ABSTRACT Landslide susceptibility analysis is a common approach for assessing the risk of slope failures in areas affected by rainfall. During the rainy season, the city of Salvador, Brazil, often experiences slope failures and landslides. This research applied the infinite slope model using GIS as a screening method to assess landslide susceptibility in Salvador. The GIS-based model calculated the slope stability factor of safety by integrating spatial data, including soil parameters, slope angles, and depths of water penetration, to identify areas prone to landslide during rainfall events. Soil parameters were obtained from laboratory analyses of 392 samples collected from various locations in Salvador, which is considered an administrative subdivision organized by districts. Four water penetration depths (h = 1, 2, 3, and 4 metres) in the soil were simulated, and their influence on the slope factor of safety was analyzed. Susceptible areas for landslide were obtained by calculating the thickness of the unstable material or the weight of the surface layer as a function of water penetration depth. For h = 1 m, the total landslide-susceptible area was 1.24 km2, while for h = 4 m, the area increased to 42.75 km2. The 4-Liberdade/São Caetano district showed the largest area susceptible to landslides. The results from the GIS model were consistent with the historical landslide data and landslide susceptibility maps from Serviço Geológico do Brasil (CPRM – Brazil Geological Service) official dataset. The integration of the infinite slope model into GIS provided a valuable tool for improving landslide hazard management by screening landslide-susceptible areas at both regional and local scales.

ABSTRACT In recent years, in-person shoppers have increasingly demanded systems that minimize both travel costs and product prices. This research aims to design and implement an open-source, web-based location-based service (LBS) system to assist customers in locating stores that offer desired products at the lowest overall cost. The study focuses on Tehran, Iran, as a case study area. Unlike existing LBS and vehicle routing studies that focus on delivery or single-destination navigation, this work introduces the first open-source Web-GIS framework specifically designed for multi-stop, price-sensitive in-person shopping trips, a significantly underexplored consumer segment in dense urban environments of developing countries. The system utilizes open-source technologies to reduce development and operational costs. To optimize the shopping route, a heuristic store allocation method is proposed to reduce computational complexity from factorial O(n!) to polynomial O(n * m), where m is the average number of stores per product. Subsequently, the Travelling Salesman Problem (TSP) is solved using a Genetic Algorithm, leading to a significant reduction in total travel distance. In one test case with six product locations, the optimized route reduced the distance from 195 km (random order) to 146 km. Finally, for turn-by-turn navigation, the A* algorithm is employed, showing a 70% average reduction in response time compared to Dijkstra’s algorithm, with only minimal accuracy differences. The results indicate that the proposed framework effectively reduces shopping costs and improves efficiency for urban users. Moreover, the use of open-source Web-GIS technologies ensures scalability, maintainability, and low-cost implementation for future developments.

ABSTRACT Machine learning for remote sensing imaging relies on up-to-date and accurate labels for model training and testing. Labelling remote sensing imagery is time and cost intensive, requiring expert analysis. Previous labelling tools rely on pre-labelled data for training in order to label new unseen data. In this work, we define an unsupervised pipeline for finding and labelling geographical areas of similar context and content within Sentinel-2 satellite imagery. Our approach removes limitations of previous methods by utilizing segmentation with convolutional and graph neural networks to encode a more robust feature space for image comparison. Unlike previous approaches we segment the image into homogeneous regions of pixels that are grouped based on colour and spatial similarity. Graph neural networks are used to aggregate information about the surrounding segments enabling the feature representation to encode the local neighbourhood whilst preserving its own local information. This reduces outliers in the labelling tool, allows users to label at a granular level, and allows a rotationally invariant semantic relationship at the image level to be formed within the encoding space. Our pipeline achieves high contextual consistency, with similarity scores of SSIM = 0.96 and SAM = 0.21 under context-aware evaluation, demonstrating robust organization of the feature space for interactive labelling.

