The Impact of Climate Change on Soil Health: Analyzing Microbiome Response to Warming Using Machine Learning Models

Articles in Press

Document Type : Research Paper

Authors

1 Department of Soil Science, Faculty of Agriculture, Ferdowsi University of Mashhad Mashhad Iran

2 Assistant professor, Soil Reclamation and Sustainable Land Management Department,, Soil and Water Research Institute of Iran, Karaj, Iran

3 Computer Department, Jihad Daneshgahi, of Khorasan Razavi, Mashhad Iran

Abstract

Climate change has extensive effects on soil functioning, particularly through the temperature sensitivity of microbial respiration, which is measured by the temperature sensitivity coefficient (Q10). This study aimed to identify the factors influencing this coefficient and to evaluate the performance of machine learning algorithms. Data from 332 soil samples collected across 29 countries were used, and the temperature sensitivity coefficient was classified into three levels: low, medium, and high. Six algorithms—including Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, Gradient Boosting, and Multilayer Perceptron—were applied. Model performance was compared using accuracy, the area under the Receiver Operating Characteristic curve (ROC), and the area under the Precision–Recall curve (PRC). The results indicated that Random Forest had the highest performance, achieving an accuracy of 0.1170 ± 0, ROC of 0.8150 ± 0.080, and PRC of 0.7530 ± 0.096, outperforming the other models. The SHAP method was employed for interpreting the results as an explainable AI approach. This analysis revealed that glucose-induced respiration, fungal richness, and soil salinity were the most influential factors affecting the temperature sensitivity coefficient. These findings provide a clear picture of the combined role of soil and microbial features and demonstrate the potential of machine learning algorithms and explainable methods in analyzing soil responses to global warming.

Keywords

Main Subjects

Iranian Journal of Soil and Water Research

Articles in Press, Accepted Manuscript
Available Online from 11 October 2025

Files

Share

How to cite

Statistics