Search Results (417)

Mountainous watersheds often suffer from incomplete and spatially biased landslide inventories, which limit the reliability of conventional susceptibility modelling. This study examines whether the erosion coefficient Z of the Gavrilović Erosion Potential Method can provide a process-oriented indicator of slope-instability predisposition in the Portaikos watershed, Central Greece. The Z coefficient was derived from geospatial layers representing vegetation protection, lithological erodibility, erosion-process expression and slope gradient, using Copernicus land-cover products, tree-cover density data, Sentinel-2 imagery, FABDEM and national soil–geological information. A landslide inventory of 46 mapped occurrences from the Hellenic Survey of Geology and Mineral Exploration was then used as an independent reference layer. Erosion severity was classified into five classes and compared with the landslide distribution through Frequency Ratio analysis. Most of the basin was assigned to moderate, very slight and slight erosion classes, covering 34.7%, 30.1% and 28.2% of the area, respectively. By contrast, severe and excessive erosion occupied only 6.7% and 0.4% of the watershed, but contained a much larger proportion of the mapped landslides: 58.7% and 10.9%, respectively. This disproportion was also reflected in the Frequency Ratio analysis. When the severe and excessive classes were considered together, they occupied approximately 7.1% of the watershed but contained 69.6% of the mapped landslides, corresponding to an FR value of 9.79. The separate excessive class showed the highest FR, but it was interpreted cautiously because of its very limited spatial extent and small landslide count. These results indicate that high Z values coincide with terrain sectors where lithological weakness, steep slopes, reduced surface protection and erosion-related sediment-source conditions jointly favour slope instability. The Gavrilović Z coefficient should therefore not be interpreted as a substitute for rainfall-threshold analysis or inventory-based predictive models. Rather, it may serve as a useful first-order screening layer for field verification, spatial prioritization and ecosystem-based mitigation planning in data-scarce Mediterranean mountain watersheds. Full article

Hainan Tropical Rainforest National Park is a prime example of a “continental island” tropical rainforest and holds significant value for biodiversity conservation. However, human activities have led to frequent changes in land use and increased habitat fragmentation within the park; a precise analysis of the underlying mechanisms is necessary for ecological restoration. Consequently, drawing upon land-use data from 2000 to 2020, this study coupled multi-dimensional fragmentation metrics (CFI, AFI, and SFI) with the XGBoost-SHAP framework to systematically unravel the spatiotemporal dynamics and underlying driving mechanisms of landscape fragmentation in Hainan Tropical Rainforest National Park. Our findings revealed that the spatial configuration of fragmentation predominantly propagated along river networks and transport corridors, accompanied by a fluctuating ‘decline–rise–decline’ temporal trajectory. Notably, the XGBoost-SHAP attribution highlighted a distinct temporal shift in the dominant drivers: fragmentation was primarily mitigated (negatively driven) by NDVI between 2000 and 2010 but was subsequently exacerbated (positively driven) by GDP growth from 2010 to 2020. The findings of this study provide data support and a scientific basis for ecosystem restoration and land use planning in Hainan Tropical Rainforest National Park. Full article

In China’s national parks, Tianchuang were located on land within boundaries that are excluded from national park administration to reduce governance conflicts. According to existing research on boundary optimization of protected area, Tianchuang are routinely regarded as fixed constraints or contextual backgrounds rather than optimization targets. Based on systematic conservation planning (SCP) theory and integrated models of MaxEnt, InVEST, and Marxan, and taking Tianchuang as dynamically adjustable spatial variables, this study aims to explore the boundary optimization schemes of Tianchuang areas in Wuyishan National Park of China. The main findings are as follows: Firstly, highly suitable habitats for representative species are predominantly concentrated within the north-central region of the park, while Tianchuang patches generate significant edge effects and pose critical disturbances to ecosystem integrity. Secondly, human activity intensity demonstrates marked spatial heterogeneity, indicating distinct variations in socioeconomic impacts and management difficulties among Tianchuang patches. Thirdly, based on irreplaceability, landscape connectivity, and patch fragmentation, three optimization strategies are proposed, namely incorporation, retention, and provisional retention, balancing ecological protection and community development. This study provides a scientific basis for the differentiated management of Tianchuang areas in Wuyishan National Park and offers a practical pathway for the boundary optimization and spatial governance of protected areas in China. Full article

