Search Results (772)

As urban rail transit systems enter a stage of high-quality development, accurate short-term passenger flow forecasting has become essential for efficient operation and management. To improve the accuracy and robustness of multi-step short-term passenger flow forecasting under fluctuating demand conditions, this study develops a feature-enhanced Informer framework that embeds Complex Network Representation (CNR) into the Informer architecture to jointly capture both the topological characteristics of the subway network and spatiotemporal passenger flow dynamics. External factors, including subway schedules and land use around stations, are further integrated to enrich the input features. In addition, the ProbSparse self-attention mechanism is adopted to improve long-sequence dependency modeling, thereby enabling efficient multi-step passenger flow forecasting. Experiments were conducted on the Beijing metro passenger flow dataset from January to October 2024 to evaluate the proposed model. The dataset covered 264 stations and was aggregated at 15 min intervals. Based on historical passenger flow and multi-source features, the model predicts passenger flow over multiple future time steps. The overall evaluation metrics were calculated on the test set and averaged over all test samples and observed stations. The experimental results show that, compared with the standard Transformer model, the proposed model reduces the average prediction error by 16.59% on weekdays and 20.48% on weekends while maintaining stable predictive performance during peak hours. Sensitivity analysis and ablation studies are further conducted to evaluate the model performance across different station types and forecasting horizons. The results demonstrate that the proposed model can provide reliable decision support for intelligent urban rail transit operations, including transport capacity scheduling, passenger service improvement, and operating cost reduction. Full article

Irrigation water management in arid systems requires spatially distributed estimates of crop evapotranspiration (ET) that fixed crop coefficients cannot provide. The actual ET of flood-irrigated rice (Oryza sativa L.) on the arid northern coast of Peru was mapped with the METRIC surface energy balance model (Mapping EvapoTranspiration at high Resolution with Internalized Calibration) on Google Earth Engine (GEE). Ten cloud-free Landsat 8/9 scenes (January–July 2022) were processed over 113 ha at Ferreñafe (Lambayeque) on the 30 m product grid, onto which the 100 m native thermal observation was resampled, with internal calibration based on automatic anchor-pixel selection and hourly ERA5-Land data. Daily field-mean ET ranged from 4.2 to 8.1 mm d⁻¹, peaking during flooding and establishment and declining towards harvest. Because the same reference $E{T}_o$ underlies the METRIC internal calibration and the FAO-56 estimate, this is a comparison between two modelling approaches rather than an independent validation. Against the FAO-56 reference ET, METRIC showed a positive bias of $+0.65$ mm d⁻¹ (percent bias (PBIAS) $=+13\%$; root mean square error (RMSE) $=1.23$ mm d⁻¹; $r^2=0.57$; $n=9$, after excluding one date with anomalous reanalysis forcing), concentrated during flooding and after harvest, whereas at full canopy cover the two estimates converged. Two global ET products that share neither the METRIC formulation nor the ERA5-Land forcing reproduce the same seasonal decline once the canopy closes ($r=0.63$ and $0.91$) but stay far below in magnitude, as expected from their 500 m pixel. ET did not differ between sowing methods and varied only slightly among cultivars (∼0.3 mm d⁻¹), against marked intra-field variability. The METRIC–GEE workflow offers a low-cost, high-resolution tool for monitoring water use in data-scarce arid rice systems. Full article

Change detection (CD) is a fundamental remote sensing task that identifies surface modifications from multi-temporal imagery of the same area, with applications in urban monitoring, agriculture, forest disturbance mapping, disaster assessment, and land cover analysis. The task is complicated by radiometric and atmospheric variability, co-registration errors, seasonal cycles, and sensor heterogeneity. Deep learning has progressively superseded traditional and classical machine learning approaches through hierarchical feature extraction and end-to-end optimization. Following the PRISMA 2020 guidelines, this systematic review examines 144 primary studies identified through a structured Scopus search complemented by the authors’ prior research and citation searching, spanning three paradigms: traditional approaches (algebraic operators, transformations, probabilistic frameworks), classical machine learning (support vector machines, random forests, object-based analysis), and deep learning architectures (fully convolutional, Siamese, attention-based, Transformer, state space, diffusion-based, and weakly supervised models). We provide background on problem formulation, benchmark datasets, and evaluation metrics, alongside a taxonomy organized by paradigm and supervision mode. A quantitative comparison on dominant benchmarks reveals the strengths and limitations of current methods. Open challenges include the absence of a universal benchmark protocol, the research-to-deployment gap, and the need for label-efficient learning. This review serves as a structured reference and outlines promising directions for the field. Full article

