Potential of solar-induced chlorophyll fluorescence for monitoring long-term dynamics of soil salinity in Central Asia the Xinjiang Region China.

INTRODUCTION

Soil salinization in Central Asia and Xinjiang, China, poses serious threats to agriculture and ecosystems. Solar-induced chlorophyll fluorescence (SIF), which reflects plant photosynthetic status and stress, shows promise for monitoring salinity but remains underutilized in this region.

METHODS

This study integrated SIF-derived indices (SIFI) with soil salinity data to build a region-specific prediction model. Using a random forest algorithm, soil salinity was classified into five levels based on satellite data and ground references from 2000-2020. Model performance, seasonal sensitivity, and spatial variation were analyzed across Central Asian countries and Xinjiang.

RESULTS

SIF effectively detected salinization dynamics, with highest sensitivity in Kazakhstan and Xinjiang. April was identified as the most responsive month, with SIFI1 being the key indicator. The model achieved over 80% accuracy in typical regions and around 70% in atypical regions. Kazakhstan had the largest salt-affected area, followed by Turkmenistan and Xinjiang. Tajikistan showed high variability, while Xinjiang remained relatively stable. Most areas exhibited increasing salinity and expansion of saline lands.

DISCUSSION

These findings demonstrate the potential of SIF-based monitoring for large-scale salinity assessment. The integration of plant physiological signals with machine learning provides a valuable tool for early warning and sustainable land management in arid regions.