Search Results (135)

High-precision semantic segmentation of remote sensing imagery is crucial in geospatial analysis. It plays an immeasurable role in fields such as urban governance, environmental monitoring, and natural resource management. However, when confronted with complex objects (such as winding roads and dispersed buildings), existing semantic segmentation methods still suffer from inadequate target recognition capabilities and multi-scale representation issues. This paper proposes a neural network model, LMVMamba (LoRA Multi-scale Vision Mamba), for semantic segmentation of remote sensing images. This model integrates the advantages of convolutional neural networks (CNNs), Transformers, and state-space models (Mamba) with a multi-scale feature fusion strategy. It simultaneously captures global contextual information and fine-grained local features. Specifically, in the encoder stage, the ResT Transformer serves as the backbone network, employing a LoRA fine-tuning strategy to effectively enhance model accuracy by training only the introduced low-rank matrix pairs. The extracted features are then passed to the decoder, where a U-shaped Mamba decoder is designed. In this stage, a Multi-Scale Post-processing Block (MPB) is introduced, consisting of depthwise separable convolutions and residual concatenation. This block effectively extracts multi-scale features and enhances local detail extraction after the VSS block. Additionally, a Local Enhancement and Fusion Attention Module (LAS) is added at the end of each decoder block. LAS integrates the SimAM attention mechanism, further enhancing the model’s multi-scale feature fusion capability and local detail segmentation capability. Through extensive comparative experiments, it was found that LMVMamba achieves superior performance on the OpenEarthMap dataset (mIoU 52.3%, OA 69.8%, mF1: 68.0%) and LoveDA (mIoU 67.9%, OA 80.3%, mF1: 80.5%) datasets. Ablation experiments validated the effectiveness of each module. The final results indicate that this model is highly suitable for high-precision land-cover classification tasks in remote sensing imagery. LMVMamba provides an effective solution for precise semantic segmentation of high-resolution remote sensing imagery. Full article

Given that the increasing non-agricultural conversion of cultivated land (NACCL) endangers food security, studying the spatial and temporal variation characteristics and driving mechanisms of NACCL in Sichuan Province can offer a scientific foundation for developing local farmland preservation measures and controlling further conversion. Guided by the theoretical framework of land use transition, this study utilizes land use datasets spanning multiple periods between 2000 and 2023. Comprehensively considering population scale factors, natural geographical factors, and socioeconomic factors, the county-level annual NACCL rate is calculated. Following this, the dynamic evolution and underlying driving forces of NACCL across 183 counties in Sichuan Province are examined through temporal and spatial dimensions, utilizing analytical tools including Nonparametric Kernel Density Estimation (KDE) and the Geographical Detector model with Optimal Parameters (OPGD). The study finds that: (1) Overall, NACCL in Sichuan Province exhibits phased temporal fluctuations characterized by “expansion—contraction—re-expansion—strict control,” with cultivated land mainly being converted into urban land, and the differences among counties gradually narrowing. (2) In Sichuan Province, the spatial configuration of NACCL is characterized by the expansion of high-value agglomerations alongside the dispersed and stable distribution of low-value areas. (3) Analysis through the OPGD model indicates that urban construction land dominates the NACCL process in Sichuan Province, and the driving dimension evolves from single to synergistic. The findings of this study offer a systematic examination of the spatiotemporal evolution and underlying drivers of NACCL in Sichuan Province. This analysis provides a scientific basis for formulating region-specific farmland protection policies and supports the optimization of territorial spatial planning systems. The results hold significant practical relevance for promoting the sustainable use of cultivated land resources. Full article

