Abstract Multiple uplifts of the Qinghai-Tibet Plateau since the late Cenozoic have generated large effects on the geomorphology and environment over Asia. However, how the basin responded to the uplifted Tibetan Plateau remains unknown. Here, we examined the digital geomorphology of the Bahe River basin, the greatest first-order tributary in the southern Weihe River basin. The geomorphogenesis of the watershed, the far-field effects, and the long-range consequences of the elevation from the Tibetan Plateau were thoroughly discussed. We found that there were five layers of terraces, which were mainly controlled by tectonic activity. The geomorphological characteristics suggested that Bahe basin was influenced by a severe tectonic activity prime, with intense river erosion. The Lishan-North Qinling Mountains uplift, facilitated by the horizontal expansion of the northern Qinghai-Tibet Plateau, was also responsible for the development of the Bahe River basin.

Abstract As people’s living standards improve, tourism has become an important way for people to spend their time on leisure and entertainment. The growing number of tourists in recent years has given rise to the creation of tourism-related ancillary services. Travelers need to choose a travel route that suits their needs and expectations and do it in a way that does not cause a waste of time, whether it is an emerging self-driving tour or a traditional tour group. Therefore, the optimization of tourist routes is of great significance to the majority of tourists. Given the planning requirements of tourist attractions in the post-epidemic era, an ant colony-based optimization algorithm is proposed to resolve the planning problem of optimal tourist routes. An optimized pheromone update strategy is also proposed based on the basic ant colony optimization algorithm. The optimized ant colony algorithm tries to balance two conflicting concepts, namely, flows into tourist attractions and the carrying capacity of destinations. To analyze the performance of the proposed optimization algorithm, the effects of different optimization algorithms on the route planning of tourist attractions were compared in the experiment, and the acceleration ratio of the optimized ant colony algorithm was tested using the graphics processing unit parallel computing program. The results show that the proposed algorithm provides certain advantages and has certain potential in parallel computing. To sum up, this study provides a better scientific basis for optimal tourist route planning and has a good reference value.

Abstract Dolines are depressions in the soluble ground that indicates the degree of karstification. They may also act as connection points (vulnerability spots) between the surface and underground for the transmission of runoff, sediments, and pollutants. The delineation of these spots (dolines) is a crucial step in environmental management through land use planning to protect the karst underground, which is rich in flora and fauna. This requirement can benefit from a cost-effective, accessible, and non-invasion high-resolution investigation generating digital elevation models (DEMs) from unmanned aerial vehicle (UAV) imagery and automated object detection techniques. This study examines the capabilities of UAV-based DEM in detecting dolines across 50 km2 in the environmentally protected area of river Vermelho (APANRV – Área de Proteção Ambiental das Nascentes do Rio Vermelho). Initially, an automatic objects (doline and no-doline) detection algorithm was applied to the DEM, followed by a visual inspection to differentiate doline from possible dolines in orthomosaic photos, topographic profiles, and shaded UAV-based relief (digital terrain model; DTM and DSM). For the redundancy checking, a cluster analysis with four tests was conducted. The objects generated from the best clusters and morphological analysis were gathered in the same base for visual inspection. Out of a total of 933 objects identified, 41% were obtained from the DSM base, 25% from the perimeter-to-area ratio, and 34% through convergence between the two-analyses. Subsequently, the resulting doline typologies are discussed in reference to their proximity to hydrogeological features and their impacts on underground vulnerability. The findings aligned with the previous research as dolines were highly concentrated near sites where carbonates come in contact with siliciclastic sediments.

Abstract The geochemical sampling work in the difficult and dangerous areas is very hard; hence, it can be greatly improved by combining with the remotely sensed data. Thus, a retrieval model is proposed by Kernel Principal Component Analysis and Artificial Bee Colony (ABC) optimized Support Vector Machine (SVM) models based on Landsat 8 remotely sensed data and the geochemical data in the study area. The analysis results show that the geochemical data delineate the areas with relatively enriched elements, but indicate the low-abnormal ore (chemical) points, and the anomalies delineated by the inversion data are better for this purpose, for better indication. At the same time, the distribution and intensity of the corresponding abnormal areas found that the abnormal areas delineated by the inversion data basically contain the abnormal areas delineated by the original data, and the anomalies located at the ore spots are obviously enhanced; it shows that the SVM model of ABC Optimization can establish the relation between geochemistry data and remote sensing data, can supply the original data effectively, and can also provide the direction for the next prospecting work.

Abstract The landslide susceptibility assessment is an essential part of landslide disaster risk identification and prevention. However, the binarization of the hidden layer limits the parameterization ability of the conditional probability of visible layer, making the training process of restricted Boltzmann machine more difficult and further limiting the accuracy and efficiency of deep belief network (DBN) model in landslide susceptibility assessment. Therefore, this study proposed a landslide susceptibility assessment method based on Auto-Encoder (AE)-modified DBN. Zhouqu County, Gansu Province in the People’s Republic of China, was selected as the study area. Historical landslides in Zhouqu County were identified using small baseline subset interferometric synthetic aperture radar technology and optical image. Landslide factors were screened based on a geographical detector and stepwise regression method. The Logcosh loss function and determinant coefficient R 2 index were used to evaluate the training process of the AE model, and the balanced cross entropy loss function was used to evaluate the entire network training process. In addition, the area under the curve (AUC) of the synthetical index model (SIM), support vector machine (SVM), and multilayer perceptron (MLP) were compared and evaluated. The results indicated that the proposed model could significantly improve the accuracy of landslide susceptibility assessment. The AUC value of the proposed model was 0.31, 0.12, and 0.11 higher than that of SIM, SVM, and MLP, respectively. Therefore, the improved DBN model based on AE proposed is reliable for early landslide identification and prediction.

