A key challenge in precision agriculture is acquiring reliable spatial soil information under varying sampling densities, from sparse surveys to intensive monitoring. The individual predictive soil mapping (iPSM) method performs well in data-scarce conditions but neglects spatial proximity, limiting its predictive accuracy where spatial autocorrelation exists. To overcome this, we developed an enhanced framework, iPSM-Spatial Distance (iPSM-SD), which systematically integrates spatial proximity through multiplicative (MUL) and additive (ADD) strategies. The framework was validated using two contrasting cases: sparse soil organic carbon density data from Yunnan Province (n = 118) and dense soil organic matter data from Bayi Farm (n = 2511). Results show that the additive model (iPSM-ADD) significantly outperformed the original iPSM and benchmark models, including random forest, regression kriging, geographically weighted regression, and multiple linear regression, under sufficient sampling, achieving an R² of 0.86 and reducing RMSE by 46.6% at Bayi Farm. It also maintained robust accuracy under sparse sampling conditions. The iPSM-SD framework thus provides a unified and adaptive tool for digital soil mapping across a wide range of data availability, supporting scalable soil management decisions from regional assessment to field-scale variable-rate applications in precision agriculture.

Spatial prediction requires determining how sampled observations support estimates at unsampled locations. Existing methods commonly represent these relationships in geographic space, environmental feature space, or through a global statistical relationship between predictors and the target property. We propose a dual-space comprehensive similarity (CS) framework that combines target-oriented environmental similarity (ES) with geographic spatial proximity (SP). ES uses partial least squares screening and principal component transformation, whereas SP derives from scaled geographic distance. Their product gives high influence only to observations supported in both spaces. We embedded CS in individual predictive soil mapping (iPSM), creating iPSM-CS, and evaluated it using 547 soil organic carbon density observations from the Tibetan Plateau. Repeated experiments were conducted using training datasets of 100, 200, and 400 samples, with an independent 27-sample validation set and a 20-sample test set. The 500-sample scenario was evaluated separately using fixed neighbourhood sizes determined before final testing. Across the three repeated sample-size scenarios, iPSM-CS generally produced higher correlation coefficients (r) and lower root mean square error (RMSE) and mean absolute error (MAE) values than iPSM-ES and the benchmark models. In the 500-sample evaluation, iPSM-CS achieved r = 0.79, RMSE = 2.44 kg·m⁻², and MAE = 1.93 kg·m⁻², compared with r = 0.55, RMSE = 3.34 kg·m⁻², and MAE = 2.52 kg·m⁻² for iPSM-ES. Top-1 CS and Top-1 ES were significantly negatively associated with relative prediction error in the independent test set, while CS provided a clear stratification of errors across the predefined similarity intervals. The results support the use of dual-space similarity for spatial prediction, although its transferability requires evaluation across additional regions, attributes, and spatial validation designs.

Climate extremes in the Jinsha River Basin have changed during a period of regional warming and cascade hydropower development, but the climatic contribution of reservoir impoundment remains uncertain. We analyzed 20 indices derived from daily observations at 15 meteorological stations during 1980–2022. Regional and station-level trends were assessed, and reservoir associations were explored using matched before–after–control–impact (BACI) contrasts, distance–trend analyses. The station aggregates showed a consistent warming pattern: summer days (SU) and warm-day frequency (TX90p) increased by 0.50 and 0.72 reference-period standard deviations per decade, respectively, whereas frost days (FDs) and cold-night frequency (TN10p) decreased by 0.24 and 0.33. The five compound dry–hot indices increased by 0.24–0.43 standard deviations per decade, while the joint warm–dry severity index (SCEI) declined by 0.34, indicating more severe monthly warm–dry conditions. Precipitation trends were less spatially coherent. Autocorrelation-aware tests retained the full-network trends, although station-level support for maximum event duration weakened. BACI contrasts were mixed and uncertain, and the full-period distance gradient was absent in the common pre-impoundment period. These analyses describe the sampled station network but do not establish a basin-wide causal effect of reservoir impoundment.