Soil organic matter (SOM) is the main source of soil nutrients, which are essential for the growth and development of agricultural crops. Hyperspectral remote sensing is one of the most efficient ways of estimating the SOM content. Visible, near infrared, and mid-infrared reflectance spectroscopy, combined with the partial least squares regression (PLSR) method is considered to be an effective way of determining soil properties. In this study, we used 54 different spectral pretreatments to preprocess soil spectral data. These spectral pretreatments were composed of three denoising methods, six data transformations, and three dimensionality reduction methods. The three denoising methods included no denoising (ND), Savitzky–Golay denoising (SGD), and wavelet packet denoising (WPD). The six data transformations included original spectral data, R; reciprocal, 1/R; logarithmic, log(R); reciprocal logarithmic, log(1/R); first derivative, R’; and first derivative of reciprocal, (1/R)’. The three dimensionality reduction methods included no dimensionality reduction (NDR), sensitive waveband dimensionality reduction (SWDR), and principal component analysis (PCA) dimensionality reduction (PCADR). The processed spectra were then employed to construct PLSR models for predicting the SOM content. The main results were as follows—(1) the wavelet packet denoising (WPD)-R’ and WPD-(1/R)’ data showed stronger correlations with the SOM content. Furthermore, these methods could effectively limit the correlation between the adjacent bands and, thus, prevent “overfitting”. (2) Of the 54 pretreatments investigated, WPD-(1/R)’-PCADR yielded the model with the highest accuracy and stability. (3) For the same denoising method and spectral transformation data, the accuracy of the SOM content estimation model based on SWDR was higher than that of the model based on NDR. Furthermore, the accuracy in the case of PCADR was higher than that for SWDR. (4) Dimensionality reduction was effective in preventing data overfitting. (5) The quality of the spectral data could be improved and the accuracy of the SOM content estimation model could be enhanced effectively, by using some appropriate preprocessing methods (one combining WPD and PCADR in this study).

Understanding the long-term dynamics and driving factors behind small and micro wetlands is critical for their management and future sustainability. This study explored the impacts of natural and anthropogenic factors on the spatiotemporal evolution of these areas in Wuxi area using the support vector machine (SVM) classification method and the geographic detector model based on Landsat satellite image data from 1985 to 2020. The results revealed that: (1) Natural wetlands were prominent in Wuxi area, with an average proportion of 70%, and although they exhibited a downward trend over the last ten years, the scale of natural small and micro wetlands increased 1.5-fold—from 4349.59 hm² in 1985 to 10,841.59 hm² in 2020. (2) The small and micro wetlands in Wuxi area had obvious seasonal variations, with most being 0.1–1 hm² and 1–3 hm², respectively. From the perspective of spatial distribution, they were primarily distributed in Yixing district, which accounts for 34% of Wuxi area. (3) The distribution of small and micro wetlands was systematically affected by natural and human activities. The main factors that affected the distribution of small and micro wetlands were the average annual temperature and GDP, with the interactions between all factors being nonlinear and bi-linear. The influences of natural factors on small and micro wetlands were weakened, with human activities steadily emerging as the dominant factor that affected their distribution. The results of this study can provide supportive data and a scientific basis for the ecological restoration and protection of wetlands.

Time-series monitoring of the land subsidence in the Yangtze River coastal area is crucial for maintaining river stability and early warning of disasters. This study employed PS-InSAR and SBAS-InSAR techniques to monitor the land subsidence along the Yangtze River in Nanjing, using a total of 42 Sentinel-1A images obtained between April 2015 and November 2021. The accuracy of both methods was compared and validated, while a comprehensive analysis was conducted to ascertain the spatial distribution characteristics and underlying causes of land subsidence. The maximum deviation between the two methods and six leveling point data did not exceed ±5 mm. Within the 5 km buffer zone on either side of the Yangtze River in Nanjing, four subsidence funnels were identified. Analysis of the factors contributing to land subsidence in this area indicates that underground engineering construction and operation, increasing ground building area, and geological condition all have certain correlations to the land subsidence. The results obtained through PS-InSAR and SBAS-InSAR technologies revealed a high degree of consistency in monitoring outcomes, and the latter method exhibited superior monitoring accuracy than the former one in this area. This study holds significant implications for guiding the scientific management of urban geohazards along the Yangtze River.