Coastal cities are highly sensitive and vulnerable to climate change risks. A scientifically grounded assessment of urban climate change risk and its driving mechanisms is essential for strengthening urban climate adaptation capacity and supporting sustainable development. Taking Ningbo as the study area, this paper constructs a risk assessment system comprising five categories of extreme climate indicators, namely heat, rainstorm, drought, humidity, and strong wind, based on the China Surface Climate Normals Dataset for 1981–2010 and meteorological observations from the National Centers for Environmental Information (NCEI) for 2015–2024. Using 30 m resolution land-use data for 2023, three land-use sensitivity indicators are extracted: the proportion of built-up land, the proportion of green space and forest land, and the proportion of water area. The CRITIC objective weighting method is then applied to construct an integrated climate change risk index and identify the spatial pattern of climate change risk in Ningbo. On this basis, the high-risk area identification performance of Logistic Regression, Random Forest, and XGBoost is compared. The optimal XGBoost model is selected and combined with the SHAP method to systematically reveal the direction, relative importance, and nonlinear threshold relationships through which land-use characteristics affect the formation of high-risk areas. The results show that urban climate change risk in Ningbo exhibits a pronounced spatial differentiation pattern, with higher risk in the northeastern coastal and central–eastern areas and lower risk in the western and southwestern areas. Insufficient green space and forest land buffering is the most important factor affecting the formation of high-risk areas. All three land-use variables have clear nonlinear thresholds. The critical turning points for identifying high-risk areas are 20.0% built-up land, 2.0% green space and forest land, and whether there is a water body or not. Significant interaction effects are observed among land-use variables, among which the interaction between built-up land and green space/forest land is the most prominent. These findings provide methodological support and empirical evidence for climate change risk assessment and climate-adaptive spatial planning regulation in coastal cities. Full article

Flooding is one of the most frequent and damaging natural disasters, accounting for nearly half of global disasters and posing a major challenge in Indonesia, where floods represent approximately 77% of all nationally recorded disaster events. Rapid urbanisation, land-use change, and climate-induced extreme rainfall have intensified flood risks nationwide. However, existing vulnerability assessments remain fragmented and localised, limiting their relevance for national-scale adaptation planning. This study develops a measurable and explainable framework for assessing urban flood vulnerability across Indonesia using cloud-based geospatial data and interpretable machine learning. The approach integrates CEMS-GLOFAS (flood hazard), WorldPop (population exposure), SRTM (topography), and ESA WorldCover (land cover) datasets within Google Earth Engine (GEE). Flood vulnerability is quantified through a modified Flood Vulnerability Index (FVI) combining hazard, exposure, and physical vulnerability components. The Extreme Gradient Boosting (XGBoost) model predicts FVI values, while SHapley Additive exPlanations (SHAP) and Partial Dependence Plots (PDPs) enhance model transparency and identify the influence of key variables such as flood depth, population density, and elevation. The model achieved high predictive accuracy (R² = 0.89; RMSE = 0.04728 FVI units, dimensionless) and revealed substantial spatial heterogeneity across 514 districts, with the highest FVI (0.75–0.85) in Banda Aceh, Mojokerto, Pasuruan, Samarinda, and Merauke. The integration of GEE and explainable AI offers a transparent, scalable framework to support data-driven flood risk mitigation and urban climate resilience in Indonesia. Full article