Data-driven models are increasingly used for flood and landslide susceptibility mapping in rapidly urbanizing regions, particularly in the Global South. In this context, land use and land cover (LULC) is routinely adopted as a conditioning factor, although its geomorphological meaning and temporal consistency with hazard inventories are seldom evaluated. Information Value (IV) models have been widely applied to landslide susceptibility, but their use for flood susceptibility in complex coastal cities remains limited. This study evaluates IV-based flood and landslide susceptibility in Vitória, Brazil, a predominantly insular city characterized by sharp geomorphological contrasts and high population density. Two LULC datasets with different levels of urban detail were tested as conditioning factors alongside topographic and hydrological variables. Eighteen models were constructed for each hazard and validated using Area Under the Curve (AUC) metrics and expert judgement. Morphological attributes were the most important predictors: slope alone achieved an AUC of 0.90 for landslides, whereas elevation reached 0.71 for floods. The inclusion of LULC increased AUC values to 0.94–0.95 for landslides and 0.85–0.89 for floods, but also introduced spatial and temporal biases associated with stationary and coarsely classified land cover. Our findings highlight the limitations of incorporating LULC as a conditioning factor without temporal harmonization with hazard inventories or adequate urban class disaggregation. We argue that, in complex urban settings, LULC is more appropriately interpreted as a proxy for exposure and vulnerability than as a dominant predisposing factor, and that its use in predictive models should be critically assessed to avoid misleading conclusions. Full article

Soil erosion on sloping cropland is a major threat to agricultural sustainability and ecological security in mountainous regions, yet its spatial distribution and landscape-level structural characteristics remain insufficiently quantified. Taking Zhaotong in northeastern Yunnan, China, as a typical mountainous agricultural region in the upper Yangtze River Basin, this study integrated Sentinel-2 imagery, high-resolution reference data, field survey information, the Google Earth Engine platform, a random forest classifier, the Chinese Soil Loss Equation, and landscape pattern metrics to assess soil erosion on sloping cropland. The land use classification achieved an overall accuracy of 90.70% and a Kappa coefficient of 0.88, providing a reliable basis for sloping cropland extraction. Sloping cropland covered 4218.77 km², accounting for 84.64% of total cropland area, but contributed 1.84 × 10⁷ t·yr⁻¹ of annual soil loss, equivalent to 97.51% of total cropland erosion. The mean erosion rate of sloping cropland reached 4260.50 t·km⁻²·yr⁻¹, and 95.84% of sloping cropland exceeded the soil loss tolerance threshold. County-level analysis revealed strong spatial heterogeneity, with high erosion risks concentrated in northern and eastern mountainous counties. Intensive, Severe, and Extreme erosion occupied only 26.14% of the sloping cropland area but contributed 62.21% of total soil loss. Landscape metrics further showed that Moderate erosion had the highest patch density and edge density, indicating a critical fragmentation stage in erosion development. These findings support a tiered conservation strategy in which high-intensity patches are prioritized for immediate sediment reduction, while fragmented Moderate-erosion (2500–5000 t·km⁻²·yr⁻¹) areas receive preventive management. The proposed framework provides a useful approach for identifying erosion hotspots and supporting targeted soil and water conservation in mountainous agricultural landscapes. Full article

Understanding how land-use change affects habitat fragmentation and connectivity is essential for assessing landscape degradation and conservation effectiveness in protected areas globally. It is particularly acute in tropical protected areas where anthropogenic pressures are intensifying. This study investigated long-term landscape dynamics, forest fragmentation, and structural connectivity in the Bena Mulumbu Hunting Domain, a Category VI protected area located in the Miombo woodland region of southeastern Democratic Republic of the Congo. Landsat imagery acquired in 1995, 2005, 2015, and 2025 was classified using the Random Forest algorithm into six land-cover classes (Miombo woodland, savanna, agricultural land, mining areas, built-up/bare land, and water bodies) to quantify land-cover changes over 30 years. Landscape composition was assessed using the percentage of landscape (PLAND), Shannon diversity metrics, and transition analyses. At the same time, fragmentation and structural connectivity of Miombo woodland were evaluated along an edge-to-core gradient (0–2 km, 2–4 km, 4–6 km, and >6 km) using landscape metrics. Results showed that savanna remained the dominant land-cover type throughout the study period. However, the landscape underwent progressive reorganization characterized by recurrent transitions among Miombo woodland, savanna, and agricultural land, leading to increased spatial heterogeneity. Fragmentation analyses revealed significant spatial differences in total core area among zones (Kruskal–Wallis: H = 8.12, p = 0.044); however, after normalization by zone area, no consistent edge-to-core gradient was observed for core habitat proportion, indicating that raw differences primarily reflect zone size rather than a systematic ecological gradient. Despite increasing fragmentation, structural connectivity remained high across the hunting domain. The CONNECT index increased significantly from the edge toward the core zone (p = 0.003), highlighting better-connected forest networks in interior sectors. These findings suggest that the Bena Mulumbu Hunting Domain is experiencing an intermediate stage of landscape transformation, where forest fragmentation is evident but has not yet resulted in widespread connectivity loss. Maintaining existing forest cores and connectivity corridors should therefore be prioritized to prevent further degradation of ecological integrity. These findings challenge the assumption that landscape degradation in protected tropical Miombo woodlands necessarily follows a simple edge-to-core gradient. Full article

