Search Results (6,192)

Land-cover-based ecosystem service valuation commonly assigns a uniform value coefficient to pixels within the same land-cover class, thereby overlooking ecological heterogeneity in mountainous protected areas. This study developed an improved remote sensing ecological index (IRSEI) and applied it as a spatial adjustment factor in the equivalent factor method to assess ecosystem service value (ESV) changes in the Qionglai Mountains section of Giant Panda National Park, China, between 2017 and 2022. IRSEI integrated the normalized difference vegetation index, wetness, normalized difference built-up and soil index, land surface temperature, and cumulative dynamic habitat index derived from the fraction of absorbed photosynthetically active radiation. DHI-cum was included as a proxy for annual cumulative vegetation productivity and habitat energy availability rather than a direct measure of biodiversity or giant panda habitat quality. Total ESV increased from 3.27 × 10⁸ CNY in 2017 to 4.25 × 10⁸ CNY in 2022, representing an increase of 9.79 × 10⁷ CNY, or 29.98%. Water bodies contributed the largest absolute increase, rising by 4.79 × 10⁷ CNY, or 42.45%, whereas farmland showed the highest relative increase of 45.35%. Woodland remained the dominant contributor to total ESV. Spatially, ESV was higher in the northern and southern parts and lower in the central region. All corrected sensitivity coefficients were below one, indicating that total ESV responded inelastically to ±50% perturbations of individual land-cover value coefficients. The framework improves within-class spatial differentiation of ESV and may support targeted management of mountainous protected areas, although field-based habitat and biodiversity data are needed for further validation. Full article

Net primary productivity (NPP) is a core diagnostic variable of terrestrial carbon cycling, yet consistent annual NPP simulation remains challenging in alpine heterogeneous regions where topography, hydrothermal gradients, vegetation structure, and nutrient constraints interact. Remote-sensing products such as MODIS provide valuable observational constraints, whereas light-use-efficiency models such as CASA retain process transparency and scenario transfer capability. This study develops a CASA-based, MODIS-constrained framework for annual NPP simulation over the Gannan Plateau. The framework preserves a locally parameterized CASA baseline and adds a geographically weighted regression (GWR) residual-alignment layer trained on CASA–MODIS residuals during 2005–2013. The fitted correction relationship was then applied to the 2014–2020 temporal transfer period and evaluated in a 2030s SSP scenario transfer experiment. MODIS was treated as the correction target rather than ground truth, and GLASS was adopted as an independent product-level benchmark. During 2014–2020, the GWR-corrected product showed improved pooled pixel-level agreement with the MODIS-constrained target relative to parameter-localized CASA, with R² increasing from 0.438 to 0.708 and RMSE decreasing from 91.4 to 68.7 g C m⁻² yr⁻¹. Residual Moran’s I also decreased, indicating weaker residual spatial organization after correction. Product-level comparison with GLASS showed a moderate, directionally consistent improvement relative to uncorrected CASA, although this comparison was not interpreted as ground-truth validation. The 2030s scenario transfer experiment indicated that the correction layer changed the spatial expression of NPP divergence among SSP pathways. Overall, the proposed framework provides a process-model-preserving and observation-constrained approach for improving agreement between annual NPP estimates and the MODIS-constrained target in alpine heterogeneous regions, while its applicability remains subject to product uncertainty, spatial dependence, and future nonstationarity. Full article