This study evaluates the resource, technical, economic, and CO₂ mitigation potentials of terrestrial hydropower, wind power, and photovoltaic (PV) power in China under historical and future SSP(Shared Socioeconomic Pathways) climate scenarios. By integrating hydro-meteorological observations, land-use information, digital elevation data, nature-reserve constraints, and CMIP6 climate outputs, we estimate renewable-energy potentials through a consistent national-scale screening framework and cost–supply curve analysis. The results show clear spatial heterogeneity among the three energy sources. Hydropower potential is concentrated mainly in the Yangtze River basin, Pearl River basin, and Southwestern International Rivers. Wind-power potential is relatively high in northwestern, northeastern, and plateau regions, while PV potential is particularly large in northwestern, northern, northeastern, and selected southeastern regions. Under the adopted assumptions, PV shows the largest resource and technical potential, followed by wind power and hydropower; however, this ranking reflects resource potential rather than comprehensive deployment superiority. Practical development is also constrained by ecological flow requirements, land-use competition, grid integration, storage demand, transmission capacity, curtailment risk, and regional demand matching. The findings provide a national-scale comparative reference for renewable-energy planning and CO₂ mitigation, while highlighting the need for future work that incorporates dynamic land use, system-level integration costs, detailed turbine or power-curve modeling, and dynamic grid-emission factors. Full article

The Curve Number (CN) remains one of the most widely applied parameters for estimating direct surface runoff. However, its conventional application based on watershed-aggregated tabulated values conceals hydrological variability in regions with contrasting soils and steep topographic gradients. A recurring limitation of distributed CN approaches is the absence of independent hydrological validation; most machine learning models are trained and evaluated against the same SCS-USDA lookup values used to construct the training target, a circular scheme that measures statistical agreement rather than physical credibility. This study develops a reproducible geospatial workflow for distributed CN estimation across continental Ecuador, combining official MAG land use, soil surface texture natural drainage, and topographic slope layers at 1:25,000 scale with a Random Forest regression model at 10 m spatial resolution. The CN reference raster was derived from official geo-pedological layers and independently validated, not against tabulated assumptions, but against observed hydrological behaviour. Field hydraulic characterization across four dominant land cover classes in the Guamote microwatershed (Chimborazo Province), combined with HEC-HMS (US Army Corps of Engineers, Davis, CA, USA) rainfall-runoff modelling over 41 years (1981–2021), confirmed a mean annual discharge of 0.1568 m³ s⁻¹ consistent with the tabulated CN assignments. To our knowledge, this is the first nationally distributed CN map with field-anchored hydrological benchmarking for an Andean country. The Random Forest model achieved an RMSE = 10.4, an R² = 0.42, and an NSE = 0.41, a performance consistent with published field-based CN estimation studies and expected given the inherent scatter of the SCS-USDA method under real-world conditions. Zonal CN comparisons confirmed a mean absolute error below 5 CN units across the Andean highland and Amazon watersheds; the Guamote watershed showed a mean ∆CN below 4 units against the field-calibrated model. Land use and surface texture emerged as the dominant CN predictors, with natural drainage providing critical discrimination in volcanic and poorly drained soil environments. The resulting 10 m national CN map offers a physically grounded, spatially explicit parameterization layer for distributed hydrological modeling and water resources planning across data-scarce Andean and tropical territories, with direct relevance for flood risk screening, irrigation planning, watershed conservation, and climate adaptation under SDG 6, SDG 11, SDG 13 and SDG 15. Full article

