Land use transition is essentially one of the manifestations of land use/cover change (LUCC). Although a large number of studies have focused on land use transitions on the macro scale, there are few studies on the micro scale. Based on the data of two high-resolution land use surveys, this study used a land use transfer matrix and GeoDetector model to explore the spatial-temporal patterns and driving forces of land use transitions at the village level in Pu County over a ten-year period. Results show that Pu County has experienced a drastic process of land use transition. More than 80% of cropland and grassland have been converted to forest land, and over 90% of the expansion of built-up land came from the occupation of forest land, cropland, and grassland. The driving forces of land use transition and its magnitude depended on the type of land use. The implementation of the policy of returning farmland to forest, or grain-for-green (GFG) was the main driving force for the large-scale conversion of cultivated land to forest land in Pu County. In the context of policy of returning farmland to forests, the hilly and gully regions of China’s Loess Plateau must balance between protecting the ecology and ensuring food security. Promoting the comprehensive consolidation of gully land and developing modern agriculture may be an important way to achieve a win-win goal of ecological protection and food security.

Vegetation on the Loess Plateau, China, has continuously improved thanks to certain ecological restoration (ER) strategies, including the integrated soil conservation project that began in the late 1970s and the “Grain for Green” project that began in the 1990s. The experience of these strategies in different geomorphological regions is of great value to ER worldwide. In this study, the evolution of the land-use transition (LUT) pathway and ecosystem service value (ESV) in four geomorphological regions of the Beiluo River Basin was analyzed using geo-informatic Tupu and the equivalent factor method with data from 1975 to 2015. The results indicated that, from 1975 to 2015, the proportion of forestland in the Beiluo River basin increased by 18.27%, while the areas of shrub, grassland, cultivated land, and water decreased by 1.03%, 0.16%, 18.23%, and 0.26%, respectively. In the past 40 years, the overall ESV of the basin increased by USD 3.209 billion (54.16%). The landform, vegetation cover, LUT, and ESV analysis indicated that the main ecological functions of the loess hilly and gully (LHG), loess plateau gully (LPG), rocky mountain (RM), and terrace and plain (TP) regions are soil and water conservation (SWC), SWC and food production, regulation and food production, respectively. ER projects enhanced the main ecological function of individual regions. In detail, the transition of “cultivated land → grassland” enhanced SWC function in the LHG region, and the transition of “grassland (shrub) → forestland” enhanced the regulating services of the RM and LPG regions. Moreover, the transition of “cultivated land to grassland” did not seriously lower the food production services of the TP and LPG regions, owing to the increase in grain yield per unit area. However, there were alternating transitions between cultivated land and ecological land types, implying a game between the peasant households’ demands and the ER strategies. Conflicting demands between local households and the public necessitate precision ER strategies, including land planning, ecological compensation, training and employment for local residents, etc.

Soil erodibility (K factor) and saturated hydraulic conductivity (Ks) are essential indicators for the estimation of erosion intensity and can potentially influence soil nutrient losses, making them essential parameters for the evaluation of land reclamation quality. In this study, 132 soil samples from 22 soil profiles were collected to measure soil physicochemical properties (e.g., particle size distribution, bulk density and soil nutrient content) and calculate the K factor and Ks of reclaimed soils across the South Dump of the Pingshuo opencast coalmine in the Loess Plateau, China. Geostatistical analysis and the kriging interpolation were employed to quantify the spatial variations in the K factor and Ks in different layers. The results show that the K factor at 0–10 cm is obviously lower than that of other soil layers due to the external input of organic matter, while the Ks tends to decrease along with soil depth. Horizontally, the K factor at 0–10 cm and 50–60 cm shows a decreasing tendency from west to east, while that of other soil layers seems not to show any spatial distribution pattern along latitude or longitude. Meanwhile, the Ks at 0–10 cm presents a striped distribution pattern, while that of other soil layers shows a patchy pattern. On the other hand, the independent-sample t-test and Spearman’s correlation analysis were carried out to determine the effects of soil erodibility on total nitrogen (TN), soil organic matter (SOM), available phosphorus (AP) and potassium (AK). Overall, the K factor is negatively correlated with TN (r = −0.362, p < 0.01) and SOM contents (r = −0.380, p < 0.01), while AP and AK contents are mainly controlled by Ks. This study provides insight on the optimization of reclamation measures and the conservation of soil nutrients in reclaimed land of similar ecosystems.

Fine land cover classification (FLCC) of complex landscapes is a popular and challenging task in the remote sensing community. In complex surface-mined areas (CSMAs), researchers have conducted FLCC using traditional machine learning methods and deep learning algorithms. However, convolutional neural network (CNN) algorithms that may be useful for FLCC of CSMAs have not been fully investigated. This study proposes a multimodal remote sensing data and multiscale kernel-based multistream CNN (3M-CNN) model. Experiments based on two ZiYuan-3 (ZY-3) satellite imageries of different times and seasons were conducted in Wuhan, China. The 3M-CNN model had three main features: (1) multimodal data-based multistream CNNs, i.e., using ZY-3 imagery-derived true color, false color, and digital elevation model data to form three CNNs; (2) multisize neighbors, i.e., using different neighbors of optical and topographic data as inputs; and (3) multiscale convolution flows revised from an inception module for optical and topographic data. Results showed that the proposed 3M-CNN model achieved excellent overall accuracies on two different images, and outperformed other comparative models. In particular, the 3M-CNN model yielded obvious better visual performances. In general, the proposed process was beneficial for the FLCC of complex landscape areas.

Determining the regulating factors of soil nutrient variations can guide the implementation of land reclamation measures in opencast coalmine regions. In this study, 132 soil samples were collected at 22 sample sites in the South Dump of Pingshuo opencast coalmine, and soil physicochemical properties were separately measured to obtain the related soil information. Geostatistical analyses were employed to analyze the spatial distribution patterns of soil organic carbon stocks (SOCD), total nitrogen stocks (TND), available phosphorus stocks (APD), and available potassium stocks (AKD) at 0–60 cm. The results showed that the spatial distributions of these soil nutrient stocks were characterized by moderate (TND) to strong (SOCD, APD, and AKD) spatial dependence. Meanwhile, the values of SOCD (16.4–60.1 Mg ha⁻¹) and TND (1.9–15.5 Mg ha⁻¹) were much higher than those of APD (0.022–0.095 Mg ha⁻¹) and AKD (0.31–1.40 Mg ha⁻¹). The statistical analyses indicated that the influence of afforestation on SOCD, TND, APD, and AKD was not significant, and the dynamic variations of soil nutrient contents were mainly regulated by soil pH in the South Dump. The findings of this study can provide some scientific guidance for soil nutrient management in the opencast coalmine regions of similar ecosystems.

The pollution of ammonia-nitrogen in water became more serious with the development of the industry. In this paper, hydroxyapatite (HAP) was prepared from waste fish scales; the XRD, FTIR, SEM, TEM and Raman were used to characterize the prepared samples, and the prepared HAP samples were studied as degradation photocatalyst and adsorbing material for the removal of ammonia-nitrogen. The ammonia-nitrogen removal efficiency (ANRE, Removal Rate = (Initial Concentration–Current Concentration)/Initial Concentration × 100%) of the samples was also discussed. The results showed that the HAP sample, which was calcinated at 800 °C for 2 h, had the best ANRE, which was 41%. The UV-Vis, EIS, and PL were used to analyze the reason for the different ANRE. The reason for the high ANRE was attributed to the samples’ superior adsorption capability resulting from their large specific surface area, as well as their superior properties for photocatalytic degradation. This was a meaningful work to explore a treatment of ammonia-nitrogen by HAP.