Search Results (860)

Accurate cloud detection is an important preprocessing step for subsequent remote sensing data processing. Traditional threshold cloud detection methods have a complex process and require a large number of threshold tests. In recent years, deep learning has been widely applied to cloud detection. However, annotating training datasets for deep learning models typically requires substantial human effort and time investment. Consequently, there are few existing manually annotated cloud detection datasets, and MODIS cloud detection datasets are particularly scarce. To overcome this limitation, we proposed a cloud detection method that combines radiative transfer simulations with deep learning. We first produced a simulated cloud detection dataset using a radiative transfer model and some existing remote sensing products, and then proposed a neural network for training the cloud detection model. Compared with other deep learning models for cloud detection, our method has achieved satisfactory results on the simulated dataset overall. Furthermore, we conducted cloud detection experiments on real satellite imagery. For comparative analysis, we trained other deep learning models on a real satellite image dataset and compared their performance with that of models trained on our simulated dataset. The cloud detection results on real satellite images demonstrate that the models trained on the simulated dataset we proposed achieve performance comparable to those trained on real remote sensing datasets. Specifically, for MODIS data, we compared our results with the official MODIS cloud mask product, MOD35. The results indicate that our method achieves lower false detection rates on mixed surfaces of snow and bare land. Full article

Increasing temperature and decreasing precipitation threaten the extent, persistence, and dynamics of snow across spatial scales, particularly ephemeral snow in Mediterranean mountain regions. This study estimates ephemeral snow cover and snow water equivalent (SWE) in the Sierra de San Pedro Mártir, Baja California, Mexico, using open-access datasets and remote sensing. Camera trap images and limited in situ data were used to calibrate the normalized difference snow index (NDSI) for snow detection and to estimate SWE and topographic effects on SWE from 2002 to 2023, encompassing wet, dry, and normal years. The optimal NDSI threshold for snow detection was 6.4 for MODIS Terra and 5.3 for MODIS Aqua, substantially lower than thresholds commonly reported for seasonal snowpacks in forested regions. In wet years, snowfall contributed up to 20% of annual precipitation, compared with ~13% in dry years. In normal years, the average SWE is 70 mm (24% of annual precipitation). SWE increased by 30% (91 mm) during wet years and decreased by 21% (55 mm) during dry years. Eastness (aspect) was the only statistically significant topographic predictor of SWE for MTerra, with higher SWE values observed on west-facing slopes. This study provides the first quantitative assessment of ephemeral SWE dynamics in a Mexican Mediterranean mountain system and establishes a framework for monitoring marginal snowpacks under increasing climatic variability. Full article

Agriculture is a major global source of methane (CH₄), and accurate emission estimates are essential for refining national greenhouse gas inventories and supporting climate-resilient policies. This study develops a high-resolution estimation framework for CH₄ emissions from Korean rice paddies by integrating multi-source datasets, including Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation indices, European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis Version 5 (ERA5)-Land meteorological variables, and Harmonized World Soil Database (HWSD) soil properties. Using CH₄ flux observations from four global rice ecosystems (Italy, Japan, South Korea, and USA), we constructed parallel daily and hourly machine learning models using an automated machine learning (AutoML) framework to compare their performance and process-level interpretability. The daily model demonstrated high predictive accuracy with correlation coefficients (CC) of 0.897 in 5-fold cross-validation and 0.819 in Leave-One-Year-Out (LOYO) cross-validation. Shapley Additive Explanations (SHAP) analysis revealed that while soil temperature is the dominant predictor for daily emissions (explaining ~50% of the variance), variable importance shifts significantly at finer resolutions. The hourly model exhibited a more complex multivariate structure. In this high-resolution context, although Normalized Difference Vegetation Index (NDVI) remains constant diurnally, its importance strengthens as a critical regulator of emission sensitivity, interacting with hourly meteorological fluctuations to capture short-term dynamics. The resulting 500 m daily gridded maps provide a robust foundation for national inventory refinement and spatially targeted mitigation planning. Our findings suggest that while the daily model offers optimal computational efficiency for long-term monitoring, the hourly model is superior for mechanistic understanding and detecting episodic emission events. This multi-resolution framework establishes an empirical basis for selecting appropriate temporal scales in operational greenhouse gas monitoring systems. Full article

