ABSTRACT. Tactile maps play a crucial role in improving spatial accessibility, orientation, and independent mobility for visually impaired individuals. Despite increasing academic and technological interest in tactile cartography, current studies reveal significant inconsistencies in design principles, Braille usage, scale decisions, color coding, tactile texture and tactile symbols. The lack of standardized guidelines limits the usability and transferability of tactile maps in different spatial phenomena and among user groups. This article presents a systematic review of academic research on tactile maps designed for visually impaired users; focusing particularly on tactile map design approaches, Braille integration, information density, tactile symbols and thematic content. The review encompasses peer-reviewed journal articles and conference proceedings published between 2020 and 2025, obtained from the Web of Science, Scopus, Google Scholar, and IEEE Xplore databases. 162 studies were selected and structured according to the tactile map thematic framework (design, scale, symbols, labeling, user testing) The findings demonstrate an increasing emphasis on user-centered and participatory design approaches. However, it remains limited by the consideration of various user profiles (e.g., congenital or acquired visual impairment). Braille labeling is widely used, but its effectiveness largely depends on users' Braille literacy levels, leading various studies to propose different tactile or auditory solutions. Scale selection, generalization and tactile sign differentiation emerge as the most critical challenges in depicting complex urban environments, particularly as they negatively impact the readability of tactile details and needs on the map. This study contributes to cartographic research by synthesizing key trends, methodological gaps, design limitations in tactile map and legend design approaches, tactile signs, Braille labeling, scale decisions etc., providing evidence-based information for the development of more usable and standardized tactile maps. The results aim to provide a practical guide for cartographers, GIS professionals, and researchers working in the field of accessibility and disability studies.

ABSTRACT. Cartographic generalization is a complex process directly related to changes in the scale of mapping. Cartographic generalization as a process consists of two components: 1) the cartographic form of abstraction, which is determined by the degree of abstraction, i.e., the spatial scale; 2) the cartographic form of generalization, which is determined by the degree of generalization, i.e., the scale of content. The subject of the cartographic form of abstraction is the concrete space (the order of the relative positions of objects and phenomena in reality, both in relation to each other and in relation to the spatial reference system), whereas the subject of the cartographic form of generalization is the cartographic content. Like the cartographic form of abstraction, the cartographic form of generalization depends on changes in the map scale.

This work utilizes a previously developed and repeatedly tested cartographic methodological system, specifically cartographic generalization, in the process of mapping the state of the natural environment (in this case, landscapes) from a case study in Georgia. Mapping and GIS analysis of the environmental conditions dynamics are conducted in three stages: 1) Mapping the dynamics of landscape-ethological phenomena (LEP), 2) Mapping landscape-ethological situations (LES), 3) Mapping the dynamics of landscape-ethological scenarios (LESC). The work explores the features of dynamic generalization and temporal scale. It also calculates the degree of spatiotemporal synthesis that occurs during the mapping of environmental dynamics.

This work is methodological in nature and can be used in mapping the dynamics of natural environment states in other mountainous regions.

ABSTRACT. This paper reviews recent methodological advances in the use of remote sensing for mapping agricultural landscapes and water resources. Accurate and timely information on land use dynamics and hydrological variability is essential for sustainable resource management, food security, and climate change adaptation. The increasing availability of Earth observation data from the Copernicus Programme, particularly from Sentinel-1 and Sentinel-2, has significantly improved monitoring capabilities across spatial and temporal scales. The study synthesizes key methodological approaches, including multi-temporal analysis, machine learning classification, and multi-source data fusion, which enable enhanced mapping of crop dynamics, irrigation patterns, surface water bodies, and hydrological variability. The integration of optical and radar data is highlighted as a critical factor for improving robustness under diverse environmental conditions. The paper also discusses current challenges such as data integration, model transferability, and validation constraints, and outlines future directions for developing scalable and operational Earth observation-based monitoring systems.

