University of Tabriz Water and Soil Science 2008-5133 29 2 2019 06 22 Mapping Soil Salinity Using Remote Sensing Technology (Case Study: South Lands of West Azarbaijan province) Mapping Soil Salinity Using Remote Sensing Technology (Case Study: South Lands of West Azarbaijan province) 115 128 9275 FA KH Salimi M.Sc. Graduate of Agroecology, Mahabad Branch, Islamic Azad University, Mahabad, Iran N Ahmadi Sani Assist. Prof., Faculty of Agriculture and Natural Resources, Mahabad Branch, Islamic Azad University, Mahabad, Iran N Jalilnejad Ph.D. Student of Plant Breeding, Science and Research Branch, Islamic Azad University, Tehran, Iran Journal Article 2019 09 16 Soil salinity is one of the most important limiting factors in crop production and more than 50% of the world irrigated lands have different levels of salinity. Also, one of the most important agricultural problems in Iran is soil salinty where West Azerbaijan province is no exception from this problem. The aim of present study was investigating the possibility of surface soil salinity mapping in an area of 68000 ha in the south of West Azarbaijan province using remote sensing. A systematic random sampling method was used to prepare a ground truth map of surface soil salinity (including 147 soil samples from a depth of 0-15 cm). In this research, IRS-P6 satellite data (multispectral and panchromatic bands) were used. Geometric correction of images was done using digital elevation model and road and river maps. Various indices were generated, using ratioing and principal component analysis methods, and were used in the classifications. After selection of suitable training samples, classification was performed with different number of classes by a supervised method and different algorithms. The results of the accuracy assessment showed that the highest overall accuracy percentage and Kappa coefficient (87.96% and 0.77) were related to the classification of the two classes with 4-band combination and the maximum likelihood algorithm. Also, the results revealed high potential of remote sensing and IRS-P6 data for surface soil salinity mapping on a local scale. Soil salinity is one of the most important limiting factors in crop production and more than 50% of the world irrigated lands have different levels of salinity. Also, one of the most important agricultural problems in Iran is soil salinty where West Azerbaijan province is no exception from this problem. The aim of present study was investigating the possibility of surface soil salinity mapping in an area of 68000 ha in the south of West Azarbaijan province using remote sensing. A systematic random sampling method was used to prepare a ground truth map of surface soil salinity (including 147 soil samples from a depth of 0-15 cm). In this research, IRS-P6 satellite data (multispectral and panchromatic bands) were used. Geometric correction of images was done using digital elevation model and road and river maps. Various indices were generated, using ratioing and principal component analysis methods, and were used in the classifications. After selection of suitable training samples, classification was performed with different number of classes by a supervised method and different algorithms. The results of the accuracy assessment showed that the highest overall accuracy percentage and Kappa coefficient (87.96% and 0.77) were related to the classification of the two classes with 4-band combination and the maximum likelihood algorithm. Also, the results revealed high potential of remote sensing and IRS-P6 data for surface soil salinity mapping on a local scale.

https://water-soil.tabrizu.ac.ir/article_9275_ec829f89e3e56e3dd0ae455a3a6ff72d.pdf