Air pollution remains a major public health and environmental challenge in China, driven by complex non-linear interactions among anthropogenic activities, meteorological conditions, and topographic features that go beyond simple linear relationships. This study presents a comprehensive spatio-temporal assessment of key air pollutants (CO, NO₂, SO₂, and PM_2.5) and their relationships with Total Column Ozone (TCO) across China’s provinces from 2019 to 2023. Multi-source high-resolution satellite data from Sentinel-5P/TROPOMI, the China High PM_2.5 dataset, MODIS, and ERA5-Land reanalysis were integrated. A tiered analytical framework was applied, combining linear Pearson correlations, non-linear Spearman rank correlations, and interpretable XGBoost machine learning with SHAP values. Results reveal a distinct seasonal “seesaw” pattern, with primary pollutants peaking during winter stagnation and TCO reaching maximum levels in late winter and spring. Non-linear analyses uncover critical threshold effects, including exponential increases in PM_2.5 and SO₂ when surface temperatures drop below 0 °C, very strong SO₂-TCO coupling (ρ = 0.93), and significant pollutant trapping in low-elevation regions (CO-elevation ρ = −0.82). These findings support the development of precision environmental policies with dynamic, temperature-threshold-based emission controls and topography-specific strategies to effectively mitigate air pollution in China. Full article

Permafrost thickness serves as a critical indicator of hydrogeological conditions in cold regions and significantly influences the safety of engineering infrastructure. Due to the combined effects of climate, ecology, and human activities, the thermal characteristics and spatial distribution of permafrost in the Greater and Lesser Khingan Mountains of Northeast China exhibit high complexity, rendering existing permafrost thickness estimation methods largely inapplicable in this region. We developed an integrated estimation framework that bridges the gap between limited deep ground temperature measurements and regional-scale mapping. To overcome the scarcity of deep borehole (>20m) data, a physical-statistical inversion method was employed to derive permafrost base depths from shallow borehole temperature profiles, thereby expanding the foundational dataset to 104 representative sites. Integrating these ground observations with satellite-derived products (e.g., MODIS NDVI) and auxiliary environmental covariates (e.g., DEM-based topography and gridded climatic data), a Random Forest algorithm (RF) was applied to generate a 1 km-resolution permafrost thickness distribution map across Northeast China with a classification accuracy of 0.74. The results indicate that the average permafrost thickness in the study area is 47.71 ± 10 m, exhibiting a spatial pattern of thicker in the north and west, thinner in the south and east, and greater in mountainous areas than in plains. The top three influencing factors of permafrost thickness are atmospheric precipitation, surface thawing degree days (TDDs), and topographic position index (TPI), revealing that the thickness of discontinuous permafrost in northeastern China is primarily governed by local factors such as soil moisture, represented by the thick permafrost existed under a small patch of ground surface. This study provides a new methodological framework for estimating permafrost thickness in regions with limited ground temperature gradient measurement in deep boreholes. Full article

Droughts are one of the most significant hazards that affect human life due to the imbalanced distribution of water across the world. Some parts of the world are usually dry, and meteorological conditions affect these regions rapidly. In water-scarce regions, droughts significantly put at risk socio-economic stability and food security, which may cause a major challenge to sustainable development. Therefore, a precise definition of drought and the identification of early warning signals can help to minimize the negative effects of droughts, especially in terms of agriculture. In this study, drought signals of three major agricultural provinces of Turkey, namely Antalya, Şanlıurfa, and Konya, were investigated. For this purpose, the Standard Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), Evaporative Demand Drought Index (EDDI), and Vegetation Condition Index (VCI) were computed for each province. A composite score index was proposed for the evaluation of multiple indices together. All datasets were obtained from remote-sensing products to ensure reproducibility. A dataset for the 2003–2023 period was used. The monthly precipitation derived from CHIRPS data and potential evaporation (PEV) data were obtained from the ERA5-Land. Therefore, the SPEI and EDDI values were calculated by using ERA5-Land PEV values but not the evapotranspiration. The Normalized Difference Vegetation Index (NDVI) values for each province were obtained from the MODIS/Terra MOD13A3 v061. The Mann–Kendall test and Sen’s slope were applied to the computed time series to detect the trends. As a result, the dry and wet periods were identified for each province individually. The VCI was found to have an increasing trend for all tested provinces. Overall, from a future perspective, the most vulnerable province in terms of meteorological drought was indicated to be Antalya. Full article