ABSTRACT. The earthquakes that hit Kahramanmaraş on 6 February 2023 not only caused severe physical destruction but also significantly altered urban energy consumption patterns. This study analyses the post-disaster dynamics of natural gas consumption in Malatya by integrating time series forecasting with Geographic Information Systems (GIS) to explicitly capture both the temporal and spatial impacts of the event. Monthly data from 2019 to 2023 were used, including total natural gas consumption, subscriber counts by customer category, mean temperature, Heating Degree Days (HDD), and Cooling Degree Days (CDD), public holidays, and lagged consumption variables. A Long Short-Term Memory (LSTM) neural network was employed to model pre-earthquake demand behavior and to generate a counterfactual scenario representing consumption levels under disaster-free conditions. Seasonal patterns and long-term dependencies were incorporated through trigonometric transformations, moving averages, and lag-based feature engineering. The analysis indicates that natural gas consumption during the February–December 2023 period would have reached approximately 329.5 million m³ under normal conditions. In contrast, the observed consumption was 246.6 million m³, corresponding to an overall decline of nearly 25%. The magnitude of this reduction varied markedly across customer groups, with industrial zones exhibiting relatively limited decreases compared to residential and service-sector subscribers, which experienced substantially sharper declines. Beyond the temporal assessment, GIS-based spatial analysis enabled neighborhood-level mapping and cartographic visualization of demand changes and recovery patterns. The results reveal that post-disaster normalization did not occur uniformly across the city, with distinct spatial clusters exhibiting prolonged suppression of consumption. These spatial disparities highlight the significant role of cartographic representation and spatial analytics in interpreting post-disaster energy demand behavior. Overall, the study demonstrates that large-scale disasters introduce both temporal disruptions and spatial heterogeneity into urban energy systems. The combined use of LSTM-based forecasting and GIS provides a robust and transferable framework for spatially explicit impact assessment, supporting post-disaster tariff planning, infrastructure recovery prioritization, supply security, and risk-informed decision-making for natural gas distribution companies.

ABSTRACT. Within the broader context of European and global heritage conservation, the some cultural heritage sites of Bulgaria are located in regions having complex geological settings. The Eastern Rhodopes are among the most archaeologically significant areas in the country, into which found are rock‑cut monuments, sanctuaries, fortifications, and burial complexes. Geologically, in the region of Eastern Rhodopes volcano‑sedimentary sequences and high‑grade metamorphic complexes are differentiated. In the mentioned region documented are faults, shear zones, and fracture systems that document several deformation phases. On the other hand the remotely sensed data along with the processing techniques included modern Geographic Information Systems (GIS) offer reliable monitoring the spatial and temporal changes of the landscape. For this study using the Spectral Angle Mapper (SAM) classification approach a time series of optical multispectral data from the Sentinel‑2 and Landsat satellite missions were elaborated to obtain vegetation trajectories in the investigated area.For surface deformation and slope instability multi‑temporal Synthetic Aperture Radar (SAR) data from the Sentinel‑1 mission were also processed using the ESA Sentinel Application Platform (SNAP), which provides a well-established pipeline framework for interferometric processing. Information on surface motion was obtained implemting Small Baseline Subset (SBAS) approach for SAR analysis, which reduces the temporal decorrelation and atmospheric disturbances. The combined use of SBAS InSAR results with SAM‑derived land cover classifications, geological maps, and other additional data within a GIS allowed assessing the landslide susceptibility in the area of cultural heritage site Vishegrad fortress situated on a steep rocky hill above the Arda River thus contributing to evaluate the geological risk.

ABSTRACT. Water erosion is a major geomorphological process contributing to landscape degradation and requiring an interdisciplinary approach for its monitoring and quantification. This study presents an integrated UAV- and GIS-based analyses for detecting and measuring topographic changes on an eroded slope. A gully located in a low-mountain environment in the Eastern Rhodopes, Bulgaria, was surveyed using an unmanned aerial vehicle (UAV) in October 2020, April 2021, and March and November 2023. The obtained multi-temporal 3D point clouds were processed and analysed using to assess erosion and sediment accumulation. Changes in gully morphology were further evaluated through high-resolution digital terrain models analyses in GIS environment. Most of the investigated elevation changes range between −5 and +5 cm for the period April 2021–March 2023, which is consistent with field measurements obtained from ground control markers. The cross-sectional and longitudinal profiles reveal sediment accumulation of up to approximately 5 cm in the upper gully channel and up to 12 cm in the downstream sections, resulting from the downslope transport of weathered material from unvegetated slopes and along the gully profile. The geospatial analysis also indicates a lateral shift of the main flow line toward the left gully slope. Overall, the results demonstrate the high potential of integrated UAV–GIS thechnologies for detailed monitoring and quantification of water erosion, and can contribute to a better understanding of water erosion processes at a local scale.

