As an important reserve land resource and ecological safety zone, saline-alkali land plays a key role in national food security during the global food crisis. Soil pH is an important indicator of saline-alkali land degradation, and its expansion seriously affects regional food security and sustainable ecosystems. However, limited mechanistic understanding hinders the formulation of land management strategies and affects the storage of soil organic matter. Therefore, identifying the spatial variation of soil properties affected by environmental factors is crucial to promote future agricultural practices. A total of 9,405 soil samples (0-100 cm) were collected from 527 sites in the Junggar Basin to measure pH levels. Geostatistical analyses were conducted to elucidate the spatial distribution patterns of pH. Environmental impact pathways were determined through the application of Linear regression (LR), Redundancy analysis (RDA), Random forest(RF), and Structural equation modeling(SEM). The results showed that (1) the mean soil pH in arid desert areas was 8.64, alkaline, with a tendency to increase and then decrease with soil depth (R²=0.82, P<0.05); sandy and saline deserts were significantly higher than gravelly deserts by a factor of 1.05, and small arbors deserts were significantly higher than shrub deserts by a factor of 1.03 (P<0.05). (2) The exponential model was the optimal model for soil pH in desert rangeland, which is mainly influenced by structural factors. The study area was completely alkaline, with heavy alkaline sporadically distributed. (3) Different statistical methods all show that monthly precipitation, soil organic carbon density, shrub vegetation cover, and biomass are significant influences on soil pH, and the structural equation model (SEM) clearly showed the direct and indirect pathways of environmental factors on soil pH, with monthly precipitation modifies soil pH changes by regulating vegetation traits and was a key driver of desert soil pH distribution in arid zones. The research results provide a data foundation for formulating effective land management strategies, which can help improve the sustainable development of grasslands in desert areas.