Vegetation phenology is a sensitive indicator of climate change, and can help understand the response of vegetation cycles to climate, which is important for understanding the impact of global climate change on terrestrial ecosystems. In this study, based on the enhanced vegetation index (EVI) time-series data, derived from the moderate-resolution imaging spectroradiometer (MODIS) data, the climate parameters were extracted using the Savitzky–Golay (S–G) filtering method to explore the spatial and temporal variation characteristics of the vegetation phenology in an agro-pasture ecotone in China, from 2000 to 2020. In addition, the response characteristics of the vegetation phenology to the climate elements (temperature and precipitation) were also analyzed. The results showed the following: (1) The start of the growing season (SOS) was widely advanced, and that was caused by climate change. The end of the growing season (EOS) was delayed, and the length of the growing season (LOS) was gradually extended with a large interannual fluctuation in the SOS and the LOS in the region; (2) the SOS showed significant negative correlations with the air temperature and precipitation. Precipitation was mainly positively correlated with the EOS, but there was no significant difference in the correlation between temperature and the EOS. In general, pre-season precipitation is the main driver of the vegetation phenology, while the influence of temperature on the phenology is less obvious; (3) in the semi-arid area and arid area, the phenology was mainly influenced by precipitation. The response of the vegetation phenology to the temperatures in different temperature zones was found to be regular, showing high spatial differences. In general, the higher the cumulus temperature, the lower the negative effect of the temperature on both the SOS and EOS. These results may provide new reference to study the non-systematic changes of the vegetation phenology in response to climate change.

The spatial–temporal simulation of fire disasters and evacuation route planning are important research fields for urban emergency responses and are primary tasks that answer complex questions after fires break out. The increasing demand for refined building information models will sharply increase the calculated and analyzed quantity. This demand presents a challenge for fire emergency responses based on massive building information. In this paper, the principle of the realistic worst case (RWC) is introduced into fire simulation and evacuation route planning. Taking the library of the Nanjing Forestry University as the study object, the spatial–temporal characteristics of the influential environmental factors of the fire are simulated, such as the meteorological elements, building structure, and building skin. The scenario zones that are relatively prone to fire are selected using an overlay analysis across the four seasons. Then, the risk threshold for evacuating personnel is analyzed in the fire zone according to international standards and firefighting criteria. Specific parameters are determined based on the analysis of the above. The growing trends for fires across the four seasons are simulated with scenario zones as the starting positions and incorporate factors such as temperature, carbon monoxide, and smoke. Lastly, a life safety assurance path (LSAP) for personnel evacuation is designed based on an indoor road network and path search algorithm. The evacuation planning result is compared with the traditional shortest-time path and shortest-distance path. Based on the study results, fire scenario zones can improve the speed and operating efficiency of spatial–temporal simulation models of fire and can also support path planning and design for emergency responses.

This paper presents a novel framework to achieve 3D semantic labeling of objects (e.g., trees, buildings, and vehicles) from airborne laser-scanning point clouds. To this end, we propose a framework which consists of hierarchical clustering and higher-order conditional random fields (CRF) labeling. In the hierarchical clustering, the raw point clouds are over-segmented into a set of fine-grained clusters by integrating the point density clustering and the classic K-means clustering algorithm, followed by the proposed probability density clustering algorithm. Through this process, we not only obtain a more uniform size and more homogeneous clusters with semantic consistency, but the topological relationships of the cluster’s neighborhood are implicitly maintained by turning the problem of topology maintenance into a clustering problem based on the proposed probability density clustering algorithm. Subsequently, the fine-grained clusters and their topological context are fed into the CRF labeling step, from which the fine-grained cluster’s semantic labels are learned and determined by solving a multi-label energy minimization formulation, which simultaneously considers the unary, pairwise, and higher-order potentials. Our experiments of classifying urban and residential scenes demonstrate that the proposed approach reaches 88.5% and 86.1% of “m $F_1$ ” estimated by averaging all classes of the $F_1$ -scores. We prove that the proposed method outperforms five other state-of-the-art methods. In addition, we demonstrate the effectiveness of the proposed energy terms by using an “ablation study” strategy.