Post-disaster recovery is increasingly understood not as a simple return to pre-event conditions but as a dynamic reorganization of land systems, in which land cover and land use change (LCLUC) provides an operational signature of recovery trajectories. However, most existing assessments reduce recovery to a single dimension—typically vegetation greenness—which can conflate systems that differ fundamentally in their response behavior. This study develops a multi-index Recovery Resilience Index for Land Systems (RRI-LS) within Google Earth Engine and applies it to the catastrophic liquefaction zone of the 2018 Mw 7.5 Palu earthquake (Central Sulawesi, Indonesia). Combining Sentinel-2 spectral indices (NDVI, NDBI, BSI), Dynamic World land-cover labels, and a hybrid Top-of-Atmosphere/Surface-Reflectance baseline to overcome the sparse pre-event archive, we quantify three resilience dimensions—resistance, recovery, and stability—and classify recovery into qualitatively distinct pathways. The hybrid baseline is quantitatively validated: after removing a small systematic offset, the residual discrepancy between TOA- and SR-derived indices is 2.4–4.9 times smaller than the measured disturbance signal. Site-level analysis of the three principal liquefaction hotspots (Balaroa, Petobo, Jono-Oge) and a 1 km grid expansion (n = 962 cells) reveal that disturbance-affected areas did not converge on a single outcome but diverged into bounce-back, transformational, reconstructed (non-vegetated), and degraded pathways; a basin-wide re-run at 250 m (n = 14,157 cells) reproduced the same pathway hierarchy, confirming robustness to grid resolution. Initial disturbance intensity was a poor predictor of long-term recovery (R² = 0.07, p < 0.001), underscoring that recovery is multidimensional and not reducible to a single shock variable. The emergence of a reconstructed, non-vegetated pathway—where bare-soil disturbance is resolved through built surfaces rather than re-greening—demonstrates that vegetation metrics alone are insufficient in human-dominated landscapes. The framework supports a land-system perspective in which recovery is conceptualized as the establishment of new functional equilibria. Full article

Rapid urbanization along transportation corridors is a key driver of land transformation and landscape homogenization in developing mega-cities. This study analyzes land use and land cover (LULC) change within a 5 km buffer, from Gulberg to T-Chowk, along the Islamabad Expressway, Pakistan, from 2010 to 2024, using multi-temporal Landsat imagery (2010, 2015, 2020, 2024) and landscape metrics. Built-up area increased by 190.8% (51.02 to 148.34 km²) between 2010 and 2024. The expansion is accompanied by sharp declines in vegetation (−36.0%; 89.17 to 57.06 km²), barren land (−93.8%; 63.88 to 3.94 km²), and water bodies (−64.2%; 8.20 to 2.93 km²). All class-area estimates were derived from the R landscapemetricspipeline and independently verified against visual inspection of the classified rasters, confirming correct class labelling throughout. Landscape structure shifted towards homogenization, with Shannon’s Diversity Index decreasing from 1.19 to 0.74, Patch Density from 74.39 to 21.08 patches per 100 ha, and Edge Density from 219.65 to 99.78 m/ha. LULC maps achieved an accuracy greater than 96% ($\kappa >0.96$) for each classification year, and metric estimates showed cross-platform agreement between R and FRAGSTATS within ±5%. A distance-zone analysis across four successive 1.25 km bands from the Expressway showed that this expansion was pervasive rather than confined to the roadside, with the urban share of classified land rising by 43.6–47.7 percentage points in every band. A complementary 300 m × 300 m fishnet-grid multivariable logistic regression ($n=2162$ cells classified as non-urban in 2010) identified proximity to existing 2010 built-up land and location within a formal housing scheme as the strongest independent correlates of conversion to urban land (AOR = 1.88 for housing-scheme proximity; 95% CI: 1.38–2.56), with acceptable model discrimination (AUC = 0.739). These findings indicate rapid corridor-scale consolidation associated with infrastructure-led growth, reducing landscape heterogeneity and potentially weakening ecological resilience. Unlike previous city-scale studies in the Islamabad–Rawalpindi region, this study adopts a transportation-corridor perspective, combining descriptive landscape ecology with spatially explicit statistical modelling to quantify both the pattern and the spatial drivers of infrastructure-led land transformation. The study underscores the necessity for integrated, corridor-scale land-use governance—including transit-oriented development and landscape-metric-based ecological zoning—to balance urban expansion with ecosystem sustainability. Full article

