Search Results (21)

Invasive species negatively affect ecosystems, economies, and human health by outcompeting native species and altering habitats. Ailanthus altissima, also known as the tree of heaven, an invasive species native to China that has spread to North America and Europe. Commonly found in urban areas and forestland, these invasive plants cause ecological and economic damage to local ecosystems; they are also the preferred host of other invasive species. Ecological stability refers to the balance and harmony in species populations. Invasive species like A. altissima disrupt this stability by outcompeting native species, leading to imbalances, and there was a lack of research and data on the tree of heaven. To address this issue, this study leveraged deep learning and satellite imagery recognition to generate reliable and comprehensive prediction maps in the USA. Four deep learning models were trained to recognize satellite images obtained from Google Earth, with A. altissima data obtained from the Life Alta Murgia project, LIFE12 BIO/IT/000213. The best performing fine-tuned model using binary classification achieved an AUC score of 90%. This model was saved locally and used to predict the density and probability of A. altissima in the USA. Additionally, multi-class classification methods corroborated the findings, demonstrating similar observational outcomes. The production of these predictive distribution maps is a novel method which offers an innovative and cost-effective alternative for extensive field surveys, providing reliable data for concurrent and future research on the environmental impact of A. altissima. Full article

Manual forestland classification methods, which rely on predetermined scoring criteria and subjective interpretation, are commonly used but suffer from limitations such as high labor costs, complexity, and lack of scalability. This study proposes an innovative machine learning-based approach to forestland classification, utilizing a Support Vector Machine (SVM) model to automate the classification process and enhance both efficiency and accuracy. The main contributions of this work are as follows: A machine learning model was developed using integrated data from the Third National Land Survey of China, including forestry, grassland, and wetland datasets. Unlike previous approaches, the SVM model is optimized with Grid Search (GS), Genetic Algorithm (GA), and Particle Swarm Optimization (PSO) to automatically determine classification parameters, overcoming the limitations of manual rule-based methods. The performance of the SVM model was evaluated using confusion matrices, classification accuracy, and Matthews Correlation Coefficient (MCC). A comprehensive comparison under different optimization techniques revealed significant improvements in classification accuracy and generalization ability over manual classification systems. The experimental results demonstrated that the GA-SVM model achieved classification accuracies of 98.83% (test set) and 99.65% (overall sample), with MCC values of 0.9796 and 0.990, respectively, outpacing other optimization algorithms, including Grid Search (GS) and Particle Swarm Optimization (PSO). The GA-SVM model was applied to classify public welfare forestland in Kunyu City, yielding detailed classifications across various forestland categories. This result provides a more efficient and accurate method for large-scale forestland management, with significant implications for future land use assessments. The findings underscore the advantages of the GA-SVM model in forestland classification: it is efficient, accurate, and easy to operate. This study not only presents a more reliable alternative to conventional rule-based and manual scoring methods but also sets a precedent for using machine learning to automate and optimize forestland classification in future applications. Full article

Human activities and environmental changes have influenced the changes in landscape patterns, which in turn profoundly impact the variation in net primary productivity (NPP) of vegetation. Understanding the relationship between landscape patterns and NPP is of significant importance for maintaining ecosystem stability and improving the ecological environment. In this study, six land use types in the arid and semi-arid regions of Northwest China were selected, and five landscape pattern indices at the landscape level and four landscape pattern indices at the class level were used. Pearson correlation and multiple linear regression models were employed to analyze the relationship between landscape indices and NPP at a 100 km × 100 km grid scale. The results indicate that there are varying degrees of correlation between landscape pattern indices and NPP from 2001 to 2020, with different levels of variation over the 20-year period. The correlation between indices and NPP is higher at the class level than at the landscape level, and the increase in landscape abundance and fragmentation promotes an increase in NPP. At the landscape level, three landscape indices, namely NP (Number of Patches), PR (Patch Richness), and SHDI (Shannon’s Diversity Index), explain 45.4% of the variation in NPP. At the class level, NP, TE (Total Edge Length), and IJI (Dispersion and Juxtaposition Index) are the main influencing factors for NPP in cropland, forestland, and grassland. Therefore, in ecological governance, it is necessary to consider landscape pattern changes appropriately to maintain ecosystem stability. Full article