Lightning-induced forest fires represent a dominant natural ignition source in boreal and temperate ecosystems, yet their climatic and biophysical controls remain poorly understood. This study investigates the spatiotemporal patterns and environmental drivers of 646 lightning-induced forest fires across the Da Hinggan Ling region, Northeast China, during 2001–2024. Multi-source datasets from ERA5-Land, MODIS, and ETCCDI were integrated to quantify short-term meteorological variability, vegetation water status, and long-term climatic extremes. Using Random Forest and XGBoost models combined with SHAP interpretability analysis, we identified key predictors and nonlinear responses of burned area to environmental forcing. Results reveal pronounced interannual fluctuations in fire activity, with 2010, 2016, and 2022 emerging as compound extreme years characterized by co-occurring drought and heatwaves. Vegetation moisture index (NDMI), diurnal temperature range (DTR), and heatwave duration (HWDI) were the most influential variables controlling burned area variability. The total burned area and fire duration showed significant declining trends, while high burned-area fires exhibited spatial clustering in dry, low-LAI regions. These findings demonstrate that compound dry–hot conditions coupled with vegetation desiccation are the primary drivers of large lightning fires. The study provides a process-based understanding of climate–fuel–fire linkages and supports improved fire risk forecasting under a warming climate. Full article

Environmental degradation in Iraq is a critical issue that requires strong monitoring. One indication of land degradation is a decrease in or loss of vegetation cover. This study examines changes in vegetation and productivity in the Thi-Qar region from 2001 to 2022, using the normalized difference vegetation index (NDVI) and net primary production (NPP), and their response to climatic and hydrological factors. To address the gap in assessments that simultaneously quantify the influence of streamflow, rainfall, and temperature across distinct land cover classes in arid and semi-arid regions, we developed a replicable multi-source geospatial framework. We used MODIS data within the Google Earth Engine platform to perform spatiotemporal analysis. We applied models to detect NDVI trends on a pixel-by-pixel basis. This study provides the first integrated, data-driven assessment of vegetation sensitivity to streamflow versus climate in the Thi-Qar Governorate using a harmonized multi-source dataset. This combines the FAO WaPOR NPP dataset with hydrological (streamflow) and climatic (CHIRPS rainfall, MODIS LST) variables within an analytical workflow to extract anthropogenic water management from climatic drivers. The results showed variations in the NDVI and productivity in the southern and southwestern regions, indicating areas of both degradation and improvement. The analysis found that 12% of the study area showed improvement, while 56.5% of the area showed degradation. Additionally, we classified the study area as either vegetation (cropland) or non-vegetation (fallow arable land, bare areas, and sand dunes). A multiple regression model was then applied to these categories to examine the relationships between streamflow, precipitation, land surface temperature (LST), and the NDVI. The multiple regression for the entire region showed that these factors explained 45.1% of NDVI variation, with streamflow being the most significant positive driver (p < 0.001). The result showed that the NDVI in cropland and arable land was strongly positively correlated with both precipitation and streamflow (R = 0.78, R = 0.75). In contrast, bare land and dunes showed weaker relationships (R = 0.26 and 0.51, respectively). Of these factors, streamflow had the most significant influence in explaining vegetation change (partial correlation p = 0.53), indicating the importance of human management in addition to climate. Full article

