استفاده از روش طيف سنجي مرئي-مادون قرمز نزديك در مدل سازي شوري خاك اراضي مستعد توليد ريزگرد استان خوزستان

Modeling Soil Salinity in Khuzestan Lands Susceptible for Dust Production Using Visible-Near Infrared Spectroscopic Method

ﺳﻄﺢ وﺳﯿﻌﯽ از اراﺿﯽ ﺷﻮر و ﻧﯿﻤﻪ ﺷﻮر اﺳﺘﺎن ﺧﻮزﺳﺘﺎن ﺑﻪ ﻋﻠﺖ ﻋﺪم ﭘﻮﺷﺶ ﺳﻄﺤﯽ و ﻣﻘﺎوﻣﺖ ﮐﻢ ﺧﺎك در ﺑﺮاﺑﺮ ﺑﺎد ﻓﺮﺳﺎﯾﻨﺪه ﺑﻪ ﮐﺎﻧﻮنﻫﺎي ﻣﺴﺘﻌﺪ ﺗﻮﻟﯿﺪ رﯾﺰﮔﺮد ﺗﺒﺪﯾﻞﺷﺪهاﻧﺪ. ﻫﺪف از اﯾﻦ ﭘﮋوﻫﺶ ﻣﺪلﺳﺎزي ﺷﻮري ﺧﺎك ﻣﻨﺎﻃﻖ ﺣﺴﺎس ﺑﻪ ﺗﻮﻟﯿﺪ رﯾﺰﮔﺮد اﺳﺘﺎن ﺧﻮزﺳﺘﺎن ﺑﺎ روش ﻃﯿﻒﺳﻨﺠﯽ اﻣﻮاج ﻣﺮﺋﯽ و ﻣﺎدونﻗﺮﻣﺰ ﻧﺰدﯾﮏ )2500-350 ﻧﺎﻧﻮﻣﺘﺮ( ﺑﻮد. از ﻣﺪلﻫﺎي ﭼﻨﺪ ﻣﺘﻐﯿﺮه رﮔﺮﺳﯿﻮن ﺣﺪاﻗﻞ ﻣﺮﺑﻌﺎت ﺟﺰﺋﯽ، ﺷﺒﮑﻪ ﻋﺼﺒﯽ ﻣﺼﻨﻮﻋﯽ و ﻣﺪل ﺟﻨﮕﻞ ﺗﺼﺎدﻓﯽ ﺑﺮاي ﻣﺪلﺳﺎزي ﺷﻮري ﺧﺎك ﺑﻪ ﮐﺎر ﮔﺮﻓﺘﻪ ﺷﺪ. ﻃﯿﻒ ﺑﺎزﺗﺎﺑﯽ ﺧﺎك ﺑﺎ دﺳﺘﮕﺎه ﻃﯿﻒﺳﻨﺞ زﻣﯿﻨﯽ )FieldSpec( ﺗﻌﯿﯿﻦ ﺷﺪ. ﻫﻤﭽﻨﯿﻦ روشﻫﺎي ﭘﯿﺶﭘﺮدازش ﻓﯿﻠﺘﺮ ﺳﺎوﯾﺘﺰي ﮔﻮﻻي، ﻣﺸﺘﻖ اول ﺑﻪ ﻫﻤﺮاه ﻓﯿﻠﺘﺮ ﺳﺎوﯾﺘﺰي ﮔﻮﻻي )FD-SG(، ﻣﺸﺘﻖ دوم ﺑﻪ ﻫﻤﺮاه ﻓﯿﻠﺘﺮ ﺳﺎوﯾﺘﺰي ﮔﻮﻻي )SD-SG(، روش ﻧﺮﻣﺎلﺳﺎزي اﺳﺘﺎﻧﺪارد )SNV( و روش ﺣﺬف ﭘﯿﻮﺳﺘﺎر )CR(، ﺟﻬﺖ ﺣﺬف ﻧﻮﯾﺰ و اﻓﺰاﯾﺶ دﻗﺖ ﻣﺪلﻫﺎي ﭼﻨﺪ ﻣﺘﻐﯿﺮه ﻣﻮرد اﺳﺘﻔﺎده ﻗﺮار ﮔﺮﻓﺖ. ﻧﺘﺎﯾﺞ ﻧﺸﺎن داد ﮐﻪ ﻣﺪل ﺗﺮﮐﯿﺒﯽ ﺣﺪاﻗﻞ ﻣﺮﺑﻌﺎت ﺟﺰﺋﯽ- ﺷﺒﮑﻪ ﻋﺼﺒﯽ ﻣﺼﻨﻮﻋﯽ ﺑﺎ ﻣﻌﯿﺎرﻫﺎي ارزﯾﺎﺑﯽ )2/65 - 3/40 =RPDcal( در ﺑﺮآورد ﺷﻮري ﺧﺎك دﻗﺖ ﻣﻨﺎﺳﺒﯽ دارد. در ﻣﻘﺎﺑﻞ ﻣﺪل ﺗﺮﮐﯿﺒﯽ ﺣﺪاﻗﻞ ﻣﺮﺑﻌﺎت ﺟﻨﮕﻞ ﺗﺼﺎدﻓﯽ ﻧﯿﺰ ﮐﻤﺘﺮﯾﻦ دﻗﺖ )0/1-85/98= RPDcal( را ﻧﺸﺎن داد. ﭘﯿﺶﭘﺮدازش ﻃﯿﻒ اﺻﻠﯽ در دو ﻣﺪل ﺷﺒﮑﻪ ﻋﺼﺒﯽ و رﮔﺮﺳﯿﻮن ﺣﺪاﻗﻞ ﻣﺮﺑﻌﺎت ﺟﺰﺋﯽ ﺳﺒﺐ اﻓﺰاﯾﺶ ﻧﺴﺒﯽ دﻗﺖ ﻣﺪل ﺷﺪ درﺣﺎﻟﯽﮐﻪ در ﻣﺪل ﺟﻨﮕﻞ ﺗﺼﺎدﻓﯽ ﭘﯿﺶﭘﺮدازش ﺳﺒﺐ ﮐﺎﻫﺶ دﻗﺖ ﺑﺮآورد ﻣﺪل، ﻧﺴﺒﺖ ﺑﻪ ﻃﯿﻒ اﺻﻠﯽ ﺷﺪ. ﻣﺤﺪوده 2300 ،2000 ،1800،1900 و 1500 ﻧﺎﻧﻮﻣﺘﺮ ﺑﻪ ﻋﻨﻮان ﻃﻮل ﻣﻮج ﮐﻠﯿﺪي ﻣﺘﺄﺛﺮ از ﺷﻮري ﺧﺎك ﺷﻨﺎﺳﺎﯾﯽ ﺷﺪ. از ﻃﻮل ﻣﻮجﻫﺎي ﮐﻠﯿﺪي ﺑﻪدﺳﺖ آﻣﺪه، ﻣﯽﺗﻮان در ﻣﻄﺎﻟﻌﺎت دورﺳﻨﺠﯽ و ﻬﯿﻪ ﻧﻘﺸﻪ ﺷﻮري ﻣﻨﺎﻃﻖ ﺣﺴﺎس ﺑﻪ ﺗﻮﻟﯿﺪ ﮔﺮد و ﻏﺒﺎر اﺳﺘﺎن ﺧﻮزﺳﺘﺎن اﺳﺘﻔﺎده ﮐﺮد

