﻤﺎﻋﯽ ﺑﺴﯿﺎري ﻣﯽﺷﻮد. ﻫﺪف از اﯾﻦ ﭘﮋوﻫﺶ، ارزﯾﺎﺑﯽ ﮐﺎراﯾﯽ ﻣﺪلﻫﺎي ﯾﺎدﮔﯿﺮي ﻣﺎﺷﯿﻦ GLM ،CART و GAM در ﺷﻨﺎﺳﺎﯾﯽ ﻣﻨﺎﻃﻖ ﺣﺴﺎس ﺑﻪ ﺧﻄﺮ ﺳﯿﻼب در ﺣﻮﺿﻪ ﮐﺸﮑﺎن اﺳﺖ. اﺳﺘﺎن ﻟﺮﺳﺘﺎن و ﺑﻪوﯾﮋه ﺣﻮﺿﻪ ﮐﺸﮑﺎن ﺷﺎﻣﻞ: ﺳﻠﺴﻠﻪ، دﻟﻔﺎن، دوره، ﺧﺮمآﺑﺎد، ﭘﻠﺪﺧﺘﺮ و ﮐﻮﻫﺪﺷﺖ، ﺳﯿﻞﺧﯿﺰ اﺳﺖ و دﻓﻌﺎت ﺑﺴﯿﺎري دﭼﺎر ﺧﺴﺎرات ﻧﺎﺷﯽ از ﺳﯿﻞ ﺷﺪه اﺳﺖ و در ﻓﺮوردﯾﻦ 1398، ﺑﺰرگﺗﺮﯾﻦ ﺳﯿﻞ 200 ﺳﺎل اﺧﯿﺮ را ﺗﺠﺮﺑﻪ ﮐﺮده اﺳﺖ؛ در ﻫﻤﯿﻦ راﺳﺘﺎ از ﻋﻮاﻣﻞ ﻣﺨﺘﻠﻒ ﺷﺎﻣﻞ: ارﺗﻔﺎع، ﺟﻬﺖ ﺷﯿﺐ، اﻧﺤﻨﺎي زﻣﯿﻦ، درﺻﺪ ﺷﯿﺐ، ﻓﺎﺻﻠﻪ از رودﺧﺎﻧﻪ، ﺗﺮاﮐﻢ زﻫﮑﺸﯽ، ﺧﺎك، ﺳﻨﮓﺷﻨﺎﺳﯽ، ﮐﺎرﺑﺮي اراﺿﯽ و ﺷﺎﺧﺺ رﻃﻮﺑﺖ ﺗﻮﭘﻮﮔﺮاﻓﯽ اﺳﺘﻔﺎده ﺷﺪ. ﻧﻘﺸﻪ رﻗﻮﻣﯽ ﺗﻤﺎم ﻋﻮاﻣﻞ ﻧﺎمﺑﺮده در ﻧﺮماﻓﺰار ArcGIS10.5 و در ﻗﺎﻟﺐ ﭘﺎﯾﮕﺎه داده ﺗﻬﯿﻪ ﺷﺪ. ﻣﻮﻗﻌﯿﺖ 123 واﻗﻌﻪ ﺳﯿﻞ ﺛﺒﺖﺷﺪه در ﺳﺎلﻫﺎي اﺧﯿﺮ در اﯾﻦ ﺣﻮﺿﻪ، ﺟﻤﻊآوري و ﺑﻪﺻﻮرت ﺗﺼﺎدﻓﯽ در دو دﺳﺘﻪ آﻣﻮزش ﻣﺪل )86 واﻗﻌﻪ( و اﻋﺘﺒﺎرﺳﻨﺠﯽ ﻣﺪل )37 واﻗﻌﻪ( در ﻣﺪلﺳﺎزيﻫﺎ اﺳﺘﻔﺎده ﺷﺪ. ﺑﺎ اﺳﺘﻔﺎده از ﻣﺪلﻫﺎي ﯾﺎدﮔﯿﺮي ﻣﺎﺷﯿﻦ و ﻋﻮاﻣﻞ ﻣﺆﺛﺮ ﻣﺤﯿﻄﯽ، ﻧﻘﺸﻪﻫﺎي ﭘﯿﺶﺑﯿﻨﯽ ﭘﺘﺎﻧﺴﯿﻞ ﺳﯿﻞ ﺗﻬﯿﻪ ﺷﺪﻧﺪ و ﺳﭙﺲ ﺑﺎ اﺳﺘﻔﺎده از روشﻫﺎي ﻣﻨﺤﻨﯽ ﻣﺸﺨﺼﻪ AUC و ﺷﺎﺧﺺ TSS اﻋﺘﺒﺎرﺳﻨﺠﯽ ﺷﺪﻧﺪ. ﻧﺘﺎﯾﺞ ﺣﺎﺻﻞ از اﻋﺘﺒﺎرﺳﻨﺠﯽ ﻣﺪلﻫﺎ ﻧﺸﺎن داد ﮐﻪ ﻣﺪل ﯾﺎدﮔﯿﺮي ﻣﺎﺷﯿﻦ CART ﺑﺎ 0/91=AUC و ﺷﺎﺧﺺ 0/88=TSS دﻗﯿﻖﺗﺮﯾﻦ ﻣﺪل در ﭘﯿﺶﺑﯿﻨﯽ ﭘﺘﺎﻧﺴﯿﻞ ﺧﻄﺮ ﺳﯿﻞ ﺑﻮده و ﭘﺲ از آن ﻣﺪل GAM ﺑﺎ 0/87=AUC و ﺷﺎﺧﺺ 0/84=TSS و ﻣﺪل GLM ﺑﺎ 0/83=AUC و ﺷﺎﺧﺺ 0/88=TSS ﻗﺮار دارﻧﺪ. دﻗﺖ 0/91 ﻣﺪل CART ﻧﺸﺎندﻫﻨﺪه دﻗﺖ ﻋﺎﻟﯽ اﯾﻦ ﻣﺪل ﺑﺮاي ﺣﻮﺿﻪ ﮐﺸﮑﺎن اﺳﺖ. اﯾﻦ ﻣﺪل، ﻣﺴﺎﺣﺖ ﺑﯿﺸﺘﺮي از ﺣﻮﺿﻪ را ﺗﺤﺖ ﺷﺮاﯾﻂ ﭘﺘﺎﻧﺴﯿﻞ ﺑﺎﻻ و ﻣﺘﻮﺳﻂ ﺧﻄﺮ ﺳﯿﻞﮔﯿﺮي ﻧﺸﺎن ﻣﯽدﻫﺪ ﮐﻪ اﻏﻠﺐ ﻣﻨﺎﻃﻖ ﻏﺮﺑﯽ و ﻫﻤﭽﻨﯿﻦ ﻣﻨﺎﻃﻖ ﻣﺮﮐﺰي ﺣﻮﺿﻪ )ﮐﻮﻫﺪﺷﺖ، ﺧﺮمآﺑﺎد و ﭘﻠﺪﺧﺘﺮ( را ﺷﺎﻣﻞ ﻣﯽﺷﻮﻧﺪ ﮐﻪ دﻗﯿﻘﺎً ﺑﺨﺶﻫﺎﯾﯽ از ﻫﻤﯿﻦ ﻣﻨﺎﻃﻖ در ﺳﯿﻞ ﺑﺰرگ ﺳﺎل 98 ﻫﻢ زﯾﺮ آب رﻓﺘﻨﺪ و ﻻزم اﺳﺖ در اوﻟﻮﯾﺖ اول ﺑﺮﻧﺎﻣﻪرﯾﺰي و ﻣﺪﯾﺮﯾﺖ رﯾﺴﮏ ﺳﯿﻞ ﻗﺮار ﮔﯿﺮﻧﺪ.
چكيده لاتين :