ABSTRACT. This work presents a methodology for reconstructing and visualizing borehole trajectories in a three-dimensional environment with integration into a Geographic Information System (GIS). The approach aims to convert directional drilling data into continuous 3D spatial representations suitable for visualization and spatial analysis. The reconstruction process is based on directional drilling parameters including measured depth, inclination, and azimuth, which are used to compute the spatial coordinates of the borehole trajectory. The methodology was implemented using Python and scientific computing libraries to process the data and generate 3D representations of various borehole geometries such as vertical, horizontal, and multilateral wells. The reconstructed trajectories were compared with the reference dataset in order to evaluate the consistency of the generated 3D models. The resulting trajectories were then exported and visualized in a GIS environment, demonstrating the capability of integrating subsurface trajectories with surface spatial data. This study represents a preliminary stage of a broader research project aimed at developing advanced tools for subsurface trajectory modeling and spatial visualization, with future validation planned using real drilling datasets.

ABSTRACT. In the semi-arid regions of Algeria, where rainfed cereals dominate agricultural production, climate change is increasingly constraining winter wheat phenology. Climate records indicate an increase in mean air temperature of approximately +1.5 to +2 °C since the 1980s, accompanied by a decline in annual precipitation of 15–30%. Using Sentinel-2 time series data spanning 2016–2025, this study confirms these long-term climatic trends and evaluates their impacts on vegetation phenology. Pearson correlation analysis was applied to quantify relationships between the onset of effective rainfall (OER), phenological metrics, and precipitation variability using the Standardized Precipitation Index (SPI) computed at 3- to 6-month timescales (SPI-3 to SPI-6), representing short- to medium-term moisture conditions relevant to crop development. Results reveal moderate positive correlations (r ≈ 0.5) between SPI-3 and the start of season (SOS), indicating that short-term precipitation anomalies strongly influence vegetation green-up timing. In contrast, stronger correlations (r ≈ 0.7) were observed between SPI-6 and the end of season (EOS), suggesting that cumulative moisture availability predominantly regulates late-season phenological dynamics. These findings highlight the scale-dependent sensitivity of winter wheat phenology to precipitation variability and emphasize the importance of multi-timescale drought indicators for improving climate impact assessments in semi-arid agroecosystems.

ABSTRACT. As one of the most vulnerable natural environments, wetlands provide a large number of crucial ecosystem functions and services for humans and nature. Wetland mapping in the region of Oran using medium-resolution Landsat imagery is often affected by spectral confusion. Features such as open water, moist soils, temporarily flooded areas, dense aquatic vegetation, urban surfaces, and terrain shadows can easily produce classification uncertainties. In addition, the 30-m spatial resolution of Landsat data constrains the detection of small wetlands and exacerbates mixed-pixel effects. To address these limitations, a Support Vector Machine (SVM)-based classification framework was implemented using Landsat images acquired in 1987 and 2024, representing contrasting spectral conditions. The classification process was conducted in two stages, with and without an optimized training dataset, and both outcomes were compared. The enhancement consists of image composition considering seven bands, including three original Landsat spectral bands (near-infrared, red, and blue) and four supplementary features: the Normalized Difference Vegetation Index (NDVI), the Perpendicular Impervious Surface Index (PISI), elevation data derived from the SRTM digital elevation model, and a single texture band. The results demonstrate a substantial improvement in classification accuracy following the integration of the enhanced feature set as training data. A visual comparison of SVM outcomes before and after data composition clearly shows the performance of the proposed methodology. In addition, the Kappa coefficient was computed with values increasing from 0.56 to 0.80 in 2024 and from 0.50 to 0.85 in 1987. This wetland mapping approach provides a robust framework for monitoring spatio-temporal wetland dynamics and supports informed decision-making for sustainable management and conservation of wetlands in the Oran region, increasingly threatened by urban expansion and climatic variability.