As a core component of terrestrial ecosystems, forests play an irreplaceable ecological role in carbon sequestration, biodiversity conservation, and global climate regulation. Home to key global forest belts including the Congo Basin, the African continent’s forest changes directly shape regional ecological balance and sustainable development while profoundly affecting global ecological security and climate dynamics. Based on the Shared Socioeconomic Pathways (SSPs), a unified narrative framework for global socioeconomic and environmental change scenarios, this study couples techniques such as the Future Land Use Simulation (FLUS) model, dynamic degree analysis, transition matrix, K-means clustering analysis, and patch fragmentation analysis. This work aims to answer two key questions: (1) What are the spatiotemporal characteristics and dominant drivers of African woodland changes under different SSPs? (2) How do spatial clustering and fragmentation patterns vary across scenarios? It systematically predicts and analyzes the spatiotemporal characteristics, driving mechanisms, and fragmentation change patterns of African woodlands in 2030, 2050, and 2070 under five scenarios (SSP1-SSP5) with 2020 as the baseline. These five official IPCC SSP frameworks represent five distinctly divergent socioeconomic development trajectories ranging from sustainable to fossil-fuel-driven development, which are the core differentiated scenarios recommended by IPCC; full inclusion facilitates systematic comparison of varied forest feedback features across Africa’s diversified national development backgrounds. The research results show that understory forests in the SSP5 (Fossil Fuel-dominated Development) scenario exhibit a stable growth trend, with the total area transferred in significantly exceeding the area transferred out from 2020 to 2070, resulting in a net increase of 143,513 km². This growth occurs because high-income economies under this scenario invest heavily in ecological restoration and forest protection, offsetting carbon-intensive development impacts. The core forest density continues to increase and is distributed in contiguous areas; the SSP4 (uneven development) scenario regarding forest degradation is the most severe, with the dynamic rate expected to drop to −0.05% between 2050 and 2070, and a net transfer of −265,581 km². Forest fragmentation is highest, and the core density area is gradually shrinking. Cluster analysis shows that forest area remains relatively stable in most African countries, with stable countries accounting for as much as 95.49% under scenario SSP5. Regions with woodland expansion are mainly distributed in North Africa and localized parts of Southern Africa. After refinement using independent tree-density evidence, woodland expansion in South Africa is shown to be more limited and spatially heterogeneous; these newly expanded woodlands are mostly artificial plantations and alien invasive tree stands rather than native natural woodlands, mainly occurring in eastern and southeastern areas rather than in arid western regions. The spatiotemporal transfer process exhibits significant periodic differentiation, with 2030–2050 being a critical transitional period for forest change, and the differentiation effect between scenarios intensifying. Fragmentation analysis indicates that scenario SSP3 (regional rivalry, with moderate population growth and weak policy constraints) has the best forest integration and the lowest degree of fragmentation, while scenario SSP4 is most strongly affected by human activities and has the highest risk of patch fragmentation. These findings can provide a scientific basis for African countries to formulate differentiated forest protection policies and optimize ecological restoration plans, while also offering theoretical insights for continental-scale forest ecological management. Full article