Abstract Global warming has caused changes in various ecological processes and has potential to change ecosystems’ stability. In spite of comprehensive studies to investigate air temperatures under global warming, much less is known about changes in soil temperatures, particularly in deep layers. Herein, we used 30 years of soil temperature data from a temperate steppe region to assess vertical characteristics and their changes in soil temperature from the surface to a depth of 3.20 m. We determined, apparently for the first time, that the soil temperature is the lowest at 0.2 cm layer at an annual level. Furthermore, the vertical variation of soil temperature (temperature variation with soil depth) strictly conformed to composite exponential function curves, and there were two composite exponential function curves that are symmetric to each other, to represent soil temperature in a pair of months with a difference of 6 months. Parameters in the functions changed as the soil warmed over 30 years. This finding explored the pattern of soil temperature in deep layers depending on the mathematics model. Model building and understanding is beneficial for predicting vertical and temporal extensions of soil temperature and their impact on below-ground processes in regional ecosystem.

Abstract The current cultivated land quality (CLQ) evaluation method fails to consider the determination of soil nutrients, resulting in the low efficiency of soil nutrient message extraction. In an effort to effectively solve the above problems, combined with the spatial variation of soil nutrients (SVSN) at the field scale, a CLQ evaluation method is proposed. The soil nutrients were determined according to the soil spatial variation analysis, and the soil color was standardized. The characteristic bands were determined by soil fertility and nutrients such as nitrogen, phosphorus, and potassium, and the soil nutrient message was preprocessed. On this basis, the soil nutrient message extraction model was constructed. According to the damage principle of subsoiling shovel, the limit value of the membership function of the CLQ exponent is determined, and the weighted sum method is used to calculate the CLQ exponent, so as to realize the SVSN and the evaluation of CLQ at the field scale. The experimental results show that the dimensional autocorrelation of soil bulk density and soil water content in different soil layers is high and that of capillary porosity, non-capillary porosity, and total porosity in different soil layers is strong.

Abstract Heavy metal pollution of groundwater can have severe potential impacts on human health and the environment. The aim of this study was to evaluate the spatial distribution, pollution, and health risk of heavy metals in groundwater in the main pepper production area of China. A total of 67 groundwater samples were collected, and the concentrations of cadmium (Cd), chromium (Cr), copper (Cu), nickel (Ni), manganese (Mn), and zinc (Zn) elements were analyzed. The results indicated that the average concentrations of Cd, Cr, Cu, Mn, Ni, and Zn in groundwater are lower than the limits of the Standard for Groundwater Quality of China (Class III). However, the maximum concentrations of Mn, Cd, and Ni surpass the Standard for Groundwater Quality of China. The spatial distribution patterns of the concentrations and pollution levels of the six heavy metals in groundwater to be substantially heterogeneous. Furthermore, the pollution level of heavy metals in groundwater decrease in the following order: Mn > Cd > Ni > Cr > Zn > Cu, and the collected groundwater samples are found to be slightly polluted by heavy metals. Overall, the non-carcinogenic risk of the investigated metals, instigated primarily by oral ingestion of groundwater, are found to be higher than the acceptable range for children, and Cd poses the most significant health risk among the investigated metals.

Abstract Land cover in tropical marine climate zones is important for global climate change. The existing analysis of land cover product consistency mainly focuses on a continental or national scale and rarely takes different geographical zones (such as tropical marine climate zones) as examples to carry out micro-interpretation from the perspective of ecology from the grid scale. In fact, some types of land cover under different zones have poor accuracy due to the standard of cognition and the complexity of the spatial pattern of ground objects. In addition, land cover and its change in tropical Marine climate zones will affect the greenhouse effect, energy balance, water transport, and so on, thus affecting climate change on a regional or even global scale. Therefore, this article presents an evaluation based on GLOBCOVER, CCI LC, and MCD12Q1 data using Malaysia as a case study, through area composition similarity, field sample point validation, and landscape indices. The results showed that (1) the area correlation coefficient between GLOBCOVER and CCI LC is the highest at 0.998. (2) The CCI LC had the highest OA and kappa coefficient of 59.01% and 0.4957, while the GLOBCOVER product had the lowest OA and kappa coefficient of 49.24% and 0.3614, respectively. (3) The consistency of the water landscape index is high between the CCI LC and GLOBCOVE data, the consistency of the artificial surfaces landscape index is high between the CCI LC and MCD12Q1 products, and the consistency of the grassland/shrubland landscape index is high between the GLOBCOVE and MCD12Q1 products. The results of microscopic landscape patterns show that the three product landscape patterns are generally more consistent in East Malaysia than in West Malaysia. The low accuracy of grassland, bareland, and shrubland is the key reason for the wide variation in landscape patterns between the three products.

Abstract Three-dimensional (3D) city modeling is an essential component of 3D geoscience modeling, and window detection of building facades plays a crucial role in 3D city modeling. Windows can serve as structural priors for rapid building reconstruction. In this article, we propose a framework for detecting window lines. The framework consists of two parts: an improved stacked hourglass network and a point–line extraction module. This framework can output vectorized window wireframes from building facade images. Besides, our method is end-to-end trainable, and the vectorized window wireframe consists of point–line structures. The point–line structure contains both semantic and geometric information. Additionally, we propose a new dataset of real-world building facades for window-line detection. Our experimental results demonstrate that our proposed method has superior efficiency, accuracy, and applicability in window-line detection compared to existing line detection algorithms. Moreover, our proposed method presents a new idea for deep learning methods in window detection and other application scenarios in current 3D geoscience modeling.