Flood is a phenomenon that causes a lot of environmental and socio-economic damage. The purpose of this study is to evaluate the efficiency of CART, GLM and GAM machine learning models in identifying flood risk areas in the Kashkan basin. Lorestan province and especially Kashkan basin, including: Selseleh, Delfan, Doreh, Khorramabad, Poldakhtar and Kuhdasht, is flooded and has suffered flood damage many times and in April 2019, experienced the largest flood of the last 200 years. In this regard, various factors including: height, slope direction, land curvature, slope percentage, distance from the river, drainage density, soil, lithology, land use and topographic moisture index were used. The digital map of all the mentioned factors was prepared in ArcGIS10.5 software and in the form of a database. The location of 123 flood events recorded in recent years in this basin was collected and randomly used in two categories of model training (86 cases) and model validation (37 cases) in modeling. Using machine learning models and environmental factors, flood potential prediction maps were prepared and then validated using AUC characteristic curve methods and TSS index. The results of model validation showed that CART machine learning model with AUC = 0.91 and TTS = 0.88 index was the most accurate model in predicting flood risk potential, followed by GAM model with AUC = 0.87 and TSS index = 0.84 and GLM model with AUC = 0.83 and TSS index = 0.88. Accuracy 0.91 CART model indicates the excellent accuracy of this model for the Kashkan basin. This model shows a larger area of the basin under high potential and moderate flood risk conditions, which include most of the western areas as well as the central areas of the basin (Kuhdasht, Khorramabad and Poldakhtar), which are exactly parts The same areas were flooded in the great flood of 2019 and it is necessary to be in the first priority of flood risk planning and management in this basin.