Standardized drought indices such as the Standardized Precipitation Index (SPI) are widely used in both monitoring and climate-change impact assessments. However, SPI values are not uniquely defined unless the reference climatology used for standardization is explicitly stated and justified−a methodological issue that becomes critical under non-stationary climate conditions. Here, we present a methodological assessment of how reference-climatology strategy affects SPI-based drought projections under climate change, using Türkiye’s 26 major basins as a hydroclimatically diverse testbed. These assessments inform sustainable water resources planning, agricultural adaptation, and climate-resilient infrastructure design under non-stationary climate. Daily precipitation projections from 56 GCM-RCM pairs (EURO-CORDEX EUR-11, 0.11° (approximately 12 km at the mid-latitudes of the study domain); CMIP5 RCP8.5) were bias-corrected against ERA5-Land and aggregated to basin means. We computed SPI-9 and compared two commonly used reference strategies: (i) a fixed historical baseline (1970–2005), applied consistently to both historical and future periods (fixed-baseline SPI); and (ii) a period-specific baseline (period-specific SPI; future SPI values are standardized to the climatology of the future evaluation period itself). Using the same climate simulations, the two strategies yield markedly different drought projections. At the country scale, end-of-century drought time reaches 458 months under the fixed-baseline strategy, whereas the period-specific strategy indicates 393 drought months. Corresponding severity summaries are likewise stronger under fixed-baseline standardization. The contrast is even stronger in several Mediterranean basins, where fixed-baseline standardization produces persistently severe drought conditions. These results show that SPI-based drought projections are substantially sensitive to the choice of reference-climatology strategy, and that the same climate ensemble can support materially different drought narratives depending on how anomalies are standardized. Because the two strategies differ in both reference-timing and calibration-window length (36 versus 95 years), the headline contrast should be interpreted as a combined effect rather than as a pure baseline-timing result. In the present implementation, the period-specific strategy uses a single future calibration period (2006–2100), so the comparison should be interpreted as a stress test of reference framing under non-stationary climate rather than as an equal-length baseline experiment. An equal-length late-baseline sensitivity check (1970–2005 versus 2065–2100; both spanning 36 years) shows that the fixed-to-late-baseline contrast is larger than the fixed-to-period-specific contrast in 25 of 27 spatial units, including a 3.0-fold amplification at the national scale, indicating that the reference-timing effect persists when calibration-window length is held constant. Because the analysis is based on a CMIP5-driven RCP8.5 ensemble, the numerical projections should be interpreted as a high-end stress-test envelope rather than as the most likely outcome. We therefore recommend that drought projection studies explicitly report the reference-climatology strategy, justify the calibration window, and distinguish between analyses designed to quantify change relative to a historical climate and analyses designed to describe anomalies relative to an evolving future climate. These methodological choices have direct implications for sustainable water resources management and drought-risk preparedness in water-stressed Mediterranean systems, and contribute to broader sustainability targets such as Sustainable Development Goal 6 (Clean Water and Sanitation), SDG 13 (Climate Action), and SDG 15 (Life on Land). Full article

Urban canopy mitigates urban heat, yet how the spatial configuration of high- and low-canopy layers shapes population heat exposure across a national urban system remains insufficiently understood. Drawing on a panel of 369 Chinese prefecture-level cities for 2000, 2005, 2010, 2015, and 2020, this study constructs a population-weighted thermal-exposure metric—the Human Heat Exposure Index (HEI)—and stratifies urban vegetation into high- and low-canopy classes based on Chinese Land Cover Dataset (CLCD) land-cover types. Multiscale Geographically Weighted Regression (MGWR) and Extreme Gradient Boosting (XGBoost) with SHapley Additive exPlanations (SHAP)-based interpretation are combined to identify spatially varying associations and nonlinear marginal effects of stratified canopy patterns on HEI. HEI shows a persistent south–high, north–low spatial structure, with Global Moran’s I stable at approximately 0.85 throughout the study period. High-canopy edge density and cohesion are increasingly associated with reduced heat exposure in densely built regions, while low-canopy mean patch area and edge density retain explanatory power across all years through near-surface evapotranspirative regulation. The marginal cooling effect of vegetation strengthens appreciably only above an Normalized Difference Vegetation Index (NDVI) of approximately 0.6, and the apparent inflection ranges for impervious surface proportion and standardized solar radiation lie near 25% and 0.4, respectively. These findings suggest that in cities with high impervious loads, cooling-network connectivity and within-zone canopy configuration matter more than additional canopy area alone, and that planning targets should be calibrated to climate zone, city type, and existing surface conditions. Full article