Air pollution remains a major public-health concern, and exposure to particulate matter (PM), particularly PM₁₀ (with a diameter ≤ 10 µm), is associated with adverse respiratory and cardiovascular outcomes. Most research relies on a singular model for PM₁₀ surface estimation. This study is an assessment of a national-scale, daily PM₁₀ estimation framework for Lithuania (2019–2024), using a hybrid machine-learning method that combines Random Forest (RF) and extreme gradient boosting (XGBoost) algorithms. Hourly PM₁₀ observations were aggregated from 18 monitoring stations to obtain daily means and temporal means. The predictors integrated meteorological factors, such as temperature, wind, humidity, and precipitation, to determine satellite-based atmospheric composition from Sentinel-5P Tropospheric Monitoring Instruments (TROPOMI). Atmospheric components include nitrogen dioxide (NO₂), carbon monoxide (CO), sulfur dioxide (SO₂), ozone (O₃), formaldehyde (HCHO), and the absorbing aerosol index (AI). Moderate-Resolution Imaging Spectroradiometers (MODIS) were used to record land-surface temperature and static spatial descriptors, such as elevation, land cover, Normalized Difference Vegetation Index (NDVI), population, and road proximity. The dataset was partitioned temporally into training (70%), validation (20%), and testing (10%). The hybrid model achieved an improved accuracy, compared with single-model baselines, reaching a coefficient of determination (R²) of 0.739 in validation and R² = 0.75 in the tested dataset. Mean absolute error (MAE) was 3.15 µg/m³, and root mean square error (RMSE) was 3.98 µg/m³. The results indicate a slight tendency to overestimate PM₁₀ concentrations at lower concentration levels. Feature-importance analysis revealed that short-term temporal persistence is the key to daily PM₁₀ prediction, while meteorological variables provide secondary contributions. Temporal evaluation, using consecutive two-year windows, revealed a consistent improvement in predictive performance from 2019–2020 to 2023–2024, while station-level analysis showed moderate-to-strong agreement between the predicted and observed PM₁₀ concentrations across monitoring stations, with R² ranging from 0.455 to 0.760. This provides decision-support capabilities for air-quality management, the evaluation of mitigation measures, and integration of air-pollution considerations into sustainable urban planning strategies assessing public-health protection. Full article

Twin wildfires burned over 9500 ha in Seferihisar, İzmir, western Türkiye, on 29—30 June 2025 under extreme fire weather conditions. This study reconstructs the spatiotemporal progression of the fires and examines the drivers of contrasting behaviors and burn severity. Multi-source datasets—Sentinel-2 imagery, VIIRS/MODIS thermal detections, MTG images and thermal detections, aerial photos, and ground data—were integrated to delineate progression polygons and compute rate of spread (ROS), fuel consumption (FC), and fire-line intensity (FI). Kuyucak fire showed rapid early growth, burning 3554 ha in 2.5 h (mean ROS of 5.0 km h⁻¹; mean FI of 37,789 kW m⁻¹), driven by strong northeasterly winds of 40–50 km h⁻¹, steep terrain, dense Pinus brutia fuels, and very low dead fine-fuel moisture (<6%). Kavakdere fire advanced more slowly (mean ROS of 1.6 km h⁻¹) across open grassland and cropland, yielding lower FC and FI. Synoptic analysis revealed a strong pressure-gradient-induced northeasterly wind regime linked to a mid-tropospheric geopotential height dipole between Central Europe and the Eastern Mediterranean, while WRF simulations indicated a dry boundary layer and enhanced low-level winds during peak spread. Sentinel-2 dNBR burn severity mapping showed substantial spatial variability tied to fuel and topography contrasts. Findings demonstrate how twin ignitions under similar weather conditions can produce divergent outcomes, underscoring the need for terrain- and fuel-aware strategies during extreme Mediterranean fire outbreaks. Full article

