Search

Soil salinization has a significant impact on agricultural production and ecology. There is an urgent demand to establish an effective method that monitors the spatial and temporal distribution of soil salinity. In this study, a multi-indicator soil salinity monitoring model was proposed for monitoring soil salinity in Bachu County, Kashgar Region, Xinjiang, from 2002 to 2022. The model was established by combining multiple predictors (spectral, salinity, and composite indices and topographic factors) and the accuracy of the four models (Random Forest [RF], Partial Least Squares [PLS], Classification Regression Tree [CART], and Support Vector Machine [SVM]) was compared. The results reveal the high accuracy of the optimized prediction model, and the order of the accuracy is observed as RF > PLS > CART > SVM. The most accurate model, RF, exhibited an R2 of 0.723, a root mean square error (RMSE) of 2.604 g·kg−1, and a mean absolute error (MAE) of 1.95 g·kg−1 at a 0–20 cm depth. At a 20–40 cm depth, RF had an R2 value of 0.64, an RMSE of 3.62 g·kg−1, and an MAE of 2.728 g·kg−1. Spatial changes in soil salinity were observed throughout the study period, particularly increased salinization from 2002 to 2012 in the agricultural and mountainous areas within the central and western regions of the country. However, salinization declined from 2012 to 2022, with a decreasing trend in salinity observed in the top 0–20 cm of soil, followed by an increasing trend in salinity at a 20–40 cm depth. The proposed method can effectively extract large-scale soil salinity and provide a practical basis for simplifying the remote sensing monitoring and management of soil salinity. This study also provides constructive suggestions for the protection of agricultural areas and farmlands.

To obtain timely, accurate, and reliable information on wheat yield dynamics. The UAV DJI Wizard 4-multispectral version was utilized to acquire multispectral images of winter wheat during the tasseling, grouting, and ripening periods, and to manually acquire ground yield data. Sixteen vegetation indices were screened by correlation analysis, and eight textural features were extracted from five single bands in three fertility periods. Subsequently, models for estimating winter wheat yield were developed utilizing multiple linear regression (MLR), partial least squares (PLS), BP neural network (BPNN), and random forest regression (RF), respectively. (1) The results indicated a consistent correlation between the two variable types and yield across various fertility periods. This correlation consistently followed a sequence: heading period > filling period > mature stage. (2) The model’s accuracy improves significantly when incorporating both texture features and vegetation indices for estimation, surpassing the accuracy achieved through the estimation of a single variable type. (3) Among the various models considered, the partial least squares (PLS) model integrating texture features and vegetation indices exhibited the highest accuracy in estimating winter wheat yield. It achieved a coefficient of determination (R2) of 0.852, a root mean square error (RMSE) of 74.469 kg·hm−2, and a normalized root mean square error (NRMSE) of 7.41%. This study validates the significance of utilizing image texture features along with vegetation indices to enhance the accuracy of models estimating winter wheat yield. It demonstrates that UAV multispectral images can effectively establish a yield estimation model. Combining vegetation indices and texture features results in a more accurate and predictive model compared to using a single index.