Payments for Ecosystem Services (PES) are increasingly used to promote conservation, but evidence of their biophysical outcomes remains mixed because participation is rarely assigned at random. This study assessed the association between participation in the Yo Protejo ¡Agua para Todos! PES program and satellite-based vegetation dynamics in Cundinamarca, Colombia, during 2015–2024. Annual NDVI, EVI, and SAVI values were derived from Landsat 8 and 9 imagery processed in Google Earth Engine. The empirical strategy combined fixed-effects difference-in-differences models, two spatial comparison groups, propensity-score adjustment, inverse probability of treatment weighting, property-clustered inference, event-study diagnostics, differential-trend models, and property-year robustness analyses. Pre-treatment and covariate-balance diagnostics indicated imperfect counterfactual comparability, particularly for the external 500–1000 m buffer-ring controls. No vegetation index produced uniformly positive and robust associations across specifications. NDVI showed a small positive association during program implementation under the 500 m neighboring-control definition, but this result was not preserved after differential-trend adjustment or property-year aggregation. EVI produced positive point-level estimates under the external control, although this specification retained substantial residual imbalance and the estimates became negative after property-year aggregation. SAVI provided no robust positive evidence and showed negative post-implementation differentials in some external-control models. Overall, the estimated associations were sensitive to the vegetation metric, comparison-group definition, temporal phase, inferential level, and spatial aggregation. The results should therefore be interpreted as specification-dependent adjusted associations rather than definitive causal effects. Full article

Green and Blue Infrastructure (GBI) is increasingly addressed within climate adaptation and mitigation policies as a strategic operational measure for reducing climate-related impacts in urban environments. However, GBI effectiveness relies strictly on the biophysical condition of natural assets and on the capacity to monitor and interpret ecosystem processes over time, specifically vegetation quality and its physiological response to climatic stressors within complex urban fabrics. This variability emphasizes the need to integrate ecosystem performance into decision-making through digital frameworks capable of quantifying and accounting for ecological resources across temporal scales. In this context, Remote Sensing (RS) technologies provide a structured informational basis for assessing vegetation health and surface thermal patterns in relation to climatic stress thresholds and human exposure. This paper presents a policy-aligned geospatial evidence framework to bridge the gap between environmental monitoring and urban climate action. By integrating high-resolution multispectral remote sensing with heterogeneous spatial datasets and climate models, the framework enables the multitemporal assessment of ecological conditions to inform where GBI measures can support SECAP implementation, project refinement, and monitoring activities. Developed within the Horizon Europe KNOWING project and applied to the Naples East district, Italy, the framework was operationalized within the city’s Sustainable Energy and Climate Action Plan (SECAP). The application produces scenario-oriented outputs for interpreting the potential contribution of GBI and NbS measures to outdoor heat-stress reduction under SECAP conditions. Its practical value lies in translating biophysical data into reusable GIS/WMS layers that connect ecological performance, climate exposure, socio-energetic vulnerability, and planned urban transformations, thereby supporting SECAP implementation, project refinement, and monitoring. Full article

Ginger (Zingiber officinale) fields in northern China are often covered by plastic mulch film and shade netting, so Sentinel-2 records management materials as well as the crop canopy. Because these materials and their deployment differ among production systems, models calibrated locally may transfer poorly. We defined three observation windows for the main study counties: spring film mulching, summer shade-net coverage, and autumn exposed-canopy greening. Within each window, spectral bands, vegetation indices, and gray-level co-occurrence matrix (GLCM) textures were extracted from single Sentinel-2 scenes across three northern Chinese counties with contrasting practices. A cross-regional Random Forest Gini ranking, weighted by the similar county model-sample totals, retained four variables per feature family per stage (36 of 120 variables across the three-stage stack). The spectral-index-texture scheme achieved within-county F1 scores of 95.00% in Changyi, 94.83% in Qingzhou, and 95.58% in Fengrun. Leave-one-county-out (LOCO) model fitting returned a mean F1 of 94.73% compared with 95.14% for the within-county splits. Because the fixed-feature protocol was selected using all three counties, this LOCO test evaluates county-held-out classifier fitting rather than a fully nested feature-selection pipeline. Independent field verification with Global Positioning System (GPS) points ranged from 85.80% to 89.95%, and village-level area estimates agreed with remote-sensing totals (R² = 0.869). The 36-variable protocol performed similarly to the full 120-variable input (95.14% vs. 95.24% mean F1), and the selected features were stable across the three tested weighting rules. In Funing County, where shade nets are absent, omitting the shading stage on the basis of local agronomic practice yielded 93.75% accuracy against 96 independent GPS points. The results support management timing as a practical guide for ginger mapping within the tested northern production systems; wider climatic validation and fully nested transfer tests are still needed. Full article