This article presents an experimental study on the hydrodynamics of coolant flow within the pressure vessel of a small modular reactor (SMR) cooled with water, including areas such as the annular downcomer, bottom chamber, and core-simulating channels that are being developed for use in land-based nuclear power plants. This paper describes the experimental setup and test model, measurement techniques used, experimental conditions under which this research was conducted, and results obtained. This study was conducted at the Nizhny Novgorod State Technical University (NNSTU) using a high-pressure aerodynamic testing facility and a scale model that included structural components similar to those found in loop-type reactors. Experiments were performed with Reynolds numbers (Re) ranging from 20,000 to 50,000 in the annular downcomer space of the test model. Two independent techniques were used to simulate the non-uniform flow field in the pressure vessel: passive impurity injection (adding propane to the airflow) and hot tracer (heating one of the reactor circulation loops). The axial velocity field at the inlet to the reactor core was also investigated. This study provided information about the spatial distribution of a tracer within the coolant flow in the annular downcomer and bottom chamber of the pressure vessel. Data on the distribution of the contrasting admixture are presented in plots. The swirling nature of the coolant flow within the pressurized vessel was analyzed. It was shown that the intensity of mixing within the bottom chamber of the pressure vessel is influenced by the presence of a central vortex. Parameters associated with the mixing of admixtures within the model for the pressure vessel were estimated. Additionally, the possibility for simulating flow with different temperature mixing processes using isothermal models was observed. Full article

Forest parks are vital terrestrial ecosystems that provide multiple ecosystem services (ESs) to both society and nature, including carbon storage, water conservation, soil retention, and tourism-related cultural services. These services are essential for maintaining ecological security and supporting socio-economic development. However, little is known about how ESs vary across forest parks situated in different karst landforms, and integrated re-search on the combined effects of climate, vegetation, karst surface characteristics, and tourism remains limited. In this study, we examine forest parks in Guizhou Province, China, selecting four key ESs—water conservation, soil retention, carbon storage, and cultural services associated with tourism—and evaluate their levels through a comprehensive ecosystem services index (CES). We apply a structural equation model (PLS-SEM) to disentangle how climate, vegetation, karst surface features, and tourism activities drive spatial heterogeneity in CES. The results reveal significant differences among karst land-form units: carbon storage is relatively low in karst plateaus and gorges, whereas water conservation is highest in non-karst areas. Together, the four categories of driving factors explain 71.6–74.2% of the variance in CES, with climate emerging as the dominant contributor to spatial variation. For individual services, the principal drivers differ: normalized difference vegetation index (NDVI) and tourist numbers are jointly shaped by karst surface characteristics and climate, while multi-year average spring precipitation is the most influential factor across forest parks. This study provides new evidence of the socio-ecological mechanisms regulating ESs in karst mountain forestscapes and offers a scientific reference for enhancing and regeneratively managing ecosystem services in these fragile regions. Full article

We document and analyze an emblematic case study of non-indigenous peasant autonomy and re-peasantization in Sa Kaeo province in the Issan region of Thailand, using a mostly qualitative, single case-study methodology. The Kok Edoi autonomous community, whose members engage in community forest management and increasingly in agroecological farming, was founded more than twenty-five years ago as the product of a land occupation by landless peasants associated with the national Thai social movement, the Assembly of the Poor (AoP), which is part of the global peasant movement, La Via Campesina (LVC). Partially inspired by opportunities given to the community and to AoP by LVC to learn and gain inspiration from Latin American experiences such as the Zapatistas in Mexico, Kok Edoi autonomy exemplifies how the exchange of social movement knowledge and experience can help shape and strengthen local struggles, and it is also suggestive of autonomy as an alternative pathway of resistance and sustainable development in Thailand. We review the literature on territorial autonomy, re-peasantization, and community forestry and autonomy in Thailand and the world. Situating Kok Edoi in Thai history concerning policies and conflicts around land and forests, we examine the type, dimensions, and facets of autonomy and re-peasantization present in Kok Edoi to demonstrate how these factors contribute to the community being considered an emblematic case of peasant autonomy, peasant land occupation, peasant management of and livelihood derived from natural resources, more autonomous alternative markets, collective accumulation, and political training and mobilization that contributes to a class-based national movement. This is novel in an academic literature that has to date focused principally on indigenous autonomy, largely in Latin America. Full article

