Monitoring Spatiotemporal Dynamics of Soil Salinization Across the Qingtongxia Irrigation District Using Multi-Temporal Landsat Tm/Oli Data and Multi-Parameter Optimization

18 Pages Posted: 24 Feb 2025

See all articles by AWei SONG

AWei SONG

affiliation not provided to SSRN

Yuxing Liu

Yangtze University

Qiancheng Gao

Yangzhou University

Xiangping Wang

affiliation not provided to SSRN

Wenping Xie

Chinese Academy of Sciences (CAS)

Junhua Zhang

Ningxia University

Rongjiang Yao

affiliation not provided to SSRN

Abstract

The Qingtongxia Irrigation District, situated on the alluvial plain of the Yellow River in the Ningxia Hui Autonomous Region, faces severe soil salinization caused by both natural and anthropogenic factors. This issue reduces crop yields, exacerbates ecological and land degradation, and severely limits sustainable development of agriculture in the region. To address this challenge, this study proposed a precise soil salinization retrieval method based on multi-information fusion, enabling quantitative analysis of the spatiotemporal evolution and patterns of soil salinization.This study integrated remote sensing imagery with ground sampling data to develop several soil salinity prediction models including Multiple Linear Regression (MLR), Stepwise Regression (SR), Partial Least Squares Regression (PLSR), Support Vector Regression (SVR), and Random Forest Regression (RFR). The results indicated that the RFR model outperformed the other models in terms of salinity prediction accuracy. To further improve precision, certain data smoothing techniques, such as Moving Average (MA), Savitzky-Golay (SG), Median Filtering (MF), and Gaussian Filtering (GF), were applied to compare the model performance under different treatments. The findings revealed that the RFR model combined with GF smoothing achieved the highest predictive accuracy, making it well suited for precise surface salinity retrieval in the Qingtongxia Irrigation District. Using an optimized model, this study estimated the spatial distribution of soil salinity from 1989 to 2023. The results demonstrated a general decline in soil salinity across most areas, with notable reductions along the central northeast-southwest axis and in the western and eastern regions, while a few localized areas in the north exhibited increased salinity. This study provides a scientific foundation for predicting, preventing, and managing soil salinization in Qingtongxia Irrigation District, contributing to the sustainable application of soil resources, ensuring food security, and supporting sustainable agricultural development.

Keywords: Qingtongxia Irrigation District, Soil Salinization, Remote Sensing Retrieval, Spatiotemporal Distribution

Suggested Citation

SONG, AWei and Liu, Yuxing and Gao, Qiancheng and Wang, Xiangping and Xie, Wenping and Zhang, Junhua and Yao, Rongjiang, Monitoring Spatiotemporal Dynamics of Soil Salinization Across the Qingtongxia Irrigation District Using Multi-Temporal Landsat Tm/Oli Data and Multi-Parameter Optimization. Available at SSRN: https://ssrn.com/abstract=5152372 or http://dx.doi.org/10.2139/ssrn.5152372

AWei SONG (Contact Author)

affiliation not provided to SSRN ( email )

Yuxing Liu

Yangtze University ( email )

China

Qiancheng Gao

Yangzhou University ( email )

88 Daxue Road (South)
Yangzhou
Jiangsu, 225009
China

Xiangping Wang

affiliation not provided to SSRN ( email )

Wenping Xie

Chinese Academy of Sciences (CAS) ( email )

Chinese Academy of Sciences
Beijing, 100190
China

Junhua Zhang

Ningxia University ( email )

489 Helanshan West Rd
Xixia
Yinchuan
China

Rongjiang Yao

affiliation not provided to SSRN ( email )

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
19
Abstract Views
121
PlumX Metrics