Soil microorganisms play crucial roles in maintaining material circulation and energy flow in desert ecosystems. However, the structure and function of soil microorganisms in different forestlands are currently unclear, restricting the use of sand-fixing plants and the understanding of forest ecosystem functions. In this study, Artemisia ordosica, Caragana korshinskii, and Salix psammophila, three types of sand-fixing forests widely distributed in the Mu Us Sandy Land, were used to explore the effects of sand-fixing forests on soil physicochemical properties, soil enzyme activity, soil microbial biomass, microbial community structure, and inter-microbial species relationships. Soils of forestlands showed higher soil organic carbon (SOC), total phosphorus (TP), and total nitrogen (TN) contents than bare sandy land. The SOC in bare sandy soil was only 0.84 g kg⁻¹, while it remained 1.55–3.46 g kg⁻¹ in forestland soils. The TN in bare sandy land soil was 0.07 g kg⁻¹, which was significantly lower than that in forestland soils (0.35–0.51 g kg⁻¹). The TP in bare sandy soil was 0.18 g kg⁻¹, significantly lower than that in forestland soils (0.46–0.69 g kg⁻¹). Afforestation of bare sandy land improved soil microbial carbon and nitrogen contents and increased microbial enzyme activities of acid phosphatase and N-acetyl-β-D-glucosaminidase. Significant differences were observed between the three forestlands and bare sandy land in terms of soil microorganisms and community composition. With the establishment of a sand-fixing forest, the alpha diversity of soil bacteria significantly improved, whereas that of soil fungi remained stable. The bacterial community comprised 33 phyla, 106 classes, 273 orders, 453 families, and 842 genera. While five fungal phyla were detected by OTUs at a similarity of 97%, bacterial and fungal community structures were affected by the organic carbon content, sand particle content, soil pH, total nitrogen, and total phosphorus contents of soils. This study is helpful for vegetation construction and protection on sandy lands from the perspective of plant-microbe interactions. Full article

Natural vegetation on both sides of the Tarim River Basin (TRB) is the only barrier—a critical ecological niche—between the economic belt in the artificial oasis and the Taklimakan Desert. To understand the impact of human activities on the TRB, we explored the spatial and temporal variations in land use/land cover change (LUCC) and landscape pattern evolution from 2000 to 2020. These variations were simulated for 2030 with the 20 years of data using the cellular automata–Markov model and geographical information system analyses. The results predicted substantial LUCCs in the lower reaches of the Tarim River (TRlr), with 3400 km² (20.29%) of the total area (16,760.94 km²) undergoing changes. Wetland, artificial land, grassland, farmland, and forestland areas increased by 578.59, 43.90, 339.90, 201.62, and 536.11 km², respectively, during the period from 2020 to 2030. The only decreases were in the Gobi/other deserts and bare soils (1700.13 km²). We also determined current and future changes in TRlr landscape pattern indices at the class and landscape levels. Combined with a field survey and hydrological data, theoretical support for effective land use management strategies is provided. The findings offer a scientific basis for future ecological civilization construction and sustainable development in the TRB. Full article

Understanding the spatiotemporal changes in landscape patterns and their driving factors in Inner Mongolia can benefit land use and ecological environment management in this region. This study used the county landscape index and multiple regression analysis to reveal the temporal and spatial evolutions of landscape patterns and their driving factors in Inner Mongolia from 2000 to 2015 with multitemporal land use data. The results showed that (1) grassland was the main landscape type in Inner Mongolia. Grassland and unused land decreased, and cropland expanded from 2000 to 2015. Grassland degradation has slowed since 2005. (2) At the class level, the dominance of grassland decreased, and the degree of landscape fragmentation of cropland, forestland, and grassland increased gradually. At the landscape level, the landscape shape was more complex, the landscape connectivity was worse, and the landscape diversity gradually enhanced. (3) This study revealed that climatic factors influenced the evolution of landscape patterns, and human activities were the key driving factors of landscape-level metrics. The results of this study provide scientific bases for land management strategies. Full article