Since its launch in 2015, the Sentinel-2 mission has become a cornerstone of moderate-resolution vegetation monitoring, enabling spatially explicit and temporally dense observations of terrestrial ecosystems. Its combination of 10–20 m spatial resolution, a revisit interval of less than five days, and a spectral configuration including red-edge and Short-Wave Infrared (SWIR) bands has transformed optical vegetation monitoring beyond coarse-resolution greenness products. This review synthesises the use of Sentinel-2 for vegetation monitoring, with emphasis on phenology and growth dynamics, biomass and carbon estimation, vegetation stress detection, and associated methodological developments. A systematic Scopus search identified 1700 publications, of which 1097 studies were retained following thematic and methodological screening. The results reveal rapid growth in Sentinel-2-based research after 2018, reflecting its transition into a widely adopted data source supported by cloud-based processing platforms and harmonised data products. Research output is concentrated in a limited number of journals and regions, with Europe and Asia dominating contributions, while other regions remain underrepresented. Phenology and growth monitoring, biomass and carbon assessment, and vegetation stress analysis emerged as the principal application domains. Across these themes, methodological development has shifted from vegetation indices towards machine learning, hybrid radiative-transfer modelling, and multi-sensor data fusion. The reviewed evidence indicates that no single methodological approach consistently outperforms others; rather, performance depends on the target variable, ecosystem characteristics, and the treatment of observational uncertainty. Sentinel-2 has transformed vegetation monitoring by enabling spatially explicit assessment of vegetation phenology, biomass, carbon dynamics, and stress across ecosystems. However, important challenges remain, including uncertainty propagation, limited sensitivity to early physiological stress, the absence of thermal observations, and uneven validation across ecosystem types. Future progress will depend on uncertainty-aware retrieval frameworks, physically informed hybrid models, multi-sensor integration, and expanded calibration and validation across underrepresented ecosystems. Full article

Kazakhstan’s Ili River delta nourishes a unique wetland ecosystem in arid Central Asia. The delta is dominated by common reed [Phragmites australis (Cav.) Trin. ex Steud.], an ecologically and economically significant species that is sensitive to water levels. We used machine learning methods, including the Random Forest algorithm, to classify common reed-containing wetland vegetation and water surfaces in the delta based on satellite data from 2000 to 2023. The remote sensing results were integrated with field-based geobotanical studies conducted between 2017 and 2019. These studies, which facilitated identification of three classes of vegetation: meadow, marsh and aquatic, also revealed a good correspondence between predicted and measured common reed biomass per unit area. The long-term dynamics of four wetland communities with a common reed content of ≥25% were analyzed with respect to significant interannual variability in the flow of the Ili River. The highest water inflows and water surface areas in the delta were recorded in 2002, 2010 and 2016, and in each case, a statistically significant expansion of common reed was observed one year later. Common reed areas subsequently declined—rapidly or after a lag of several years. Temporal expansion and contraction of these areas following pulses of water differed substantially from that of overall wetland vegetation as measured previously. The identified patterns have important scientific and practical significance for assessing the stability of the surrounding wetlands, preservation of the delta environment, and sustainable use of common reed. Full article

The objective of this study was to predict seed and oil yields in safflower (Carthamus tinctorius L.) using UAV-based multispectral imagery and machine learning algorithms. The study was conducted during the 2024 growing season under varying irrigation levels, fertilization practices, and applications of plant growth-promoting rhizobacteria. The experiment included four irrigation levels, fertilized and unfertilized conditions, and bacterial treatments consisting of Bacillus pumilus, Bacillus albus, their mixture, and a non-bacterial control. Sixty-two vegetation indices were calculated from multispectral images acquired during the harvest maturity period and used to predict seed and oil yields. To identify the most informative features, Mutual Information, Recursive Feature Elimination, and LASSO feature selection methods were applied; subsequently, Linear Regression, Decision Tree, Random Forest, Support Vector Regression, K-Nearest Neighbors, and XGBoost regression algorithms were compared. The results showed that Linear Regression combined with Mutual Information-based feature selection was the most successful approach for predicting both seed and oil yields. According to the 5-fold cross-validation results, average values of $R=0.8600$, $MAE=0.2870$, and $RMSE=0.3765$ were obtained for seed yield prediction, while average values of $R=0.8710$, $MAE=0.0876$, and $RMSE=0.1101$ were obtained for oil yield prediction. Full article

