Kaimin Wang

Timely and accurate tree species mapping is crucial for forest resource inventory, supporting management, conservation biology, and ecological restoration. This study utilized Sentinel-1 and Sentinel-2 data to classify five dominant tree species in Chengde and Beijing. To effectively capture the influence of multi-temporal data, data were acquired in March, June, September, and December 2020, extracting various features, including bands, spectral indices, texture features, and topographic variables. The optimal input variable combination was explored using 1519 field survey samples for training and testing datasets. Classification employed Random Forest, XGBoost, and deep learning models, with performance evaluated through out-of-bag estimation and cross-validation. The XGBoost model achieved the highest accuracy of 81.25% (kappa = 0.74) when using Sentinel-1 and Sentinel-2 bands, indices, texture features, and DEM data. Results demonstrate the effectiveness of using Sentinel data for tree species classification and emphasize the value of machine learning algorithms. This study underscores the potential of combining synthetic aperture radar (SAR) and optical data for large-scale tree species classification, with significant implications for forest monitoring and management.

Long-term exposure to high concentrations of fine particles can cause irreversible damage to people’s health. Therefore, it is of extreme significance to conduct large-scale continuous spatial fine particulate matter (PM_2.5) concentration prediction for air pollution prevention and control in China. The distribution of PM_2.5 ground monitoring stations in China is uneven with a larger number of stations in southeastern China, while the number of ground monitoring sites is also insufficient for air quality control. Remote sensing technology can obtain information quickly and macroscopically. Therefore, it is possible to predict PM_2.5 concentration based on multi-source remote sensing data. Our study took China as the research area, using the Pearson correlation coefficient and GeoDetector to select auxiliary variables. In addition, a long short-term memory neural network and random forest regression model were established for PM_2.5 concentration estimation. We finally selected the random forest regression model (R² = 0.93, RMSE = 4.59 μg m⁻³) as our prediction model by the model evaluation index. The PM_2.5 concentration distribution across China in 2021 was estimated, and then the influence factors of high-value regions were explored. It is clear that PM_2.5 concentration is not only related to the local geographical and meteorological conditions, but also closely related to economic and social development.

Grassland desertification stands as an ecological concern globally. It is crucial for desertification prevention and control to comprehend the variation in area and severity of desertified grassland (DGL), clarify the intensities of conversion among DGLs of different desertification levels, and explore the spatial and temporal driving factors of desertification. In this study, a Desertification Difference Index (DDI) model was constructed based on albedo-EVI to extract desertification information. Subsequently, intensity analysis, the Geo-detector model, and correlation analysis were applied to analyze the dynamics and driving factors of desertification. The results showed the following: (1) Spatially, the DGL in Xilingol exhibited a zonal distribution. Temporally, the degree of DGL decreased, with the proportion of severely and moderately desertified areas decreasing from 51.77% in 2000 to 37.23% in 2020, while the proportion of nondesertified and healthy areas increased from 17.85% in 2000 to 37.40% in 2020; (2) Transition intensities among different desertification levels were more intense during 2000–2012, stabilizing during 2012–2020; (3) Meteorological factors and soil conditions primarily drive the spatial distribution of DDI, with evapotranspiration exhibiting the most significant influence (q-value of 0.83), while human activities dominate interannual DDI variations. This study provides insights into the conversion patterns among different desertification levels and the divergent driving forces shaping desertification in both spatial and temporal dimensions in Xilingol.

Nighttime light data is a mainstay method in confirming and supporting other traditional economic data indicators, which in turn influence business and policy-making decisions. Accuracy in and clear definition of economic data and its related indicators are thus of great importance not only for analysis of urban development and related policies seeking sustainable development, but also plays a key role in whether or not these policies are successful. Discovering and recognizing the applications and limitations of nighttime light and other peripheral data could prove helpful in future data analysis and sustainable development policy. One possible limitation could exist in GDP decrease, and whether or not nighttime light would decrease accordingly, as most studies show nighttime light increase confirms economic growth, thus affecting the reliability of the data’s correlation with economic data. This study utilizes nighttime light data in a cross-reference with GDP data during instances of global GDP shrinkage over the years of 2007–2017, split between 2007–2012 for the DMSP dataset and 2013–2017 for the VIIRs dataset. It seeks to establish through linear regression whether or not yearly average nighttime light data products show a positive correlation even during periods of economic decline, thereby providing a clearer understanding of the strengths and limitations of NTL as an economic validation indicator. Analysis shows that both years of global GDP decrease in turn also displayed global nighttime light decrease, in addition to linear regression giving satisfactory results pointing to a positive correlation over the timespan. The VIIRS data series resulted in higher regression coefficients, which is in line with the results of previous literature.

Gross primary productivity (GPP) is an important parameter that represents the productivity of vegetation and responses to various ecological environments. Using the Mann–Kendall methods, Pearson correlation, and the Geodetector, this study investigated the spatiotemporal variation and driving factors of GPP from 2000 to 2020. The results showed that (1) in terms of spatial distribution, GPP showed a trend of “low-high-low” regions, with low values for grassland and arable land and a high value for forest land. The growth trend is fast in forest areas, while the growth trend is not obvious in cultivated areas. The regions with significant growth accounted for 68.73% of the whole region. (2) The whole region shows a growth rate of 2.07 g C∙m⁻²∙yr⁻¹, showing obvious seasonality, with a slow growth trend in spring and autumn and a fast growth trend in summer. (3) The driving factors of GPP spatial differentiation in the Beijing-Tianjin-Hebei region were land surface temperature, land use type, and nighttime light data, while precipitation and downward surface shortwave radiation show no strong explanatory power for the spatial differentiation of GPP, which means that these two factors have less driving force on the spatial differentiation of GPP. The interaction of LUCC with the other factors presents two-factor enhancement, while the LST interaction with the other three factors presents non-linear enhancement. This study could provide a theoretical basis for the sustainable development of the Beijing-Tianjin-Hebei Region.