Changes in land use and land cover (LULC) are the leading contributors to the decline and loss of ecosystem services in the world. The present study covered the Central Rift Valley lakes basin in Ethiopia, focusing on the valley floor and the East and West escarpments, to analyze changes in LULC and to estimate associated losses in ecosystem service values (ESVs). Covering both upstream and downstream areas in the basin, the study addressed major gaps in existing studies by connecting the sources and sinks of material (e.g., sediment and water) in source-to-lake systems. Additionally, the study facilitated the identification of critical areas for conserving natural resources and reversing the decline of associated ESVs in the Central Rift Valley. A post-classification comparison approach was used to detect LULC changes between 1973 and 2020 using four Landsat images from 1973, 1990, 2005 and 2020. The value transfer valuation method was used to estimate the changes in ESVs due to LULC changes. Among the seven major identified LULC classes, farmlands, settlements, and bare lands showed positive changes, while forestlands, grasslands, shrublands and waterbodies showed negative changes over the last 47 years. The expansion of farmlands, for example, has occurred at the expense of grasslands, forestlands and shrublands. The changes in LULC over a period of 47 years resulted in a total loss of US $62,110.4 × 10⁶ in ESVs. The contributors to the overall loss of ESVs in decreasing order are provisioning services (US $33,795.1 × 10⁶), cultural services (US $28,981.5 × 10⁶) and regulating services (US $652.9 × 10⁶). The results imply that addressing the degradation of land and water resources is crucial to reversing the loss of ecosystem services and achieving the national Sustainable Development Goals (SDGs) related to food and water security (SDGs 2 and 6) and life on land (SDG 15). Full article

Cypress (Cupressus funebris Endl.) is a major tree species planted for forestland restoration in low-fertility soil and in areas where rocky desertification has occurred. Calcium (Ca) fertilizer can adjust the pH of soil and has an important effect on the growth of cypress. Soil and water losses are serious in Southern China, and soil acidification is increasing, which results in high calcium loss. However, the adaptability of cypress clones to different concentrations of calcium in acidic soils has not been studied. In this investigation, a potted-plant experiment was set up with three concentrations of calcium oxide (CaO) fertilizer (0, 3, and 6 g·kg⁻¹) added under local soil conditions with 0 and 3 g·kg⁻¹ nitrogen (N), phosphorus (P), and potassium (K) fertilizer. The effects of CaO on the growth, root development, and nutrient uptake and utilization efficiency of cypress clones were analyzed. The growth, root development, and nutrient absorption and utilization of cypress differed when calcium fertilizer was applied to acidic soils with different degrees of fertility. In the soil with 0 g·kg⁻¹ NPK fertilizer, the 3 and 6 g·kg⁻¹ CaO treatments significantly increased the clonal growth of cypress seedling height, basal diameter, and dry-matter weight. In addition, the length, surface area, and volume of the roots less than 2.0 mm of root diameter also significantly increased, indicating that the fine cypress roots were somewhat able to adapt to differing Ca levels under lower fertility conditions. Moreover, the efficiency of N, P, and Ca accumulation was highest in the 3 g·kg⁻¹ CaO treatment. After adding 3 g·kg⁻¹ CaO fertilizer to the soil with 3 g·kg⁻¹ NPK fertilizer, only the root dry-matter weight increased significantly, indicating that root development (including root length, surface area, and volume) in the D1–D3 diameter classes (≤1.5 mm in diameter) was significantly elevated. When CaO application reached 6 g·kg⁻¹, the seedling height, basal diameter, and dry-matter weight of each organ decreased, as did the length, surface area, and volume of the roots in the all diameter classes, indicating that the addition of excessive CaO to fertile soil could inhibit the growth and root development of cypress. In Ca-deficient low-quality acidic soils, adding CaO fertilizer can promote the development of fine roots and the uptake and utilization of N, P, and Ca. The results of this study provide a basis for determining the optimal fertilization strategy when growing cypress in acidic soils in Southern China. Full article