UAV-based remote sensing excels in rapid response, high timeliness, simple operation, and high degrees of automation, and has been widely applied for geological hazard monitoring. Deep learning methods based on unitemporal UAV images can only analyze the static appearance of a scene, while bitemporal change detection can capture the dynamic evolution of hazards; however, due to diverse geological landforms and topography, environmental noises such as vegetation cover, and dynamic weather conditions, change detection of geological hazards from UAV images based on traditional deep learning technology is not always effective. Therefore, there is an urgent need to utilize large vision-language models (LVLMs) to further improve the accuracy and robustness of the change detection model. Motivated by this, this paper proposes a novel remote sensing model for geological hazard monitoring, referred to as CLIFF (CLIP-BIT-EfficientNet), based on the multi-modal LVLM Contrastive Language–Image Pre-training (CLIP), the change detection network Bitemporal Image Transformer (BIT), and the classification network EfficientNet, along with corresponding datasets and model fine-tuning strategies. The proposed transfer fusion module bridges the CLIFF and BIT networks by aligning their feature distributions and dimensions, allowing the general knowledge of the LVLM and the task-specific knowledge of the learnable branch to reinforce each other. Furthermore, this integrated pipeline addresses the scarcity of labeled hazard data by allowing the BIT to train on larger public datasets, while fine-tuning EfficientNet on smaller hazard-classification datasets within the change area, making the approach more efficient and reliable than direct classification methods. Experimental results show that the proposed CLIFF algorithm outperforms state-of-the-art deep learning algorithms such as LightCDNet and ChangeFormer, with an IoU of 75.74% and an F1 score of 0.8689 for change detection. Meanwhile, CLIFF has an overall accuracy rate of 86.89% in identifying geological hazards along gas pipelines, such as crude oil spills, collapses, landslides, and floods, with per-class accuracies of 87.32% and 86.17% for crude oil spills and landslides, respectively. Full article

High-density rodent populations cause severe habitat degradation and ecological imbalance in natural grasslands through intense foraging and burrowing activities. However, dynamically monitoring these small mammals and assessing their large-scale damage using traditional ground surveys alone is challenging. In this study, we evaluated rodent damage severity in alpine meadows and typical steppe by proposing an integrated framework that combines ground, unmanned aerial vehicle (UAV), and satellite data. Using data from 36 plots per grassland type, we extracted a suite of ecological parameters, including aboveground biomass, vegetation cover, community height, rodent burrow density, and plant diversity metrics, to construct a plot-scale Rodent Damage Index ($RDI$). This $RDI$ was then linked with a satellite-derived Remote Sensing Ecological Index ($RSEI$) to model and map damage severity at the regional scale. Separate linear regression models were developed for the two grassland types. The alpine meadow model exhibited better model fit and predictive performance (fitting $R^2$ = 0.762, RMSE = 0.136; LOOCV $R^2$ = 0.729, RMSE = 0.145, 95% CI: 0.580–0.886) than the typical steppe model (fitting $R^2$ = 0.574, RMSE = 0.198; LOOCV $R^2$ = 0.478, RMSE = 0.214, 95% CI: 0.279–0.787), highlighting that the predictive relationship model performance differs significantly between grassland types. Our findings demonstrate that integrating multi-source, cross-scale spatial data is an effective approach for assessing rodent damage. Furthermore, these results indicate that rodent damage assessment should be grassland-type-specific to ensure accuracy and support targeted rodent damage management planning. Full article