ABSTRACT. Heavy rainfalls or rapid snowmelt are widely recognized as main triggering factors for flood occurrences in some cases being beyond the natural drainage capacity of a single catchment. This is of particular importance in mountainous regions characterized by rugged topography, where numerous fluvial valleys have developed through prolonged erosional processes due to mentioned favorable conditions for their development. Additional factors engraving the situation are the active tectonics, the steep and eroded slopes, the limited vegetation cover subsequent from poorly applied afforestation measures and poor riverbed maintenance as well as loose soil cover result from loose colluvial and proluvial deposits. All said is valid to high degree for territory of the Eastern Rhodopes which makes the region naturally prone to rapid flood events that cause geomorphological changes. Such events are highly related to the increase of the geological hazard in this area e.g. landslides development. The primary objective of this paper is to propose general framework for assessment of the mentioned hazard based on freely available remotely sensed data processed within GIS environment producing as final result hazard maps. The principal source of the data and products originate from the satellites missions Sentinel-1 and Sentinel-2 of the Copernicus Programme complemented by additional geospatial datasets such as bedrock and soil maps, soil moisture maps, digital elevation models. The outputs of this research are hazard maps illustrating the susceptibility to flood-triggered geological processes, thus providing better understanding and improved management of natural hazards in the region.

ABSTRACT. The integration of artificial intelligence and remote sensing data into Geographic Information Systems (GIS) is transforming the capabilities of Earth observation and spatial analysis. Within this context, accurate deformation monitoring using Differential Interferometric Synthetic Aperture Radar (DInSAR) remains challenged by spatial and temporal decorrelation in interferograms. Recent studies (e.g., Zhou et al., 2023) have shown the potential of deep learning to mitigate atmospheric and topographic disturbances in InSAR time series. In this work, we propose an innovative workflow that combines optical–SAR data fusion, a hybrid U-Net/Transformer architecture, and full GIS integration, aiming to generate enhanced coherence maps and extract reliable deformation time series. The processing chain includes precise geometric co-registration, extraction of optical features (NDVI, NDBI, texture indices), multimodal patch construction (SAR amplitude + optical bands + indices + raw coherence), and a dual-branch deep model that predicts coherence through nonlinear inter-modal correlations, inspired by the DeepInSAR framework. Furthermore, self-supervised learning and Physics-Informed Neural Networks (PINNs) are incorporated to reduce dependency on labeled data and to enforce physical consistency (phase periodicity, signal energy conservation). These mechanisms improve model generalization across different sites and extreme events, in line with recent developments in physics-informed geohazard modeling (Moeineddin et al., 2023). The predicted coherence maps act as quality indicators for guiding phase unwrapping and DInSAR inversion, significantly improving deformation accuracy and reliability. Final products are integrated within a GIS environment (GeoPackage / WFS / WMS) for multi-layer spatial analysis (enhanced coherence, displacement, DEM, land cover), vulnerability zoning, and decision support in land and infrastructure monitoring. This unified approach, supported by recent advances in multimodal data fusion and AI-based geospatial analytics (Sensors, 2025), illustrates the convergence of Earth observation, deep learning, and geoinformatics for more accurate and operational deformation mapping.

ABSTRACT. Air pollution caused by fine particulate matter and the increasing emissions of greenhouse gases is among the leading environmental problems with a direct impact on public health, quality of life, and sustainable development. Their effective assessment and management require not only reliable data but also appropriate tools for spatial analysis and visualization. In this context, cartography and geographic information systems (GIS) play a key role in presenting, analysing, and interpreting environmental indicators.

This paper focuses on the development of interactive online cartographic products that visualize the current state and trends of air pollution by fine particulate matter and greenhouse gas emissions across different countries and regions over a selected period. Particular emphasis is placed on demonstrating the capabilities of modern web-based cartography as an effective tool for the analysis and communication of environmental information for both specialists and the general public.

The methodological approach is based on the use of statistical data obtained from reliable international sources, which have been pre-validated, standardized, and spatially linked to administrative units. The data were processed and analysed in a GIS environment. A comparative spatial analysis was carried out in order to identify differences and trends in pollution levels among individual countries and regions.

The results are presented through interactive web maps and applications that enable visual comparison, easier interpretation, and the exploration of spatial patterns of environmental indicators, demonstrating the high potential of web-based cartography as a tool for environmental analysis, informed decision-making, and raising public awareness of air pollution issues.