This study aimed to re-evaluate a historical war supply route within the context of cultural tourism, to revitalize its natural, historical, and cultural values, and to integrate it with existing hiking and trekking routes. Remote sensing (RS) and geographic information system (GIS) technologies were utilized, and land surveys were conducted to support the analysis and validate the existing data. Data for slope, one of the most critical factors for hiking route selection, were generated, and the optimal route between the starting and destination points was identified using least cost path analysis (LCPA). Historical, touristic, and recreational rest stops along the route were mapped with precise coordinates, and both the existing and the newly generated routes were assessed in terms of their accessibility to these points. Field validation was carried out based on the experiences of expert hikers. According to the results, the length of the existing hiking route was determined to be 15.72 km, while the newly developed trekking route measured 17.36 km. These two routes overlap for 7.75 km, with 9.78 km following separate paths in a round-trip scenario. It was concluded that the existing route is more suitable for hiking, whereas the newly developed route is better suited for trekking. Full article

Watershed ecosystem resilience (RES) plays a vital role in supporting ecosystem sustainability. However, comprehensive assessments and investigations into the complex mechanisms driving RES remain limited, particularly in ecologically sensitive basins. To address this gap, this study proposes a multidimensional RES evaluation framework tailored to watershed-specific natural characteristics. The framework integrates five core dimensions: ecosystem resistance, ecosystem recovery capacity, ecosystem adaptability, ecosystem services, and ecosystem vitality. RES patterns under 2030 different future scenarios were simulated using the PLUS model combined with CMIP6 climate projections. Spatial and temporal dynamics of RES from 2010 to 2020 were quantified using Geodetector and Partial Least Squares Path Modeling, offering insights into the interactions among natural and anthropogenic drivers. The results reveal that RES in the Upper Yangtze River Basin exhibits a spatial gradient of “high in the east and west, low in the middle” with an overall 2.80% decline during the study period. Vegetation coverage and temperature emerged as dominant natural drivers, while land use change exerted significant indirect effects by altering ecological processes. This study emphasizes the importance of integrated land-climate strategies and offers valuable guidance for enhancing RES and supporting sustainable watershed management in the context of global environmental change. Full article

Historic centres are extremely complex parts of contemporary cities, particularly from morphological, architectural, and cultural points of view, where a significant proportion of the land area may be occupied by built heritage sites that require protection and conservation. These urban contexts are also characterised by scarce green and public open spaces endowment, a high proportion of private property, and high levels of natural risk exposure. From a climate change adaptation perspective, Nature-based Solutions (NbS) have emerged as measures to manage urban ecosystems to address environmental and societal challenges. To overcome the conflicting objectives of climate change adaptation and built heritage preservation, this study proposes a three-step methodology applied to the historic centre of Catania (Italy): (i) Land-Use/Landownership and Land Cover/Maintenance and Quality analyses; (ii) Land Transformability Assessment; (iii) Land Transformation Scenarios Assessment. According to this methodology, five Land Transformation Scenarios have been drawn up: (1) NbS full installation; (2) NbS installation with some limitations; (3) NbS installation after re-arrangement; (4) NbS installation strongly limited; (5) NbS installation not viable. This approach allowed us to identify the most feasible and suitable buildings and open spaces, while distinguishing public and private properties, to implement a more comprehensive integration of NbS and built heritage preservation in historic cities for mutual benefits. Full article