This study evaluated a cost-effective UAV-based multi-sensor approach for characterizing river corridor conditions during and after moderate floods in two mountain river corridors in Northern Utah, USA: the Logan River and Blacksmith Fork River. These study reaches included urban, rural, and agricultural areas, hydraulic structures and bridges, and fish passage structures. A DJI Matrice 300 UAV was used with two separate payloads: an AgEagle Altum-PT multispectral camera and an R3 Pro V2 two-return LiDAR system. The workflow included UAV flight planning and data collection, post-processing of the multi-spectral and LiDAR sensor data, spatial resolution and accuracy assessment, and interpretation of the resultant data. The multi-spectral post-processing produced pansharpened orthomosaics with a spatial resolution of 0.0432 m, while the UAV LiDAR produced DSM/DTM products at 0.05 m resolution. LiDAR accuracy assessment showed vertical RMSE values of approximately 0.0602 m for the Blacksmith Fork and 0.0782 m for the Logan River. The results showed that multispectral imagery and 2-band LiDAR provided a cost-effective means for detailed remote sensing with each sensor providing complementary information for flood and river corridor assessment. Multispectral imagery supported interpretation of flood extent, vegetation condition, relative turbidity, and thermal patterns, while LiDAR captured terrain and surface features such as banks, levees, floodplain surfaces, channel modifications, and structures. The integrated datasets supported maximum flood extent mapping and flood-level estimation. These datasets can support reach-scale hydraulic modeling, catchment hydrology, river corridor ecology, floodplain conditions, and real-time monitoring of floods, in addition to quantification of flood hazards or post-flood impacts for municipalities and insurers. Full article

Remote sensing has become an important tool for crop monitoring and precision agriculture, yet its applications in sugar beet production remain fragmented across sensing platforms, target traits and modelling strategies. This review synthesises the development, current applications and future directions of remote sensing in sugar beet production, with particular attention to the transition from crop monitoring to precision management. A structured search was conducted in Scopus and the Web of Science Core Collection for publications from 2003 to 2025, and 181 relevant peer-reviewed articles were retained for thematic analysis. The literature shows a clear increase in sugar beet remote sensing studies, particularly after 2015, coinciding with the availability of Sentinel-2 imagery and, from 2016 onwards, the growing use of unmanned aerial vehicle-based sensing. It also indicates a gradual shift from crop mapping and canopy monitoring towards disease detection, weed mapping, yield prediction and management-oriented applications. Current studies demonstrate the value of satellite, unmanned aerial vehicle and proximal sensing for retrieving canopy traits, assessing biotic stresses, estimating root yield and supporting field-scale management. However, sugar beet presents specific challenges because its economic value depends not only on canopy development or root biomass, but also on sucrose concentration, recoverable sugar yield, and processing quality. These quality-related traits remain less studied and are difficult to infer directly from canopy observations. Modelling approaches have evolved from vegetation-index-based empirical models towards machine learning, deep learning, multi-temporal analysis, data fusion and crop model assimilation, but issues of model transferability, ground-truth availability and operational decision support remain unresolved. Future research should strengthen multi-source observations, external validation, quality-oriented prediction and decision-support workflows to promote robust, scalable and economically meaningful remote sensing applications in sugar beet production. Full article