The combination of natural environment changes and human activities affects the growth of grasslands. In order to quantitatively assess the causes of spatial and temporal variation of grasslands in Xilingol, this study assessed the spatial and temporal evolution patterns of grassland health based on MOD13A1 long time series Normalized Difference Vegetation Index (NDVI) data from 2000–2019 using trend analysis. The geodetector model was used to explore the dominant drivers of spatial variation in grassland NDVI, combined with 34 factors covering natural environmental changes and human disturbances over the same period. The results show that the grasslands of Xilingol showed an overall recovery trend from 2000 to 2019, with an average annual NDVI growth rate of 0.0028/a, a monthly increasing rate of 0.0005/month, and 68.06% of the grassland at an average recovery level. Moisture-dominated natural climate change factors, such as Growing Season Precipitation (Prep2), Annual Mean Water Vapor Pressure (WVP), and Annual Mean Relative Humidity (RH), were the underlying cause of grassland health changes during the study period, with the highest explanatory factor being growing season precipitation (q value of 0.59 on a multi-year average). The influence of primary production value among human activities was greater, and the explanatory factor of tertiary production value showed an increasing trend. The interactions among natural and anthropogenic factors significantly enhances their explanatory credibility for NDVI, with the type of interaction dominated by the two-factor enhancement. Risk detection of the top 10 dominant drivers in terms of q statistic were carried out to obtain the threshold range of each driver in the high zone of grassland NDVI, which can provide a scientific reference for the sustainable restoration of grassland.

The large area estimation of forest canopy closure (FCC) using remotely sensed data is of high interest in monitoring forest changes and forest health, as well as in assessing forest ecological services. The accurate estimation of FCC over the regional or global scale is challenging due to the difficulty of sample acquisition and the slow processing efficiency of large amounts of remote sensing data. To address this issue, we developed a novel bounding envelope methodology based on vegetation indices (BEVIs) for determining vegetation and bare soil endmembers using the normalized differences vegetation index (NDVI), modified bare soil index (MBSI), and bare soil index (BSI) derived from Landsat 8 OLI and Sentinel-2 image within the Google Earth Engine (GEE) platform, then combined the NDVI with the dimidiate pixel model (DPM), one of the most commonly used spectral-based unmixing methods, to map the FCC distribution over an area of more than 90,000 km². The key processing was the determination of the threshold parameter in BEVIs that characterizes the spectral boundary of vegetation and soil endmembers. The results demonstrated that when the threshold equals 0.1, the extraction accuracy of vegetation and bare soil endmembers is the highest with the threshold range given as (0, 0.3), and the estimated spatial distribution of FCC using both Landsat 8 and Sentinel-2 images were consistent, that is, the area with high canopy density was mainly distributed in the western mountainous region of Chifeng city. The verification was carried out using independent field plots. The proposed approach yielded reliable results when the Landsat 8 data were used (R² = 0.6, RMSE = 0.13, and 1-rRMSE = 80%), and the accuracy was further improved using Sentinel-2 images with higher spatial resolution (R² = 0.81, RMSE = 0.09, and 1-rRMSE = 86%). The findings demonstrate that the proposed method is portable among sensors with similar spectral wavebands, and can assist in mapping FCC at a regional scale while using multispectral satellite imagery.

Accurate information on tree species is in high demand for forestry management and further investigations on biodiversity and environmental monitoring. Over regional or large areas, distinguishing tree species at high resolutions faces the challenges of a lack of representative features and computational power. A novel methodology was proposed to delineate the explicit spatial distribution of six dominant tree species (Pinus tabulaeformis, Quercus mongolia, Betula spp., Populus spp., Larix spp., and Armeniaca sibirica) and one residual class at 10 m resolution. Their spatial patterns were analyzed over an area covering over 90,000 km² using the analysis-ready large volume of multisensor imagery within the Google Earth engine (GEE) platform afterwards. Random forest algorithm built into GEE was used together with the 20th and 80th percentiles of multitemporal features extracted from Sentinel-1/2, and topographic features. The composition of tree species in natural forests and plantations at the city and county-level were performed in detail afterwards. The classification achieved a reliable accuracy (77.5% overall accuracy, 0.71 kappa), and the spatial distribution revealed that plantations (Pinus tabulaeformis, Populus spp., Larix spp., and Armeniaca sibirica) outnumber natural forests (Quercus mongolia and Betula spp.) by 6% and were mainly concentrated in the northern and southern regions. Arhorchin had the largest forest area of over 4500 km², while Hexingten and Aohan ranked first in natural forest and plantation area. Additionally, the class proportion of the number of tree species in Karqin and Ningcheng was more balanced. We suggest focusing more on the suitable areas modeling for tree species using species’ distribution models and environmental factors based on the classification results rather than field survey plots in further studies.