A broad area of saline and semi-saline lands of Khuzestan province have changed into centers susceptible to dust production due to eroded wind and lack of surface coating and low soil resistance. The objective of this study was to model the soil salinity of sensitive areas to dust production in Khuzestan Provenience usin spectrometry method of visible and near-infrared wavelengths (2500-350 nm). The least square multivariate regression model, artificial neural network and random forest model were used to estimate soil salinity. The main soil spectrum was determined using the FieldSpect machine. Also, preprocessing methods including Savitzky-Golay filter, the first derivative with the Savitzky-Golay filter (FD-SG), the second derivative with the Savitzky-Golay filter (SD-SG), the standard normalization method (SNV), and the continuum remove method (CR) were used to eliminate the noise and to increase the accuracy of the multivariate model. The results showed that the combined model partial least squares-artificial neural network model with assessment criteria (RPDcal = 3.40-2.65) has high accuracy for salinity estimation. In contrast, the combined model of least squares - random forest showed the lowest accuracy (RPDcal = 0.85-1.98). Preprocess of the main spectrum in two models (neural network and partial least squares regression) increased the relative accuracy of the model; while in the random forest model, preprocess reduced the accuracy of the model compared to the main spectrum. The ranges of 1800, 1900, 2000, 2300 and 1500 nm were recognized as "the key wavelengths" impressed by soil salinity. The key wavelengths can be used in remote sensing studies and mapping of soil salinity in areas sensitive to dust production in Khuzestan province.