ABSTRACT. Pluvial flooding, also known as surface water flooding, occurs when intense rainfall overwhelms the capacity of urban drainage systems and natural infiltration processes, leading to the accumulation of water on the land surface. This study aimed at prediction of pluvial flow depth and velocity using the rain-on-grid method in MIKE+ hydraulic model. The studied domain was represented by an upstream part of the Gidra River basin (1.51 km2), where the Píla municipality (western Slovakia) is located. We simulated an extreme 58-minute rainfall event from 7 June 2011, which hit also wider surrounding area and caused a destructive flash flood on the Gidra River. The time series of rainfall intensities were recorded at the nearby Častá rainfall station, which is operated by the Slovak Hydrometeorological Institute. Another high-resolution inputs in the 2D MIKE+ model were airborne laser scanned digital elevation model with 1 m resolution and land cover with the corresponding Manning's roughness coefficients and infiltration losses. Based on the results, the largest flood extent was recorded after 32 minutes of rainfall, when 1.1 km2 of the domain was represented by non-flood areas with flow depths <0.005 m. The flooded areas accounted for 0.41 km2, out of which most of the flooded pixels belonged to the flow depth interval from 0.01 to 0.05 m (53.5%), followed by the interval 0.005 – 0.01 m (18.0%). In case of flow velocity, the highest share of pixels was included in the flow velocity interval from 0.00 to 0.50 m/s (71.4%), followed by the interval 0.50 – 1.00 m/s (20.2%). The results are important for local flood risk management, improving resilience and adaptation to floods in the face of growing human impact on landscape and climate change. Acknowledgment: Funded by the EU NextGenerationEU through the Recovery and Resilience Plan for Slovakia under the project No. 09I03-03-V03-00085.

ABSTRACT. In recent years, interest in risk modeling and its spatial distribution has grown across all European countries, driven by the escalation of military and hybrid threats from Russia. Since Russia's full-scale invasion of Ukraine, any research related to the application of spatial analysis in defense operations has also garnered significant attention. This paper presents two examples of how spatial analysis can support defense efforts. The first example concerns hybrid activities on the Polish-Belarusian border, where since the summer of 2021, large numbers of migrants have been intentionally lured to Belarus and subsequently directed to the eastern border of the European Union to cross it illegally. Research on identifying infiltration routes used by illegal migrants across the Polish-Belarusian border will be presented. By defining three components: walk (terrain traversability), hide (concealment opportunities), and security (level of border protection), areas particularly vulnerable to this phenomenon were identified, and potential infiltration paths were mapped. These findings were compared with Border Guard data regarding the detention of illegal migrants. The results demonstrated a high level of consistency between the model’s outputs and official Border Guard records. The second example involves the use of the potential method to assess the level of threat to the civilian population in the event of an attack on critical infrastructure. As part of this research, a critical infrastructure coefficient, threat potential, and population potential were determined. Based on these metrics, the ratio of threat potential to population potential and the specific population threat potential were calculated. This allowed for the identification of areas with varying degrees of risk to civilians in the event of a missile attack on critical infrastructure facilities. Due to the sensitive nature of the subject matter, the presentation will emphasize methodological issues, keeping the visualization of actual results to a minimum or incorporating elements of data masking/obfuscation for the output data.

ABSTRACT. The study presents the development of an AI-driven agent designed to automate the analysis of geospatial data and provide decision-makers with precise, context‑aware information about terrain conditions. The proposed solution integrates a large language model with a dedicated GIS processing engine, enabling both semantic interpretation of analytical tasks and execution of advanced spatial operations. The system utilizes topographic datasets, including elevation models, land cover, transportation networks, and environmental constraints, to generate structured situation reports and terrain assessments. The research investigates two main assumptions: an AI agent can effectively integrate heterogeneous geospatial datasets and generate interpretable, decision‑supporting recommendations, and such an agent can significantly reduce the time required for situational analysis in transportation planning or field‑operation scenarios. A prototype implementation was developed and tested on representative spatial analysis tasks, such as route feasibility evaluation and identification of terrain obstacles. Preliminary results indicate that the AI agent is capable of producing coherent analytical outputs combining descriptive reasoning with quantitative GIS indicators. Moreover, test cases demonstrate a measurable acceleration of the analysis workflow compared with traditional manual GIS procedures. The contribution of this work is a functional prototype of an intelligent, agent‑based system that enhances decision‑making processes by leveraging automated interpretation and summarization of geographic information.

