No abstract

Reducing the dimension of the hyperspectral image data can directly reduce the redundancy of the data, thus improving the accuracy of hyperspectral image classification. In this paper, the deep belief network algorithm in the theory of deep learning is introduced to extract the in-depth features of the imaging spectral image data. Firstly, the original data is mapped to feature space by unsupervised learning methods through the Restricted Boltzmann Machine (RBM). Then, a deep belief network will be formed by superimposed multiple Restricted Boltzmann Machines and training the model parameters by using the greedy algorithm layer by layer. At the same time, as the objective of data dimensionality reduction is achieved, the underground feature construction of the original data will be formed. The final step is to connect the depth features of the output to the Softmax regression classifier to complete the fine-tuning (FT) of the model and the final classification. Experiments using imaging spectral data showing the in-depth features extracted by the profound belief network algorithm have better robustness and separability. It can significantly improve the classification accuracy and has a good application prospect in hyperspectral image information extraction.

Research on the service values of urban ecosystems is a hot topic of ecological studies in the current era of rapid urbanization. To quantitatively estimate the ecosystem service value in Chengdu, China from the perspectives of natural ecology and social ecology, the technologies of remote sensing (RS) and geographic information system (GIS) are utilized in this study to extract the land use type information from RS images of Chengdu in 2003, 2007, 2013 and 2018. Subsequently, a driver analysis of the ecosystem services of Chengdu was performed based on socioeconomic data from the last 16 years. The results indicated that: (1) from 2003 to 2018, the land utilization in Chengdu changed significantly, with the area of cultivated lands, forest lands and water decreasing remarkably, while the area of construction lands dramatically increased. (2) The ecosystem services value (ESV) of Chengdu decreased by 30.92% in the last 16 years, from CNY 2.4078 × 1010 in 2003 to CNY 1.6632 × 1010 in 2018. Based on a future simulation, the ESV is further predicted to be reduced to CNY 1.4261 × 1010 by 2033. (3) The ESV of Chengdu showed a negative correlation with the total population, the urbanization rate and the per capita GDP of the region, indicating that the ESV of the studied region was inter-coupled with the socioeconomic development and can be maintained at a high level through rationally regulating the socioeconomic structure.

The Three Gorges Reservoir region in China is the Yangtze River Economic Zone’s natural treasure trove. Its natural environment has an important role in development. The unique and fragile ecosystem in the Yangtze River’s Three Gorges Reservoir region is prone to natural disasters, including soil erosion, landslides, debris flows, landslides, and earthquakes. Therefore, to better alleviate these threats, an accurate and comprehensive assessment of the susceptibility of this area is required. In this study, based on the collection of relevant data and existing research results, we applied machine learning models, including logistic regression (LR), the random forest model (RF), and the support vector machine (SVM) model, to analyze landslide susceptibility in the Yangtze River’s Three Gorges Reservoir region to analyze landslide events in the whole study region. The models identified five categories (i.e., topographic, geological, ecological, meteorological, and human engineering activities), with nine independent variables, influencing landslide susceptibility. The accuracy of landslide susceptibility derived from different models and raster cells was then verified by the accuracy, recall, F1-score, ROC curve, and AUC of each model. The results illustrate that the accuracy of different machine learning algorithms is ranked as SVM > RF > LR. The LR model has the lowest generalization ability. The SVM model performs well in all regions of the study area, with an AUC value of 0.9708 for the entire Three Gorges Reservoir area, indicating that the SVM model possesses a strong spatial generalization ability as well as the highest robustness and can be adapted as a real-time model for assessing regional landslide susceptibility.