Tropical riparian ecosystems are among the most threatened biomes on Earth, yet decade-scale, spatially explicit documentation of vegetation loss and landscape fragmentation in community-based ecotourism corridors remains scarce. This study investigates whether agricultural intensification associated with ecotourism development has driven measurable vegetation decline and landscape fragmentation in the Ledok Amprong riparian corridor (Poncokusumo District, Malang Regency, East Java; 8.21 ha, 800–900 m a.s.l.), a mixed-canopy riparian system dominated by Ficus benjamina, Albizia chinensis, and bamboo (Phyllostachys spp.) interspersed with smallholder coffee (Coffea canephora) agroforestry, over the period 2015–2024. We applied multi-temporal Sentinel-2 (10 m) imagery, NDVI analysis, supervised Maximum-Likelihood classification (Overall Accuracy: 89.75%, Kappa: 0.86), and FRAGSTATS-derived landscape metrics to quantify vegetation dynamics and land cover transition. Results show that agricultural land expanded by 157.51% (1.14 to 2.93 ha), while vegetation cover declined by 28.45% (6.27 to 4.49 ha). A strong negative correlation between agricultural expansion and vegetation loss (r = −0.96, p < 0.001) confirmed agriculture as the primary driver of ecosystem degradation. Landscape fragmentation intensified markedly: patch density increased by 49.6%, the connectivity index declined by 23.5%, and the fragmentation index increased by 54.3%. These findings establish the first quantitative, decade-scale spatiotemporal baseline for this ecotourism corridor and demonstrate that community-based ecotourism development, without adequate spatial governance, accelerates rather than mitigates riparian degradation—a paradox with broad implications for tropical riparian conservation governance worldwide. Evidence-based management recommendations include 10–30 m dynamic buffer zones, a 2.5 ha targeted restoration programme, and a Payment for Ecosystem Services (PES) mechanism, offering a transferable framework for sustainable riparian ecotourism management across rapidly urbanizing tropical watersheds. Full article

The sustainable management of land resources and the formulation of land policies are closely linked to the stability and health of terrestrial ecological systems, which in turn underpin sustainable regional economic, social, and environmental development. However, land use change has a time effect on the environment and requires long-term observation to discover its impact on landscape patterns. The Yellow River Basin functions as a critical ecological barrier in northern China, where land use changes are particularly intense in the transitional zone between its middle and lower reaches. Using Landsat imagery as the data source, this study adopts the Random Forest (RF) algorithm to classify eight sets of sequential data covering a 35-year period from 1990 to 2025 in the study area. Landscape pattern metrics and transfer matrices are employed to conduct qualitative and quantitative analyses of the spatiotemporal dynamics of land use changes. Additionally, land expansion analysis strategies and the RF algorithm are applied to identify the relative importance of different driving factors. The results show that: (1) The classification accuracy based on the Google Earth Engine (GEE) cloud platform remains consistently high, exceeding 90% across all phases. (2) Patch density decreases significantly, while the largest patch index continues to decline; the Shannon diversity index shows a fluctuating upward trend, and the aggregation index exhibits a slight increase. (3) Mutual conversions among farmland, forest, and grassland are the dominant processes driving land use changes in the region. (4) The Digital Elevation Model (DEM), construction land area distribution, and distance to primary roads are the key factors influencing land use patterns, with human activities acting as the primary driver of land use type transformations in the area. Full article