Gross Primary Productivity (GPP) in grassland ecosystems is a fundamental eco-biophysical indicator for assessing carbon cycling, grazing capacity, and ecosystem responses to climatic stress. However, robust estimation of GPP in arid and semi-arid rangelands remains challenging because of pronounced spatial heterogeneity, strong climate variability, and inherent uncertainties associated with remotely sensed observations. Together, these factors constrain both modeling performance and out-of-sample generalization beyond the training domain. In this dryland grassland context, this study compares the performance of machine learning (ML) models for grassland GPP proxy-based characterization, downscaling, and predictive agreement using a multivariate dataset that integrates Sentinel-2-derived spectral and phenological features, a Moderate-Resolution Imaging Spectroradiometer (MODIS)-derived GPP proxy, and complementary climatic and geographic information. Pixel-level observations spanning multiple years are analyzed, with ordinary linear regression used as a baseline benchmark and ensemble decision-tree models, including Random Forest, Gradient Boosting, and Histogram-based Gradient Boosting (HGB), compared. Instead of relying solely on random cross-validation, model performance is systematically assessed using a combination of spatially structured validation and a leave-one-year-out scheme to explicitly examine spatial and temporal generalization. The results indicate that ensemble tree-based models outperform linear approaches, with the HGB model showing the strongest agreement with the MODIS-derived GPP proxy (R² = 0.95, RMSE = 0.035 on the test set) and maintaining stable performance across spatial and temporal validations (R² = 0.86–0.96 across years). Taken together, the findings demonstrate that integrating multi-source remote sensing data with climatic information within a rigorous validation framework enables a more reliable assessment of model generalization and gap-filling consistency with respect to a remote-sensing-based proxy target, rather than an absolute validation against ground-based measurements, thereby supporting sustainability-relevant monitoring of arid grassland ecosystems. Full article

This study aims to systematically review the existing literature on the use of data derived from remote sensing products to estimate groundwater recharge. The terms “recharge”, “remote sensing product data”, “remote sensing data”, “groundwater”, and “recharge estimation” were used as keywords in the Web of Science and Scopus databases. A total of 27 articles were analyzed, highlighting the use of different precipitation and evapotranspiration products for estimating potential recharge. This review emphasizes the potential of products such as CHIRPS and TRMM for precipitation and MODIS for evapotranspiration, as well as other remote sensing datasets that have shown good performance in their applications. The studies demonstrate the high feasibility of applying remote sensing to estimate groundwater recharge and indicate how its use can enhance the quality and reliability of the results obtained. Full article

Aerosol dry deposition is an important sink for particulate matter and a source of uncertainty in air quality modeling. Using the Weather Research and Forecasting model coupled with CUACE (WRF-CUACE), we quantified how three aerosol dry deposition schemes and satellite-based leaf area index (LAI) information affected PM_2.5 dry removal and near-surface PM_2.5 over central and eastern China in January 2022. The schemes were abbreviated as Z01, E20, and PZ10, respectively. A fourth simulation (PZ10_MLAI) used PZ10 but replaced the baseline LAI dataset with a Moderate Resolution Imaging Spectroradiometer (MODIS) constrained LAI field. Hourly PM_2.5 was evaluated with the China National Environmental Monitoring Center network. The schemes produced pronounced, size-dependent differences in deposition velocities, with a pronounced spread in the 0 to 2.5 µm average and more than one order of magnitude spread in the accumulation mode diagnostic, leading to distinct regional mean PM_2.5 dry deposition fluxes. The mean PM_2.5 flux increased by 5.9% in E20 relative to Z01 and decreased by 54.4% in PZ10. The MODIS LAI adjustment changed the PZ10 mean flux by 0.42%. The flux contrasts yielded coherent PM_2.5 responses, with E20 reducing near-surface concentrations by about 10 to 30% and PZ10 increasing them by about 20 to 60%, reaching about 80 to 100% in parts of southern China. Domain mean correlations ranged from 0.61 to 0.65 and PZ10-based simulations exhibited near-zero mean bias. Although MODIS LAI effects were modest for this winter month, local PM_2.5 differences commonly remained within about 4% and approached 6 to 10%, indicating that satellite LAI constraints can be important for multi-year and decadal applications. Full article

