Trends, turning points, and driving forces of desertification in global arid land based on the segmental trend method and SHAP model

Xiaoyu Meng Shengyu Li Khaulenbek Akhmadi Panxing He Guanpeng Dong a Key Research Institute of Yellow River Civilization and Sustainable Development & Collaborative Innovation Center on Yellow River Civilization of Henan Province,Henan University,Zhengzhou,Chinab State Key Laboratory of Desert and Oasis Ecology,Key Laboratory of Ecological Safety and Sustainable Development in Arid Land,Xinjiang Institute of Ecology and Geography,Chinese Academy of Sciences,Urumqi,Chinac Institute of Geography & Geoecology,Mongolian Academy of Sciences,Ulaanbaatar,Mongoliad College of Physical Education,Henan Normal University,Xinxiang,Chinae Key Laboratory of Geospatial Technology for the Middle and Lower Yellow River Regions,Henan University,Kaifeng,China
DOI: https://doi.org/10.1080/15481603.2024.2367806
IF: 6.397
2024-06-19
GIScience & Remote Sensing
Abstract:Desertification is a form of land degradation observed in arid, semiarid, and dry subhumid ecosystems. Assessing the global trends and drivers of desertification in arid land is crucial for developing effective land restoration policies and mitigating desertification. We aimed to evaluate global segmental trends in desertification in arid land using Moderate Resolution Imaging Spectroradiometer images from 2000 to 2022. By constructing a robust MSAVI-Albedo Desertification Distance Index (DDI), we assessed the segmented development characteristics of desertification on Google Earth Engine. Additionally, we employed the SHapley Additive exPlanations (SHAP) model and machine learning methods to analyze the individual and interactive driving mechanisms of desertification. The results indicated an overall reduction in desertification, with approximately 23.21 million km 2 (39% of the global arid region area) exhibiting negative trends in DDI. Approximately 31% of the land area showed a DDI of −0.002/y. Precipitation was consistently the primary factor influencing desertification, with an average SHAP value of 11.42. Secondary influencing factors included potential evapotranspiration, soil moisture, and vapor pressure differences. Notably, the coupling between precipitation and soil moisture exhibited the most significant impact on the desertification process, with SHAP coupling values of approximately 3.28 and 5.06 before and after the DDI turning point, respectively. These findings provide new insights into desertification on a global scale and offer valuable scientific support for promoting effective prevention and control measures.
geography, physical,remote sensing
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the trend, turning point and driving factors of land desertification in arid regions around the world. Specifically, the author aims to evaluate the dynamic changes of desertification in arid regions around the world from 2000 to 2022 by using the piecewise trend method and SHAP model, reveal its spatio - temporal evolution characteristics and internal driving mechanisms, and provide support for achieving the zero - growth land degradation target in the sustainable development goals. ### Main research questions: 1. **Systematically monitor the dynamic changes of desertification in arid regions around the world**: Use the Google Earth Engine platform to systematically monitor the arid regions around the world from 2000 to 2022. 2. **Clarify the long - term change characteristics of desertification in arid regions**: Identify the piecewise characteristics and laws of desertification at the pixel scale. 3. **Reveal the conversion mechanism of desertification in arid regions at the pixel scale**: Analyze the internal mechanism of desertification changes in arid regions from 2000 to 2022. ### Research background: - **Definition of desertification**: Desertification refers to the land degradation process that occurs in arid, semi - arid and dry semi - humid ecosystems. - **Impact**: Desertification has a profound impact on the sustainable development of human society and is one of the most urgent social, economic and ecological problems currently faced by society. - **Global distribution**: Desertification affects about one - fifth of the global population, especially in arid regions, where its impact is particularly significant. - **Insufficiency of existing research**: Although there are some regional studies, the global - scale desertification assessment is still relatively limited, especially in the use of remote sensing image data. ### Methods and techniques: 1. **Remote sensing data**: Use MODIS satellite products, including NBAR and Albedo data, to monitor desertification in arid regions around the world. 2. **Environmental factors**: Consider meteorological factors (precipitation, temperature, evapotranspiration potential, etc.), topographic factors (slope, aspect, elevation) and socio - economic factors (livestock quantity, carbon dioxide concentration). 3. **Construction of desertification index**: Construct the Desertification Distance Index (DDI) through MSAVI and Albedo to evaluate the degree of desertification. 4. **Piecewise trend analysis**: Use the binary linear regression method to detect the change points of the desertification trend, and use the Sen slope analysis method and Mann - Kendall test method for significance testing. 5. **Machine learning methods**: Use four machine learning models, namely XGBoost, LightGBM, Random Forest and Multi - Layer Perceptron, to establish a non - linear relationship model between environmental factors and the DDI change rate. 6. **Model interpretation**: Use the SHAP model to interpret the prediction results of the machine learning model and reveal the specific contributions of each feature to the prediction results. ### Main findings: - **Overall trend**: The overall degree of global desertification shows a downward trend, and about 23.21 million square kilometers (accounting for 39% of the area of arid regions around the world) show a negative trend. - **Main driving factors**: Precipitation is the most important influencing factor, followed by potential evapotranspiration, soil moisture and vapor pressure deficit. - **Interaction**: The coupling of precipitation and soil moisture has the greatest impact on the desertification process. ### Conclusion: This study provides new insights into the spatio - temporal evolution characteristics and driving mechanisms of desertification in arid regions around the world, and provides scientific support for formulating effective desertification prevention and control measures.