Landslide susceptibility mapping (LSM) serves as a fundamental technical support for geohazard prevention and mitigation across mountainous terrains. This research constructs a multi-scale terrain unit integrated modeling framework targeting complex mountainous geomorphic settings, taking Zhenping County as the research object. Multi-resolution digital elevation model (DEM) datasets, multi-source satellite remote sensing imagery (GF-2), geological vector datasets and hydrological survey data are jointly adopted as the basic data source. The r.slopeunits module embedded in GRASS GIS is utilized to automatically segment slope units, and a comprehensive composite index S, coupling slope partition quality indicator F and model prediction accuracy metric R, is proposed to adaptively optimize two critical slope-unit hyperparameters: circular variance (c) and minimum unit area (a). Four DEM spatial resolutions (15 m, 25 m, 50 m, 100 m) are systematically calibrated with 42 groups of c–a parameter combinations to screen out the optimal slope-unit segmentation scheme (c = 0.1, a = 200,000 m²). Twelve landslide predisposing covariates covering topography, hydrology, lithology, human engineering activities and land cover are selected after multicollinearity diagnosis via Variance Inflation Factor and mean utility factor contribution evaluation. Logistic regression tree (LMT), LMT-Adaboost and LMT-Random Subspace are compared by random cross-validation and spatial block cross-validation. Parameter sensitivity analysis is further carried out to quantify the stability of model outputs against DEM resolution and slope-unit parameter perturbations. The LMT-RSM ensemble achieved the highest spatial cross-validation AUC (0.954 ± 0.019), outperforming LMT (0.925 ± 0.023) and AdaBoost-LMT (0.934 ± 0.021). The DeLong test confirmed that LMT-RSM’s superiority over LMT is statistically significant (p < 0.0001). The proportion of landslides in the very high and high susceptibility zones under the LMT-RSM model reached 95.98%, demonstrating relatively excellent spatial discrimination. This study provides an operational framework combining optimized slope units, ensemble learning, and spatially explicit validation for robust LSM in complex terrain, and offers a reproducible technical pathway for landslide risk prevention in mountainous regions. Full article

Urban fringe forests deliver critical ecosystem services yet face irreversible loss and complex degradation under urbanization, while their fragmentation dynamics relative to core forests remain largely unquantified owing to the lack of a spatiotemporally consistent mapping approach. To address this gap, a multi-source remote sensing framework integrating land cover, population, nightlight, and land surface temperature data was developed to delineate urban fringe boundaries and quantify forest dynamics in Zhejiang Province via area-weighted centroids and landscape metrics, where high forest cover and a polycentric urban structure create a highly heterogeneous and dynamic fringe environment, enabling separate quantification of spatiotemporal fragmentation patterns for urban fringe and core forests from 2004 to 2024. The results are as follows: (1) The four dimensions (land, population, economy, and environment) produced spatiotemporally distinct boundaries, with multi-dimensional integration outperforming any single indicator and nighttime light being the best. Meanwhile, the total fringe area grew from 2989.04 to 3990.66 km² over two decades, with the fastest growth in 2004–2014 and the most rapid boundary shifts in 2014–2019. (2) The fringe forest proportion dropped from 24.72% to 18.37%, with the largest decline in southwestern high-forest cities. Meanwhile, Hangzhou and Ningbo fringe forests increasingly assumed the main ecological carrier role formerly held by core forests, with their centroids moved southwestward most markedly in 2009–2014 and displaying a more consistent direction than core forests. (3) Fragmentation metrics showed a higher patch density and splitting index but lower connectivity in the fringe than in the core, with intensification peaking in Zhoushan and coinciding with intensive edge expansion in 2009–2014, followed by later responses in the core. This study provides a transferable multi-dimensional remote sensing methodology for urban fringe mapping indicating that fringe forests may serve as early-warning signals of urbanization-induced forest degradation, enabling spatially targeted forest management across varied urban contexts. Full article