This study investigates the Roman land division system, centuriation, using LIDAR data and historical data to understand the landscape during the Roman period, in this case between Roman cities such as Siscia and Andautonia. LIDAR data analysis provided evidence of the preservation of the Roman centuriation system in the present day Turopoljski Lug forest. The azimuth suggests that centuriation aligned with Siscia’s ager, while the precise territorial limits between the two agers remain unclear. Additionally, the orientation of Siscia’s streets and the alignment of modern roads like Zagrebačka street suggest continuity of the Roman road system. The research also sheds light on the agricultural nature of the region in the Roman period, challenging traditional views of Turopolje as a marshy, forested area from prehistoric periods. The presence of Roman-era drainage systems and the re-evaluation of the historical landscape indicate that the region was actively cultivated. The study also discusses the abandonment of the centuriation system after the Roman period and its subsequent transformation into forested land. Future research should focus on the exact borders between the agers of Siscia and Andautonia and the ongoing influence of Roman land division on later historical landscapes. Full article

Deep neural networks (DNNs) have shown strong performance in synthetic aperture radar (SAR) image classification. However, their “black-box” nature limits interpretability and poses challenges for robustness, which is critical for sensitive applications such as disaster assessment, environmental monitoring, and agricultural insurance. This study systematically evaluates the adversarial robustness of five representative DNNs (VGG11/16, ResNet18/101, and A-ConvNet) under a variety of attack and defense settings. Using eXplainable AI (XAI) techniques and attribution-based visualizations, we analyze how adversarial perturbations and adversarial training affect model behavior and decision logic. Our results reveal significant robustness differences across architectures, highlight interpretability limitations, and suggest practical guidelines for building more robust SAR classification systems. We also discuss challenges associated with large-scale, multi-class land use and land cover (LULC) classification under adversarial conditions. Full article

Accurate and timely disaster classification is critical for effective disaster management and emergency response. This study proposes an improved ResNet-50-based deep learning model to classify seven types of natural disasters, including earthquake, fire, flood, mudslide, avalanche, landslide, and land subsidence. The dataset was compiled from publicly available sources and partitioned into training and validation sets using an 8:2 split. Experimental results demonstrate that the proposed model achieves a classification accuracy of 87% on the validation set and outperforms the traditional VGG16 model in most evaluation metrics, including precision, recall, F1-score, AUC, specificity, and log loss. Furthermore, the model effectively mitigates the gradient vanishing problem, ensuring stable convergence and robust training performance. These findings provide a practical technical reference for multi-disaster classification tasks and contribute to enhancing the efficiency of disaster response and societal resilience. Full article

Traditional urban expansion struggles to balance economic and ecological demands. Intensive development planning based on re-naturalization has become the policymakers’ choice. However, planning-oriented land use patterns and re-naturalization pathways remain difficult to determine. This study developed a spatial decision-support framework integrating ecosystem service valuation (ESV), land-use simulation, and ecological planning for Shanghai. This study assessed the spatiotemporal dynamics of ESV and simulated land use patterns and ESV for 2035 under different scenarios (inertial development, cropland protection, and ecological development). The optimal scenario and corresponding re-naturalization pathways were determined based on the principle of the optimal ESV. The results showed that ESV has declined over the past 20 years (−5.21%/5 years). High-value areas shrank significantly due to ecological space degradation. The planning-oriented ecological development scenario is the optimal scenario, with the highest ESV of CNY 189,240.29 million, which is higher than the status quo, inertia development scenario, and cropland protection scenario by 9.69%, 23.27%, and 9.53%, respectively. Taking the land use patterns under the ecological development scenario as the re-naturalization objective, 12 re-naturalization pathways totaling 686.88 km² were identified. Cropland to forestland and built-up land to cropland were the largest, accounting for 67.88% and 15.02%, respectively. This study provides valuable insights into ecological planning and re-naturalization in urbanized areas. Full article

