Dimension Reduction of remote sensing data to estimate soil organic carbon

Document Type : Research Article

Authors

Department of Geodesy and Surveying Engineering, Tafresh University, Tafresh, Iran

Abstract

In this research, the relation between soil spectral reflectance using Landsat 8 satellite data as well as SRTM Elevation data and soil organic carbon has been investigated. In the proposed method, spectral reflection of data in the main bands of Landsat 8 satellite is investigated and processed. In addition to the main bands, vegetation and lighting indices and topographic features have been studied. In this study, we try to select the effective indicators in increasing the accuracy of soil organic carbon modeling. For this purpose, in the first step, the carbon model was performed using linear regression, support vector machine regression and neural network methods. In order to implement the proposed method, 100 soil samples in East Azerbaijan province have been used. According to RMSE and R2 statistical indices, which are the basis for evaluating the models, the neural network model was selected as the final model and with the values of RMSE = 0.0798, R2= 0.8741 and RRMSE=9.5683 is more accurate than the regression method. Due to the importance of dimensionality in order to increase accuracy and reduce the complexity of calculations, a genetic algorithm was proposed in this study. This efficient algorithm increases the accuracy of soil organic carbon modeling and eliminates additional indicators. After applying the genetic algorithm on the neural network model, we were able to achieve better accuracy and the values of the baseline statistical indices were changed to RMSE = 0.043, R2 = 0.9398 and RRMSE=5.1559.

Keywords

Main Subjects

Amirkabir Journal of Civil Engineering
Volume 56, Issue 5
2024
Pages 4-4

Files

Share

How to cite

Statistics