Alaa Khallouf (1)*, Sameer Shamsham(1) and Younis Idriss(2)
(1). Department of Soil and Land reclamation- Faculty of Agriculture- Al-Baath University, Homas, Syria.
(2).Researcher, General Organization of Remote Sensing (GORS)
(*corresponding: Eng. Alaa Khallouf Email: alaakhallouf@gmail.com).
Received: 21/11/2020 Accepted: 10/05/2021
Abstract
The research aims to study of using remote sensing data, including Landsat 8 OLI and DEM (Digital Elevation Model), to predict available micronutrients (Fe, Mn, Zn, Cu, B) in Misyaf region- Hamah governorate. 56 soil samples were taken in August 2020. The Landsat 8 OLI image was processed and some vegetation indices (NDVI, SAVI, GSAVI) were derived. A Stepwise regression was set based all soil samples in adopting micronutrients- predictive models, while cross-validation method was used to evaluate the performance and the accuracy of the model based on the coefficient of determination(R2), the adjusted coefficient of determination (R2adj), mean absolute error (MAE), and root mean square error (RMSE). The results revealed that the topographic factor, particularly elevation, was the most significant factor in predicting Fe, Mn, and Cu with a positive relationship, while Zn had a different behavior, since EVI index was the most important index in predicting of boron. The performance validating the of the models show, the coefficient of determination and the adjusted coefficient of determination of the Fe model were 81.5 and 80.8%, respectively, with RMSE = 2.183 mg/kg, and MAE = 0.744. For manganese, it was found that the model with R2 = 66.1%, while the adjusted coefficient of determination = 64.7%, for calibration of the model, it was found that MAE = 0.562 and RMSE = 2.091 mg / kg.
Key words: Basalt-developed Soils, Landsat8 OLI, Micronutrients, Multiple Linear Regression, Misyaf.
Full paper in Arabic: pdf