Regional landslide susceptibility mapping commonly relies on static conditioning factors, including terrain, geology, hydrology, land cover and human activity. These factors describe long-term instability settings but cannot directly represent recent or ongoing ground deformation. Interferometric Synthetic Aperture Radar (InSAR) can provide spatially distributed deformation information, yet mountainous InSAR evidence is affected by uneven observation availability, vegetation decorrelation, terrain-induced geometric distortion and confounding between observation support and static environmental conditions. These issues make it difficult to determine whether radar-derived variables represent deformation signals or mainly indicate where observations are reliable. This study develops a scale-sensitive and confounding-audited Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) evidence representation framework for regional landslide susceptibility mapping. In Pingwu County, China, a 1607-record landslide inventory was converted into 1579 unique 30 m landslide cells, with 1393 for training and 186 for inventory-concentration validation. Fourteen static factors formed the baseline model. Deformation evidence from 98 Sentinel-1A descending acquisitions was represented using point-based line-of-sight (LOS) variables, neighbourhood component descriptors, a compressed deformation-intensity and observation-availability index, and a separated deformation intensity (DI) plus observation availability (RI) representation. Results show that the static factors already provided a strong first-order baseline. Adding SBAS-InSAR evidence did not produce uniform paired improvements across models, metrics or neighbourhood scales. In 10 km spatial-block cross-validation, random forest models using static factors, 300 m component descriptors and 300 m DI + RI features achieved similar mean area under the receiver operating characteristic curve values of 0.948, 0.949 and 0.950, with mean Matthews correlation coefficient values of 0.791, 0.795 and 0.795. Broader neighbourhood representations may expand the attainable performance boundary, but only as diagnostic evidence. SBAS-InSAR-derived information should therefore not be treated as a simple additional conditioning factor; its value depends on jointly interpreting deformation intensity, observation availability and their confounding with static context. Full article

With the acceleration of global urbanization and intensified agricultural activities, approximately 61% of the world’s wetlands have degraded over recent decades, significantly weakening their carbon sequestration capacity. The Shibalianwei Wetland, a crucial tributary system of Lake Chaohu in China, has suffered severe degradation due to land use and cover change, nutrient loading and hydrological disruption. In response, large-scale ecological restoration has been implemented since 2018. To quantify the restoration outcomes, this study integrated remote sensing, GIS, and machine learning techniques, employing the XGBoost model to evaluate and predict carbon sequestration in 2017 and 2024 based on 2010 carbon data. The results reveal that the average carbon density increased from 48.70 t ha⁻¹ in 2017 to 90.18 t ha⁻¹ in 2024, representing an overall increase of 85.2% in total carbon storage. This substantial enhancement is primarily attributed to land use transitions and ecosystem-scale restoration effects, including vegetation recovery and hydrological rehabilitation. Model validation indicated moderate prediction errors (RMSE = 0.47–0.74), with consistent performance across repeated iterations. Together with complementary MAE and R² metrics, the results suggest that the XGBoost model is capable of capturing relative spatial patterns and restoration-induced changes in wetland carbon sequestration, while retaining reasonable predictive stability under changing landscape conditions. Overall, the findings demonstrate that large-scale wetland restoration can rapidly and effectively enhance regional carbon sink capacity and highlight the potential of data-driven modeling frameworks to support wetland management and carbon-neutrality strategies. This provides important guidance for policymakers to promote sustainable land use and optimize ecosystem management under China’s dual-carbon development goals. Full article

Long-term forest landscape change in mountainous rural areas reflects interactions among land-use pressure, forest structure and human activities. This study had two primary objectives: (i) to quantify 53 years of change in forest extent, stand structure and landscape configuration in the Kümbet Planning Unit in the Eastern Black Sea Region of Türkiye and (ii) to evaluate associations of annual registered forest crime counts with election-year status and broader-scale population and real per capita income indicators while treating the crime records as contextual indicators of recorded pressure rather than as causal measures of landscape change. Forest stand-type maps from 1971, 2013 and 2024 were analyzed using LULC transition matrices and landscape metrics. Annual registered forest crime records for 1981–2025 were examined in relation to election-year status, the provincial population and real per capita income using correlation analysis and negative binomial regression, with linearly interpolated annual forest area included as an exposure offset. Total forest area declined from 6747.8 to 6298.1 ha, whereas agricultural land increased from 637.3 to 2177.7 ha. Degraded forest and open areas decreased, and 44.3% of the study area changed land-cover type. Landscape fragmentation increased between 1971 and 2013, followed by partial spatial consolidation between 2013 and 2024. Election-year status and real per capita income were not significantly associated with annual registered forest crime counts. Within the 1981–2025 study period, the provincial population showed a positive temporal association with registered forest crime counts. Each increase of 10,000 persons corresponded to an approximately 7.5% increase in the expected registered forest crime rate, but this relationship should not be interpreted as a direct local demographic effect. Registered forest crime density did not differ significantly among management-related periods. The temporal correspondence between higher crime density and stronger fragmentation was descriptive and did not establish causality. These findings demonstrate that forest conservation and landscape-restoration planning should integrate forest extent, stand structure, land-cover transitions, landscape configuration and appropriately scaled administrative indicators of recorded forest-related pressure. Full article