Against the backdrop of coordinated development in the Beijing–Tianjin–Hebei region, Hebei Province serves as an ecological safety barrier for the Beijing–Tianjin–Hebei urban agglomeration. Conducting research on land use and land cover change (LUCC) and ecosystem service value (ESV) holds significant theoretical and practical value for elucidating the mechanisms underlying ESV evolution under the combined effects of rapid urbanization and major ecological engineering projects, and for applying these findings to regional land-use planning and ecological conservation and restoration efforts. This research aligns with the United Nations Decade on Ecosystem Restoration (2020–2030). Based on land-use data from 2000, 2010, and 2020, along with 11 categories of natural and socio-economic drivers, this study systematically analyses regional LUCC and calculates ESV using locally adjusted equivalence factors. It examines the spatiotemporal evolution patterns of ESV through the analysis of local spatial autocorrelation indices (LISAs), centroid, and standard deviation ellipses, and employs a GeoDetector to measure ESV drivers. Three scenarios—a natural evolution scenario (NES), economic development scenario (EDS), and ecological protection scenario (EPS)—were established. The patch-generating Land use simulation (PLUS) model was employed to simulate LUCC for 2030 (Kappa = 0.840) and calculate ESV. Results show that from 2000 to 2020, forest land and impervious surfaces in Hebei Province continued to expand, while cropland and grassland decreased. The cumulative ESV increased by 4.85 billion yuan. Slope was the primary driver of spatial variation in ESV, and the interaction between natural and socioeconomic factors demonstrated significantly stronger explanatory power. In 2030, the total ESV under all three scenarios was lower than in 2020. The EPS reached an ESV of 344.72 billion yuan, representing a relatively suitable model that balances development and conservation. Full article

Low-carbon development (LCD) and intensive urban land use (IULU) are critical objectives for sustainable urban development. Existing studies have usually evaluated LCD or IULU separately, whereas the dynamic relationship between carbon-transition capacity and land-use intensification under rapid urbanization remains insufficiently clarified. This gap limits the ability of policymakers to design spatially differentiated and synergistic actions for achieving the Sustainable Development Goals (SDGs). This study investigates the relationship between LCD and IULU and its transformation within the sustainable development framework, using the Middle Reaches of the Yangtze River Urban Agglomeration (MRYRUA) in central China as a case study. Results indicate a strong positive correlation between LCD and IULU. Crucially, their coupling exhibited a distinct U-shape trajectory from 2005 to 2020; it decreased from 0.89 in 2005 to 0.73 in 2013 and then recovered to 0.84 in 2020, suggesting a relative weakening of the interaction followed by recoupling rather than complete decoupling. The identified U-shaped trajectory holds vital implications for other developing nations, suggesting that integrating low-carbon goals into spatial planning and land policies from the early stages of urbanization can pave the way for a faster transition to a green, intensive, and high-quality development model. Moreover, although both LCD and IULU exhibited positive trends, a widening gap was observed between provincial capitals and non-provincial cities. We, therefore, recommend integrating multi-stakeholder collaboration and implementing differentiated strategies to enhance the synergistic effects of LCD and IULU for cities at different phases of the LCD–IULU transition. Full article

Flood susceptibility in Pakistan is strongly influenced by hydro-climatic variability, land-surface conditions, topography, and recurrent floodplain exposure; however, national-scale studies often lack a comprehensive assessment that captures both spatial patterns and temporal flood-risk dynamics within a single framework. This study is one of Pakistan’s first national efforts to address the gap between flood risk assessment and prioritization through a unified geospatial assessment. This study assesses flood susceptibility across Pakistan for 2002, 2012, and 2022 using a GIS-based AHP approach by integrating climatic, environmental, topographic, hydrological, soil, LULC, and anthropogenic indicators. The study results were further analyzed through district-level assessments, risk change analysis, persistence mapping, LULC exposure assessments, and the Comprehensive Flood Risk Priority Index (FRPI). The results show that high and very high flood susceptibility zones are primarily concentrated along the Indus River corridor, lower floodplains, and coastal Sindh, accounting for more than 7% of the total land area of Pakistan. Persistent flood hotspots are identified in Rann of Kutch (66.6%), Jacobabad (65.0%), and Jafarabad (61.1%), indicating strong temporal stability of flood-prone conditions. LULC exposure analysis reveals that cropland is the dominant exposed class, with the highest district-level exposure observed in Badin (17.1%) and Larkana (10.1%). The FRPI further identifies priority flood-risk zones where susceptibility, persistence, risk change, and exposure converge, with the highest FRPI values observed in Jacobabad (0.742), Rann of Kutch (0.738), and Badin (0.711). Model validation demonstrates strong predictive performance, with susceptibility ROC-AUC values ranging from 0.85 to 0.87 and FRPI AUC reaching 0.85. The proposed framework provides a robust decision-support tool for targeted flood-risk management and climate-resilient land-use planning in Pakistan. Full article

