Search

This study addresses the critical issues of water scarcity and soil salinization impacting cotton production in South Xinjiang, China. It introduces an innovative framework for assessing regional cotton crop suitability by integrating ground-measured soil water and salt data with UAV multispectral and Sentinel-2A satellite imagery from the 2022 cotton growing season. An optimized set of vegetation indices was identified through multicollinearity analysis and full subset selection. Six advanced machine learning methods, including Random Forest (RF), were used alongside the ratio mean method to effectively upscale soil water and salt content models from the field to the regional level. A newly developed cotton suitability index was created to categorize soil water and salt conditions, resulting in detailed suitability maps for 2022 and 2023. Key findings include: (1) Model Performance: The RF model outperformed others in predicting soil water and salt content, with R² values ranging from 0.763 to 0.846 for soil moisture and 0.703–0.843 for soil salinity. It showed greater accuracy at 0–10 cm depth than 10–20 cm depth. (2) Imagery Correlation: A significant correlation was observed between UAV and Sentinel-2A imagery (R² = 0.498–0.745). Reflectivity corrections in Sentinel-2A data notably improved RF model inversion accuracy (R² gains of 0.114–0.384). (3) Suitability Analysis: The cotton suitability index maps for 2022 and 2023 indicated that most fields in Tumushuke (TMSK) were moderately suitable for cotton growth, although some areas were unsuitable. This highlights the need for additional irrigation and targeted soil water and salt management to meet cotton requirements and reduce salinity risks. Overall, this study enhances precision agriculture techniques for arid environments and provides valuable insights for managing soil salinity, supporting sustainable cotton production in challenging climates.

Landslide susceptibility prediction (LSP) is an important step for landslide hazard and risk assessment. Automated machine learning (AutoML) has the advantages of automatically features, models, and parameters selection. In this study, we proposed an AutoML-based global LSP framework at two spatial resolutions of 90 m and 1000 m, and achieved an area under the receiver operating characteristic above 0.96. The global prediction results were then validated using additional regional landslide inventories, including three countries, three provinces, and two prefecture-level datasets. Moreover, the global prediction results of 90 m are used to improve the performance of regional LSP. Specifically, the low-and very low-prone areas in the global prediction results were used as non-landslide samples for susceptibility modeling. Results demonstrated that the model achieved a better performance than original global prediction results. We believe that this study will be able to reliably promote the application of intelligent learning methods in global LSP.

This research is focused on the evaluation of the reliability of regional landslide susceptibility models obtained by exploiting inhomogeneous (for quality, resolution and/or triggering related type and intensity) collected inventories for calibration. At a large-scale glance, merging more inventories can result in well-performing models hiding potential strong predictive deficiencies. An example of the limits that such kinds of models can display is given by a landslide susceptibility study, which was carried out for a large sector of the coastal area of El Salvador, where an apparently well-performing regional model (AUC = 0.87) was obtained by regressing a dataset through multivariate adaptive regression splines (MARS), including five landslide inventories from volcanic areas (Ilopango and Coatepeque caldera; San Salvador, San Miguel, and San Vicente Volcanoes). A multiscale validation strategy was applied to verify its actual predictive skill on a local base, bringing to light the loss in the predictive power of the regional model, with a lowering of AUC (20% on average) and strong effects in terms of sensitivity and specificity.

Satellite-borne multispectral data are suitable for regional-scale grassland community classification owing to comprehensive coverage. However, the spectral similarity of different communities makes it challenging to distinguish them based on a single multispectral data. To address this issue, we proposed a support vector machine (SVM)–based method integrating multispectral data, two-band enhanced vegetation index (EVI2) time-series, and phenological features extracted from Chinese GaoFen (GF)-1/6 satellite with (16 m) spatial and (2 d) temporal resolution. To obtain cloud-free images, the Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM) algorithm was employed in this study. By using the algorithm on the coarse cloudless images at the same or similar time as the fine images with cloud cover, the cloudless fine images were obtained, and the cloudless EVI2 time-series and phenological features were generated. The developed method was applied to identify grassland communities in Ordos, China. The results show that the Caragana pumila Pojark, Caragana davazamcii Sanchir and Salix schwerinii E. L. Wolf grassland, the Potaninia mongolica Maxim, Ammopiptanthus mongolicus S. H. Cheng and Tetraena mongolica Maxim grassland, the Caryopteris mongholica Bunge and Artemisia ordosica Krasch grassland, the Calligonum mongolicum Turcz grassland, and the Stipa breviflora Griseb and Stipa bungeana Trin grassland are distinguished with an overall accuracy of 87.25%. The results highlight that, compared to multispectral data only, the addition of EVI2 time-series and phenological features improves the classification accuracy by 9.63% and 14.7%, respectively, and even by 27.36% when these two features are combined together, and indicate the advantage of the fine images in this study, compared to 500 m moderate-resolution imaging spectroradiometer (MODIS) data, which are commonly used for grassland classification at regional scale, while using 16 m GF data suggests a 23.96% increase in classification accuracy with the same extracted features. This study indicates that the proposed method is suitable for regional-scale grassland community classification.

High spatial resolution (1–5 m) remotely sensed datasets are increasingly being used to map land covers over large geographic areas using supervised machine learning algorithms. Although many studies have compared machine learning classification methods, sample selection methods for acquiring training and validation data for machine learning, and cross-validation techniques for tuning classifier parameters are rarely investigated, particularly on large, high spatial resolution datasets. This work, therefore, examines four sample selection methods—simple random, proportional stratified random, disproportional stratified random, and deliberative sampling—as well as three cross-validation tuning approaches—k-fold, leave-one-out, and Monte Carlo methods. In addition, the effect on the accuracy of localizing sample selections to a small geographic subset of the entire area, an approach that is sometimes used to reduce costs associated with training data collection, is investigated. These methods are investigated in the context of support vector machines (SVM) classification and geographic object-based image analysis (GEOBIA), using high spatial resolution National Agricultural Imagery Program (NAIP) orthoimagery and LIDAR-derived rasters, covering a 2,609 km2 regional-scale area in northeastern West Virginia, USA. Stratified-statistical-based sampling methods were found to generate the highest classification accuracy. Using a small number of training samples collected from only a subset of the study area provided a similar level of overall accuracy to a sample of equivalent size collected in a dispersed manner across the entire regional-scale dataset. There were minimal differences in accuracy for the different cross-validation tuning methods. The processing time for Monte Carlo and leave-one-out cross-validation were high, especially with large training sets. For this reason, k-fold cross-validation appears to be a good choice. Classifications trained with samples collected deliberately (i.e., not randomly) were less accurate than classifiers trained from statistical-based samples. This may be due to the high positive spatial autocorrelation in the deliberative training set. Thus, if possible, samples for training should be selected randomly; deliberative samples should be avoided.