Estimation of soil salinity using satellite images and laboratory data Case area: Barkhar city, Isfahan province

Subject Areas : Applications in biodiversity conservation and management

Sayyad Asghari Saraskanroud ^{1 *} , ehsan tavakoli ² , batool zeynali ³ , Shiva Safari ⁴ , elham mollanouri ⁵ , Aboozar Sadeghi ⁶

1 - Department of Physical Geography, Faculty of Social Science, University of Mohaghegh Ardabili
2 - Department of Physical Geography, Faculty of Social Science, University of Mohaghegh Ardabili
3 - Department of Physical Geography, Faculty of Social Science, University of Mohaghegh Ardabili
4 - MSc. Student of Remote Sensing and GIS, University of Mohaghegh Ardabili
5 - Department of Physical Geography, Faculty of Social Science, University of Mohaghegh Ardabili
6 - Department of Physical Geography, Faculty of Social Science, University of Mohaghegh Ardabili

Received: 2023-02-12 Accepted : 2025-05-12 Published : 2025-12-27

Keywords: vegetation index, salinity index, brightness index, cokriging, SVM algorithm,

Abstract :

Soil salinization and its development in arid and semi-arid areas is one of the environmental hazards. The purpose of this research is to estimate soil salinity in a specific period using remote sensing in Barkhar city of Isfahan province. In this research, to estimate soil salinity, salinity indices (NDSI, SI1, SI, SI2, SI3), vegetation index (NDVI), brightness index (BI) and spectral transmission indices with Employing empirical relationships were calculated and extracted. Then, all the indicators were classified by SVM algorithm method and the accuracy of the results was done, and finally the best indicator in soil salinity estimation was determined. In this study, 34 soil samples were checked for normality with statistical tests in Spss software and interpolation method was used. The evaluation of the classification maps of spectral indices shows that the SI3 indices applied to the images of Landsat 8, Sentinel 2 of 2020 and Landsat 8 of 2014 are the most accurate indices with Kappa coefficient values of 0.94, 0.96 and 0.87 respectively. It is a spectrum for soil salinity estimation and also the salinity maps extracted from Sentinel 2 satellite images are of higher accuracy than the maps obtained from Landsat 8 images. Therefore, by comparing the results of the index thresholding method, it was found that the indices are consistent with the results of geostatistics and the indices have sufficient accuracy to estimate soil salinity. The investigation of soil salinity in the study area during the period of time shows that soil salinity has an increasing trend, so that about 36% of the land without salinity has become saline; The biggest change in the amount of salinity has been related to highly saline lands, so that about 1067 hectares have experienced severe salinity during the studied time period.