The concentrations of dissolved N₂O in river systems at the basin scale exhibit significant spatial and temporal variability, particularly under diverse landscape conditions. This study focused on a temperate basin—the Yiluo River (YLR) basin in China—to investigate the variations in dissolved N₂O concentrations and the indirect emission factors (EF_5r) between the dry and wet seasons. The differences among tributaries were analyzed to assess the impact of land use types. The findings revealed that N₂O concentrations and saturation levels were lower during the wet season in both the main streams and tributaries. In the dry season, the N₂O concentrations were strongly correlated with NH₄⁺-N, NO₃⁻-N, and oxidation–reduction potential (ORP) (R² = 0.743, p < 0.001), while in the wet season, the N₂O concentrations were correlated with dissolved phosphorus (DP), water temperature (Tw), NH₄⁺-N, and DOC (R² = 0.640, p < 0.001). Impervious land was identified as the primary source of nitrogen in both seasons, rather than cropland. Natural land, particularly shrubland, demonstrated a notable mitigating effect on N₂O accumulation and played a significant role in reducing NO₃⁻-N levels. The YLR basin exhibited lower EF_5r values (0.005–0.052%) compared to the default value recommended by the IPCC, with a significant decrease observed during the wet season (p < 0.001). Data analysis indicated that nutrient dynamics, particularly NO₃⁻-N, the ratio of dissolved organic carbon to NO₃⁻-N (DOC/NO₃⁻-N), and the ratio of NO₃⁻-N to DP (NO₃⁻-N/DP), were significantly correlated with EF_5r. These results underscore the need to re-evaluate regional N₂O emission potentials and provide new insights into mitigating N₂O emissions through strategic land use management. Full article

Evaluating potential (PES) and realized (RES) ecosystem services can significantly improve the clarity and understanding of sustainable natural resource management practices. This study determined spatial distribution indices and assessed the economic value of both PES and RES in Northern Thailand. The geographic distribution and intensity of 17 ecological services of six land use categories (i.e., forests, agriculture, shrubland, urban land, water bodies, and barren land) were estimated for the distribution and unit values of PES and RES, by using the Co$ting Nature Model. Our results suggested that the PES and RES values were spatially consistent. The map showing the distribution of PES and RES values revealed high values in the cities of Chiang Mai, Chiang Rai, Lamphun, Lampang, Phitsanulok, and Nakhon Sawan. Nutrient cycling, soil formation, and water supply were identified as the top potential ecological services, while nutrient cycling, water supply, and soil formation were the most realized. The ecosystem service packages in Northern Thailand had a potential annual value of 36.31 billion USD per year. However, after adjusting for relative indices, the realized ecosystem services were valued at 13.44 billion USD per year, representing only one-third of the potential value. To manage resources effectively and make informed decisions, it is essential to comprehend the gap between possible and actual ecosystem services. This research underscores the financial worth of ecosystem services and emphasizes the significance of using them sustainably to enhance human well-being and conserve the environment in Northern Thailand. Full article

Vineyards are significant demanders of fertilisers, pesticides, soil tillage and water. This study assessed the environmental profile of an organic grape production system with La Rioja qualified designation of origin using a cradle-to-gate life cycle assessment (LCA). The ReCiPe method was applied to assess the environmental impacts, while the Available WAter REmaining method was used to estimate the water scarcity. Additionally, the biodiversity loss, a global issue exacerbated by agricultural practices, was evaluated along with an ecosystem service indicator, pollination, to provide a more comprehensive analysis. This study employed two functional units: one kilogram of grapes and one hectare of land. The results revealed that the environmental impacts on global warming were more than ten times lower than those reported in most studies reviewed in the literature, primarily due to the effects of direct land use changes associated with pruning waste management. The total emissions in this category were 99.51 kg CO₂ eq per hectare or 15.31 g CO₂ eq per kilogram of grapes. Agrochemical-related emissions were identified as the environmental hotspot. The water scarcity was estimated at 48.4 litres per kilogram of grapes, mainly attributed to agrochemical dispersion. The biodiversity loss was largely driven by land transformation, with plants being the most impacted taxon. However, a high abundance of pollinators was observed in spring, contributing to improved grape quality and natural pest control. These findings could help highlight the environmental benefits of organic viticulture and the good practices implemented in this pilot. Full article