Soil erosion in the Three-River Headwaters Region (TRHR) of the Qinghai-Tibet Plateau in China has a significant impact on local economic development and ecological environment. Vegetation and precipitation are considered to be the main factors for the variation in soil erosion. However, it is a big challenge to analyze the impacts of precipitation and vegetation respectively as well as their combined effects on soil erosion from the pixel scale. To assess the influences of vegetation and precipitation on the variation of soil erosion from 2005 to 2015, we employed the Revised Universal Soil Loss Equation (RUSLE) model to evaluate soil erosion in the TRHR, and then developed a method using the Logarithmic Mean Divisia Index model (LMDI) which can exponentially decompose the influencing factors, to calculate the contribution values of the vegetation cover factor (C factor) and the rainfall erosivity factor (R factor) to the variation of soil erosion from the pixel scale. In general, soil erosion in the TRHR was alleviated from 2005 to 2015, of which about 54.95% of the area where soil erosion decreased was caused by the combined effects of the C factor and the R factor, and 41.31% was caused by the change in the R factor. There were relatively few areas with increased soil erosion modulus, of which 64.10% of the area where soil erosion increased was caused by the change in the C factor, and 23.88% was caused by the combined effects of the C factor and the R factor. Therefore, the combined effects of the C factor and the R factor were regarded as the main driving force for the decrease of soil erosion, while the C factor was the dominant factor for the increase of soil erosion. The area with decreased soil erosion caused by the C factor (12.10×10 3 km 2 ) was larger than the area with increased soil erosion caused by the C factor (8.30×10 3 km 2 ), which indicated that vegetation had a positive effect on soil erosion. This study generally put forward a new method for quantitative assessment of the impacts of the influencing factors on soil erosion, and also provided a scientific basis for the regional control of soil erosion.

After the “5·12” Wenchuan earthquake in 2008, collapses and landslides have occurred continuously, resulting in the accumulation of a large quantity of loose sediment on slopes or in gullies, providing rich material source reserves for the occurrence of debris flow and flash flood disasters. Therefore, it is of great significance to build a collapse and landslide susceptibility evaluation model in Wenchuan County for local disaster prevention and mitigation. Taking Wenchuan County as the research object and according to the data of 1081 historical collapse and landslide disaster points, as well as the natural environment, this paper first selects six categories of environmental factors (13 environmental factors in total) including topography (slope, aspect, curvature, terrain relief, TWI), geological structure (lithology, soil type, distance to fault), meteorology and hydrology (rainfall, distance to river), seismic impact (PGA), ecological impact (NDVI), and impact of human activity (land use). It then builds three single models (LR, SVM, RF) and three CF-based hybrid models (CF-LR, CF-SVM, CF-RF), and makes a comparative analysis of the accuracy and reliability of the models, thereby obtaining the optimal model in the research area. Finally, this study discusses the contribution of environmental factors to the collapse and the landslide susceptibility prediction of the optimal model. The research results show that (1) the areas prone to extremely high collapse and landslide predicted by the six models (LR, CF-LR, SVM, CF-SVM, RF and CF-RF) have an area of 730.595 km2, 377.521 km2, 361.772 km2, 372.979 km2, 318.631 km2, and 306.51 km2, respectively, and the frequency ratio precision of collapses and landslides is 0.916, 0.938, 0.955, 0.956, 0.972, and 0.984, respectively; (2) the ranking of the comprehensive index based on the confusion matrix is CF-RF>RF>CF-SVM>CF-LR>SVM>LR and the ranking of the AUC value is CF-RF>RF>CF-SVM>CF-LR>SVM>LR. To a certain extent, the coupling models can improve precision more over the single models. The CF-RF model ranks the highest in all indexes, with a POA value of 257.046 and an AUC value of 0.946; (3) rainfall, soil type, and distance to river are the three most important environmental factors, accounting for 24.216%, 22.309%, and 11.41%, respectively. Therefore, it is necessary to strengthen the monitoring of mountains and rock masses close to rivers in case of rainstorms in Wenchuan county and other similar areas prone to post-earthquake landslides.

No abstract

Gridded precipitation data with a high spatiotemporal resolution are of great importance for studies in hydrology, meteorology, and agronomy. Observational data from meteorological stations cannot accurately reflect the spatiotemporal distribution and variations of precipitation over a large area. Meanwhile, radar-derived precipitation data are restricted by low accuracy in areas of complex terrain and satellite-based precipitation data by low spatial resolution. Therefore, hourly precipitation models were employed to merge data from meteorological stations, Radar, and satellites; the models used five machine learning algorithms (XGBoost, gradient boosting decision tree, random forests (RF), LightGBM, and multiple linear regression (MLR)), as well as the CoKriging method. In the north of Guangdong Province, data of four heavy rainfall events in 2018 were processed with geographic data to obtain merged hourly precipitation data. The CoKriging method secured the best prediction of spatial distribution of accumulated precipitation, followed by the tree-based machine learning (ML) algorithms, and significantly, the prediction of MLR deviated from the actual pattern. All machine learning methods showed poor performances for timepoints with little precipitation during the heavy rainfall events. The tree-based ML method showed poor performance at some timepoints when precipitation was over-related to latitude, longitude, and distance from the coast.