Multi-sensor fusion of optical and synthetic aperture radar (SAR) imagery is widely used for land cover classification, yet most studies treat heterogeneous sensors as a uniform feature pool, leaving class-dependent differences in sensor contribution insufficiently quantified. To address this gap, this study integrates optical and SAR imagery within a Random Forest (RF) classifier in Google Earth Engine (GEE) and applies a combined SHapley Additive exPlanations (SHAP)—Remove and Retrain (ROAR)—bootstrap framework to disentangle, for each class, which sensor provides which discriminative information and how faithful those attributions are. The dataset included Sentinel-1/2, Korea Multi-Purpose Satellite (KOMPSAT)-3A/5, and Landsat-8, representing a range of spatial resolutions and spectral characteristics. An RF-based machine learning (ML) model was employed to perform multi-sensor data fusion and classification. To address the inherent opacity of ML models, we employed the SHAP algorithm, an explainable artificial intelligence (XAI) method, to interpret classification decisions. SHAP analysis indicated that visible and near-infrared (NIR) bands, along with vegetation indices, were the dominant contributors to land cover classification in this study area, while SAR data provided complementary structural information for spectrally ambiguous targets, such as roads—a class-dependent role reflected in the stable rank ordering of the Sentinel-1 VV contribution across bootstrap replications, rather than in a formally significant magnitude difference. ROAR results were consistent with the top-ranked SHAP features being those on which the classifier relies, supporting a physically plausible interpretation of ML-based remote sensing classification. These attributions were robust to estimator choice (SHAP-permutation ρ = 0.89) and spatial partitioning (mean ρ = 0.97 across folds). The contribution of this study is methodological rather than algorithmic: it provides an integrated analytical framework that applies SHAP, ROAR, and bootstrap confidence intervals jointly to a specific multi-sensor land cover problem, demonstrating that interpretability and high classification accuracy can be reported together. The results demonstrate that optical and SAR fusion contribute differently across land cover classes rather than uniformly, providing practical, class-specific guidance for sensor selection in operational land cover mapping and improving the interpretability of machine learning-based mapping workflows. Full article

The black soil region of Northeast China is a key grain-production area where cropland water erosion threatens soil fertility and sustainability. We diagnosed cropland soil loss across six diagnostic years/time slices (2001, 2005, 2010, 2015, 2020, and 2024) using a Revised Universal Soil Loss Equation (RUSLE)-based remote-sensing workflow implemented in Google Earth Engine (GEE). Rainfall erosivity was derived from Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) daily precipitation, soil erodibility from SoilGrids, topography from the Shuttle Radar Topography Mission digital elevation model (SRTM DEM), vegetation cover from the Landsat normalized difference vegetation index (NDVI), and cropland extent from ESA WorldCover; alternative rainfall sources, cropland masks, and P-factor settings were used for sensitivity analyses. Under the slope-graded P-factor scenario, mean annual soil loss ranged from 1.60 to 3.07 t ha⁻¹ yr⁻¹, and the proportion of cropland exceeding T = 2 t ha⁻¹ yr⁻¹ ranged from 25.2% to 52.9%. Soil loss fluctuated among years because rainfall erosivity and cover-management effects partly counteracted each other. Risk was concentrated in sloping piedmont and hilly cropland, whereas broad plains were dominated by very slight and slight erosion. P-factor parameterization represented the largest structural uncertainty. The workflow provides regional screening evidence for field verification and conservation-practice assessment, rather than direct site-specific engineering prescriptions. Full article

Understanding the long-term spatiotemporal dynamics of soil nutrients in land consolidation areas is essential for sustainable land management. In this study, the changes in soil organic matter (SOM), available nitrogen (AN), available phosphorus (AP), available potassium (AK), and soil pH from 1980 to 2023 in the land consolidation area of Anyi County, China, were investigated. A random forest (RF) model was developed using multisource environmental variables, including climate, topography, remote sensing, vegetation, parent material, distance to rivers, and land use. The results revealed three key findings. First, while land consolidation significantly increased overall soil nutrient levels, trends for individual nutrients were highly variable: AP and AK continuously increased over the 40-year period. SOM and AN increased after consolidation but decreased in later years. This process was accompanied by systematic soil acidification. Second, land consolidation substantially restructured the spatial patterns of soil nutrients. SOM and AN became more homogeneously distributed, whereas AP and AK showed increased variability. High-value nutrient areas shifted from contiguous expansion in the early stages to local aggregation and patchy fragmentation over the long term. Third, the RF model demonstrated high predictive accuracy (R² = 0.70 for pH, 0.68 for SOM, 0.63 for AP, and 0.54 for AK), with variable importance analysis identifying distinct environmental drivers for each nutrient. These findings provide a scientific basis for precision fertilization and evidence-based land consolidation planning. Full article