Evapotranspiration (ET) links the water cycle with the energy balance and serves as a key driving process for ecosystem functioning and water resource management. Canopy conductance (Gc) plays a central role in regulating transpiration, but many models inadequately represent its regulatory mechanisms and show varying applicability across different land cover types. This study develops a remote-sensing ET estimation approach suitable for large scales and diverse land cover types and proposes an improved canopy conductance model for daily latent heat flux (LE) estimation. By integrating the canopy radiation transfer concept from the K95 model into the multiplicative Jarvis framework, an improved canopy conductance model is developed that includes limiting effects from photosynthetically active radiation (PAR), vapor pressure deficit (VPD), air temperature (T), and soil moisture (θ). Eighteen combinations of limiting functions are designed to evaluate structural performance differences. Using observations from 79 global flux sites during 2015–2023 and integrating multi-source datasets, including ERA5, MODIS, and SMAP, a two-stage parameter optimization was applied to determine the optimal limiting function combination for each land cover type. And nine sites from nine different land cover types were selected for independent spatial validation. Temporal validation within the optimization sites shows that, at the daily scale, the model achieves a Kling–Gupta efficiency (KGE) of 0.82, a correlation coefficient (R) of 0.82, and a Root Mean Square Error (RMSE) of 27.83 W/m², demonstrating strong temporal stability. Spatial validation over independent holdout sites achieved KGE = 0.84, R = 0.84, and RMSE = 22.53 W/m². At the 8-day scale, when evaluated over the holdout sites, the model achieves KGE = 0.87, R = 0.88, and RMSE = 18.74 W/m². Compared with the K95 and Jarvis models, KGE increases by about 34% and 15%, while RMSE decreases by about 38% and 12%, respectively. Relative to the MOD16 and PML-V2 products, KGE increases by about 32% and 16%, while RMSE decreases by about 33% and 17%, respectively. Comprehensive comparisons show that explicitly coupling canopy structure with multiple environmental constraints within the Jarvis framework, together with structure optimization across land cover types, can markedly improve large-scale remote-sensing ET retrieval accuracy while maintaining physical consistency and physiological rationality. This provides an effective pathway and parameterization scheme for producing ET products applicable across ecosystems. Full article

Understanding the coupling mechanisms between vegetation phenology and carbon productivity is essential for assessing ecosystem responses to climate change and guiding sustainable grassland management. This study focuses on stable alpine grasslands on the southern slope of the Qilian Mountains from 2001 to 2020, a climatically sensitive but relatively under-investigated transition zone on the northeastern Tibetan Plateau. We utilized MODIS NDVI time-series (MOD13Q1) and the latest PML V2 gross primary productivity (GPP) product at 500 m resolution to quantify changes in the start (SOS), end (EOS), and length (LOS) of the growing season. A pixel-wise linear regression approach was applied to evaluate the sensitivity of GPP to phenological metrics, explicitly characterizing how much GPP changes in response to unit shifts in SOS, EOS and LOS. Compared with previous studies that mainly described large-scale correlations between phenology and GPP or relied on coarser GPP products, this study provides a pixel-level, sensitivity-based assessment of phenology–carbon coupling in alpine grasslands using a long-term, phenology–GPP dataset tailored to the Qilian alpine region. The results revealed trends of earlier SOS, delayed EOS, and extended LOS, accompanied by a gradual increase in GPP. However, phenology–GPP coupling exhibited notable spatial heterogeneity. In mid- and low-altitude areas, extended growing seasons enhanced GPP, whereas high-altitude zones showed limited or even negative responses, likely due to climatic constraints such as cold stress and thermal–moisture mismatches. To better understand these spatial differences, we constructed a three-dimensional phenology–GPP sensitivity space and applied k-means clustering to delineate three ecological functional zones: (1) high carbon sink potential, (2) ecologically fragile regions, and (3) neutral buffers. This sensitivity-based functional zonation moves beyond traditional correlation analyses and provides a process-oriented and spatially explicit framework for ecosystem service assessment, carbon sink enhancement and adaptive land-use strategies in sensitive mountain environments. Full article