Quantifying forest fires remain a challenging task for the implementation of public policies aimed to mitigate climate change. In this paper, we propose a new method to provide an annual burned area map of Mato Grosso State located in the Brazilian Amazon region, taking advantage of the high spatial and temporal resolution sensors. The method consists of generating the vegetation, soil, and shade fraction images by applying the Linear Spectral Mixing Model (LSMM) to the Landsat-8 OLI (Operational Land Imager), PROBA-V (Project for On-Board Autonomy–Vegetation), and Suomi NPP-VIIRS (National Polar-Orbiting Partnership-Visible Infrared Imaging Radiometer Suite) datasets. The shade fraction images highlight the burned areas, in which values are represented by low reflectance of ground targets, and the mapping was performed using an unsupervised classifier. Burned areas were evaluated in terms of land use and land cover classes over the Amazon, Cerrado and Pantanal biomes in the Mato Grosso State. Our results showed that most of the burned areas occurred in non-forested areas (66.57%) and old deforestation (21.54%). However, burned areas over forestlands (11.03%), causing forest degradation, reached more than double compared with burned areas identified in consolidated croplands (5.32%). The results obtained were validated using the Sentinel-2 data and compared with active fire data and existing global burned areas products, such as the MODIS (Moderate Resolution Imaging Spectroradiometer product) MCD64A1 and MCD45A1, and Fire CCI (ESA Climate Change Initiative) products. Although there is a good visual agreement among the analyzed products, the areas estimated were quite different. Our results presented correlation of 51% with Sentinel-2 and agreement of r² = 0.31, r² = 0.29, and r² = 0.43 with MCD64A1, MCD45A1, and Fire CCI products, respectively. However, considering the active fire data, it was achieved the better performance between active fire presence and burn mapping (92%). The proposed method provided a general perspective about the patterns of fire in various biomes of Mato Grosso State, Brazil, that are important for the environmental studies, specially related to fire severity, regeneration, and greenhouse gas emissions. Full article

Estimation of the soil organic carbon (SOC) content is of utmost importance in understanding the chemical, physical, and biological functions of the soil. This study proposes machine learning algorithms of support vector machines (SVM), artificial neural networks (ANN), regression tree, random forest (RF), extreme gradient boosting (XGBoost), and conventional deep neural network (DNN) for advancing prediction models of SOC. Models are trained with 1879 composite surface soil samples, and 105 auxiliary data as predictors. The genetic algorithm is used as a feature selection approach to identify effective variables. The results indicate that precipitation is the most important predictor driving 14.9% of SOC spatial variability followed by the normalized difference vegetation index (12.5%), day temperature index of moderate resolution imaging spectroradiometer (10.6%), multiresolution valley bottom flatness (8.7%) and land use (8.2%), respectively. Based on 10-fold cross-validation, the DNN model reported as a superior algorithm with the lowest prediction error and uncertainty. In terms of accuracy, DNN yielded a mean absolute error of 0.59%, a root mean squared error of 0.75%, a coefficient of determination of 0.65, and Lin’s concordance correlation coefficient of 0.83. The SOC content was the highest in udic soil moisture regime class with mean values of 3.71%, followed by the aquic (2.45%) and xeric (2.10%) classes, respectively. Soils in dense forestlands had the highest SOC contents, whereas soils of younger geological age and alluvial fans had lower SOC. The proposed DNN (hidden layers = 7, and size = 50) is a promising algorithm for handling large numbers of auxiliary data at a province-scale, and due to its flexible structure and the ability to extract more information from the auxiliary data surrounding the sampled observations, it had high accuracy for the prediction of the SOC base-line map and minimal uncertainty. Full article