This study presents a multi-indicator remote sensing framework for assessing satellite-derived water-quality-related and trophic-state-related dynamics across four freshwater systems in Türkiye Egirdir Lake, Sapanca Lake, Catalan Dam, and Yuvacik Dam between the baseline (2015–2018) and recent (2023–2025) periods. Rather than providing a regulatory or use-specific satellite-based assessment of water-quality-related indicators, the study evaluates optically and thermally detectable surface water indicators derived from Sentinel-2 MSI and Landsat 8/9 imagery processed in Google Earth Engine. The Normalized Difference Chlorophyll Index (NDCI), the Normalized Difference Turbidity Index (NDTI), and land surface temperature (LST, applied to water surfaces) were used to detect change patterns through period-mean difference mapping (Δ-mask) and interannual time series analysis. Results reveal distinct spatial and temporal dynamics broadly consistent with the interplay of climatic, hydrological, and anthropogenic drivers. In the southern Mediterranean systems, positive ΔNDCI anomalies in littoral and inflow zones were associated with increasing summer LST, with Egirdir Lake exhibiting a statistically significant warming trend of +0.170 °C yr⁻¹ (Mann–Kendall τ = 0.53, p = 0.029), interpreted cautiously as a physically plausible signal consistent with regional climate trends, suggesting elevated thermally mediated eutrophication-related optical risk. In the northern Marmara systems, satellite-observed patterns were more strongly associated with anthropogenic nutrient loading and morphological constraints, with turbidity-related optical increases concentrated in western and marginal zones despite relatively stable thermal conditions. As concurrent in situ measurements were unavailable, cross-sensor consistency checks and literature-based benchmarking were applied as alternative validation strategies. Across all four systems, positive ΔNDCI anomalies were systematically concentrated in shallow marginal and inflow zones, while ΔNDTI patterns varied by system, underscoring the role of littoral dynamics as early indicators of optically detectable water-quality deterioration and trophic-state-related change. The proposed framework offers a scalable, cost-effective approach for freshwater quality surveillance in data-scarce environments and provides direct support for integrated water resource management under Türkiye’s National Water Plan (2026–2036). Full article

Accurate and scalable urban land cover mapping is critical for sustainable urban planning and environmental management. While deep learning models offer powerful tools for this task, their performance is often constrained by the need for vast, manually labeled datasets, which are costly and challenging to acquire for diverse urban environments. To address this limitation, we propose the Cross-Scale U-Net, an original, highly adaptable operational framework that systematically exploits the inherent scale effects of remote-sensing imagery to optimize transfer learning. By operationalizing prior theoretical findings on receptive fields, this workflow provides an actionable method for users to manipulate spatial resolution, identify an optimal scale to bridge the domain gap, and subsequently automate feature extraction with significantly reduced manual effort. Using the well-annotated ISPRS Potsdam dataset as the source domain, our framework transfers learned knowledge to classify National Agriculture Imagery Program (NAIP) data from Phoenix, AZ (2015), into four primary land cover classes. We systematically evaluated the framework’s performance across spatial resolutions ranging from 15 cm to 100 cm, achieving a peak overall accuracy (OA) of 82.45%. To assess generalizability, the model was applied in a label-free transfer scenario to NAIP imagery from Las Vegas, NV (2015), and Phoenix, AZ (2013 and 2019), consistently delivering OA values above 70%. In a comparative analysis, the Cross-Scale U-Net significantly outperformed traditional classification techniques. While our current empirical validation is focused on arid urban environments due to experimental constraints, the framework introduces a highly flexible, actionable scale-adjustment process. This approach offers a scalable workflow that can be tailored to various landscape scales—such as expanding to coarser resolutions for large-scale forests or protected areas—delivering high-fidelity maps while mitigating data scarcity. Full article