ABSTRACT. The GeoFireData project aims to systematically collect, integrate, and process geospatial data to support advanced wildfire analysis and future modeling efforts. The first phase of the project focuses on integrating heterogeneous geospatial sources – UAV-derived DEM/DSM, national forest inventory data, Ministry of Agriculture and Food (MoAF) orthophotos and DSM, and specialized layers for agricultural areas from the Land Parcel Identification System (LPIS) – into a unified analytical layer. The primary objective is to derive a statistically representative sample from the forest area for general ecosystem description and analysis, using GIS tools and statistical methods. Forest stand data layers were processed, including attributes such as forest type, age structure, mean tree height, mean diameter, and canopy height (CH). Terrain characteristics – elevation, slope, and aspect – were also incorporated. Using five stratification criteria (forest type, age class, elevation, slope, and canopy height) and GIS tools, the spatial distribution and characteristics of a representative sample were established for the “Vratsa Balkan” study area (2,132.4 ha). The sample was derived using a stratified random sampling method with proportional area allocation, in which the number of sample plots in each stratum is proportional to the stratum's accessible area within the study zone. The stratification yielded 50 actual combinations, out of a theoretical maximum of 162. A two-stage area-based generalization was applied: strata below 10 ha were merged with the nearest significant stratum while preserving forest type and age class; strata below 20 ha were subsequently merged following the same hierarchical principle, resulting in 12 final strata. Proportional allocation determined 27 sample plots (0.5 ha each). Sample representativeness was verified using a two-sample Kolmogorov-Smirnov test for elevation, slope, and CH distributions, a χ² test for the proportional representation of forest types and stand age classes, and a spatial coverage analysis – all results confirmed the representativeness of the sample across all tested indicators.

ABSTRACT. This paper investigates the contemporary design approaches in Landscape architecture and their role in improving the quality, accuracy and efficiency of design solutions. The development of digital technologies provided a wide range of specialized software tools for landscape architects. They support all design stages from feasibility studies and conceptual design to the visualization and implementation of the project. Providing spatial data as a new technological approach to landscape architecture is a key importance for vertical planning. These include vector data from ground surveys as well as raster data from terrestrial measurements, UAV, aerial and satellite remote sensing data, 3D and point cloud solutions of the territory. This data is successfully integrated into a GIS environment and through specialized software products for landscape design. The article reviews and compares the use of basic categories of software products, including CAD programs for technical drawing, 3D modeling and visualization software, GIS systems for spatial analysis, and BIM platforms for integrated design. Programs like AutoCAD, SketchUp, Lumion, Twinmotion, Revit, and ArcGIS/City Engine allow the creation of accurate plans, detailed 3D models, and photorealistic images. They facilitate the analysis of relief, vegetation, and functional zones in the vertical planning of territories. The study is focused on selecting software solutions that will enable realistic presentation of landscape projects and better communication with clients, investors, and institutions. These software products optimize the process of creating digital twins and decision-making, reducing the risk of errors in project implementation. Conclusions and recommendations have been made for the use of modern software products as a key tool in landscape architecture and a technological basis for creating sustainable, functional, ecological, and aesthetically high-quality outdoor spaces.

ABSTRACT. The pastoral landscapes from the Carpathian Mountains are the result of a continuous, traditional and sustainable economic system that has shaped the culture and identity of Carpathian nations. As for Romania, a Carpathian nation by definition, pastoralism represents a cultural reference and, as such, its proxies are seen as ancestral users of space. The cultural reference extents also to the large guardian dogs (LGD’s) present at the sheepfolds to protect the flocks. The dogs are seen as an integral part of the pastoral system providing security in an environment populated with large carnivores. However, through mountain tourism, a new category of landscape users is making its presence known more and more. From late April until the end of September, many outdoor recreational activities in the mountains overlap with the pastoral calendar, creating in some areas a potential for conflict between these two distinct categories of landscape users. The potential for conflict arises when the LGD’s come into contact with the tourist by using the same spaces intersecting the tourist trails, with incidents happening over the years. The main research hypothesis of this study states that these spaces of potential encounter are determined by the physical, environmental and logistical characteristics of the land used in the grazing system and thus by the LGD’s. The flock and the dog’s presence are not randomly distributed but follow a spatial pattern characterized by key physical and environmental predicators with different degrees of significance. The results can contribute in the process of a future predictive model build-up that can be used to map the vulnerability of tourist trails, offering a new perspective on spatial trail management in mountain areas.