National monitoring of forestlands and the processes causing canopy cover loss, be they abrupt or gradual, partial or stand clearing, temporary (disturbance) or persisting (deforestation), are necessary at fine scales to inform management, science and policy. This study utilizes the Landsat archive and an ensemble of disturbance algorithms to produce maps attributing event type and timing to >258 million ha of contiguous Unites States forested ecosystems (1986–2010). Nationally, 75.95 million forest ha (759,531 km²) experienced change, with 80.6% attributed to removals, 12.4% to wildfire, 4.7% to stress and 2.2% to conversion. Between regions, the relative amounts and rates of removals, wildfire, stress and conversion varied substantially. The removal class had 82.3% (0.01 S.E.) user’s and 72.2% (0.02 S.E.) producer’s accuracy. A survey of available national attribution datasets, from the data user’s perspective, of scale, relevant processes and ecological depth suggests knowledge gaps remain. Full article

This paper presents a new approach for rapidly assessing the extent of land use and land cover (LULC) areas in Mato Grosso state, Brazil. The novel idea is the use of an annual time series of fraction images derived from the linear spectral mixing model (LSMM) instead of original bands. The LSMM was applied to the Project for On-Board Autonomy-Vegetation (PROBA-V) 100-m data composites from 2015 (~73 scenes/year, cloud-free images, in theory), generating vegetation, soil, and shade fraction images. These fraction images highlight the LULC components inside the pixels. The other new idea is to reduce these time series to only six single bands representing the maximum and standard deviation values of these fraction images in an annual composite, reducing the volume of data to classify the main LULC classes. The whole image classification process was conducted in the Google Earth Engine platform using the pixel-based random forest algorithm. A set of 622 samples of each LULC class was collected by visual inspection of PROBA-V and Landsat-8 Operational Land Imager (OLI) images and divided into training and validation datasets. The performance of the method was evaluated by the overall accuracy and confusion matrix. The overall accuracy was 92.4%, with the lowest misclassification found for cropland and forestland (<9% error). The same validation data set showed 88% agreement with the LULC map made available by the Landsat-based MapBiomas project. This proposed method has the potential to be used operationally to accurately map the main LULC areas and to rapidly use the PROBA-V dataset at regional or national levels. Full article

The expansion of urban areas and the increase in the number of buildings and urbanization characteristics, such as roads, affect the meteorological environment in urban areas, resulting in weakened pollutant dispersion. First, this paper uses GIS (geographic information system) spatial analysis technology and landscape ecology analysis methods to analyze the dynamic changes in land cover and landscape patterns in Chengdu as a result of urban development. Second, the most appropriate WRF (Weather Research and Forecasting) model parameterization scheme is selected and screened. Land-use data from different development stages in the city are included in the model, and the wind speed and temperature results simulated using new and old land-use data (1980 and 2015) are evaluated and compared. Finally, the results of the numerical simulations by the WRF-Chem air quality model using new and old land-use data are coupled with 0.25° × 0.25°-resolution MEIC (Multi-resolution Emission Inventory for China) emission source data from Tsinghua University. The results of the sensitivity experiments using the WRF-Chem model for the city under different development conditions and during different periods are discussed. The meteorological conditions and pollution sources remained unchanged as the land-use data changed, which revealed the impact of urban land-use changes on the simulation results of PM_2.5 atmospheric pollutants. The results show the following. (1) From 1980 to 2015, the land-use changes in Chengdu were obvious, and cultivated land exhibited the greatest changes, followed by forestland. Under the influence of urban land-use dynamics and human activities, both the richness and evenness of the landscape in Chengdu increased. (2) The microphysical scheme WSM3 (WRF Single–Moment 3 class) and land-surface scheme SLAB (5-layer diffusion scheme) were the most suitable for simulating temperatures and wind speeds in the WRF model. The wind speed and temperature simulation results using the 2015 land-use data were better than those using the 1980 land-use data when assessed according to the coincidence index and correlation coefficient. (3) The WRF-Chem simulation results obtained for PM_2.5 using the 2015 land-use data were better than those obtained using the 1980 land-use data in terms of the correlation coefficient and standard deviation. The concentration of PM_2.5 in urban areas was higher than that in the suburbs, and the concentration of PM_2.5 was lower on Longquan Mountain in Chengdu than in the surrounding areas. Full article

This study analyzes the impact of land use/land cover (LULC) changes on the hydrology of the dry Afromontane forest landscape in northern Ethiopia. Landsat satellite images of thematic mapper (TM) (1986), TM (2001), and Operational Land Imager (OLI) (2018) were employed to assess LULC. All of the images were classified while using the maximum likelihood image classification technique, and the changes were assessed by post-classification comparison. Seven LULC classes were defined with an overall accuracy 83–90% and a Kappa coefficient of 0.82–0.92. The classification result for 1986 revealed dominance of shrublands (48.5%), followed by cultivated land (42%). Between 1986 and 2018, cultivated land became the dominant (39.6%) LULC type, accompanied by a decrease in shrubland to 32.2%, as well as increases in forestland (from 4.8% to 21.4%) and bare land (from 0% to 0.96%). The soil conservation systems curve number model (SCS-CN) was consequently employed to simulate forest hydrological response to climatic variations and land-cover changes during three selected years. The observed changes in direct surface runoff, the runoff coefficient, and storage capacity of the soil were partially linked to the changes in LULC that were associated with expanding bare land and built-up areas. This change in land use aggravates the runoff potential of the study area by 31.6 mm per year on average. Runoff coefficients ranged from 25.3% to 47.2% with varied storm rainfall intensities of 26.1–45.4 mm/ha. The temporal variability of climate change and potential evapotranspiration increased by 1% during 1981–2018. The observed rainfall and modelled runoff showed a strong positive correlation (R² = 0.78; p < 0.001). Regression analysis between runoff and rainfall intensity indicates their high and significant correlation (R² = 0.89; p < 0.0001). Changes were also common along the slope gradient and agro-ecological zones at varying proportions. The observed changes in land degradation and surface runoff are highly linked to the change in LULC. Further study is suggested on climate scenario-based modeling of hydrological processes that are related to land use changes to understand the hydrological variability of the dry Afromontane forest ecosystems. Full article

This study proposes a workflow for land use and land cover (LULC) classification of Advanced Land Observing Satellite-2 (ALOS-2) Phased Array type L-band Synthetic Aperture Radar-2 (PALSAR-2) images of the Brazilian tropical savanna (Cerrado) biome. The following LULC classes were considered: forestlands; shrublands; grasslands; reforestations; croplands; pasturelands; bare soils/straws; urban areas; and water reservoirs. The proposed approach combines polarimetric attributes, image segmentation, and machine-learning procedures. A set of 125 attributes was generated using polarimetric ALOS-2/PALSAR-2 images, including the van Zyl, Freeman–Durden, Yamaguchi, and Cloude–Pottier target decomposition components, incoherent polarimetric parameters (biomass indices and polarization ratios), and HH-, HV-, VH-, and VV-polarized amplitude images. These attributes were classified using the Naive Bayes (NB), DT J48 (DT = decision tree), Random Forest (RF), Multilayer Perceptron (MLP), and Support Vector Machine (SVM) algorithms. The RF, MLP, and SVM classifiers presented the most accurate performances. NB and DT J48 classifiers showed a lower performance in relation to the RF, MLP, and SVM. The DT J48 classifier was the most suitable algorithm for discriminating urban areas and natural vegetation cover. The proposed workflow can be replicated for other SAR images with different acquisition modes or for other types of vegetation domains. Full article