ABSTRACT Emerging forms of spatial data, such as sensor-based location data collected by mobile apps (MPD), have shown promise in enhancing or replacing traditional analytical methods. MPD provides detailed and timely information at lower costs, making it valuable for urban and transport planning. Multi-source MPD also offers advantages, such as greater representativeness. However, the high volume, variable frequency, and heterogeneity pose significant challenges in data processing and analysis. This study investigates the potential of multi-source MPD to estimate traffic volumes at the street level, focusing on Glasgow city-region. The evaluation covers a range of processing approach combinations. These include three MPD spatial extraction techniques —Simple Buffer (SB), Connected Street Buffer (CB), and Entire Street Buffer (ESB) —to filter relevant vehicular movement data. Additionally, four processing approaches are considered: raw counts (A1), simplified counts with commuter assumptions (A2), detailed geometric analysis (A3), and map matching (A4). The results are compared to manual traffic counts from the Department for Transport (DfT). The findings reveal that raw MPD counts (A1) lead to important biases due to uneven data volume per user. Simplified counts (A2) improve accuracy but still capture non-vehicular activities. Map matching (A4) introduces substantial improvements compared to A1 and A2. However, this approach proves less effective than A3. The geometric-temporal approach (A3) offers the most accurate estimates by incorporating movement inferences. Buffer size and built environment factors influence the methods’ performance, highlighting the need for localised buffers and built environmental controls. This research finds moderate potential for multi-source MPD to supplement traditional traffic systems, providing cost-effective and detailed traffic insights at least from a spatial variability perspective, but not for cross-temporal comparisons. Future studies can extend MPD applications for active travel and extend the analysis to other cities.

ABSTRACT To address the challenges of accurately extracting target features from complex scenes in UAV remote sensing imagery and the susceptibility of small objects to being obscured by noise, this paper proposes a lightweight detection algorithm, RE-YOLO, based on YOLOv8n. First, a multi-scale convolutional module named RFCSConv, which integrates channel and spatial attention mechanisms based on Receptive Field Attention Convolution (RFAConv), replaces the original convolution layers. This enhances feature selection and fusion at multiple scales. Second, the Efficient Squeeze-and-Excitation Module (ESEModule) is introduced into the backbone to strengthen feature representation while reducing computational overhead. Lastly, a composite loss function called Win-IoU, combining Wise-IoU (WIoU) and Inner-IoU, is proposed to dynamically adjust gradient contributions based on anchor quality. Experimental results on the VisDrone2019 dataset demonstrate that RE-YOLO achieves 29.7% mAP@0.5 with only 3.2MB of parameters and a real-time speed of 150 FPS. The algorithm also generalizes well across the HRSID and CARPK datasets, achieving 91.8% and 94.3% mAP@0.5 respectively.

ABSTRACT This paper reviews representation learning for geospatial data, focusing on methods for automatically extracting meaningful features from diverse data types. By simplifying tasks and improving accuracy, representation learning has emerged as a powerful tool for geospatial analysis. Due to its generalizability and scalability, representation learning provides an effective approach to processing geospatial data, which is inherently diverse and unstructured. We summarize the representation learning methods for different geospatial data types, including locations, points of interest (POIs), trajectories, spatial interactions, remote sensing imagery, and street view imagery. Treating each data type as a distinct modality, we emphasize the potential of multi-modal representation learning to advance the understanding of geographical phenomena and propose an LLM-guided framework as a potential solution. The review concludes by highlighting the need for further research to improve multi-modal data alignment and enhance the interpretability of feature representations, particularly in complex and dynamic geographical environments.

ABSTRACT Urban studies have a long tradition of examining the regularity of urban structure by modelling urban population density functions and probing the theoretical or behavioural foundation behind it. Previous studies commonly used census data in areal units such as census tracts or census block groups, which varied a great deal in area size and shape and led to the zonal and scale effects, commonly referred to as the modifiable areal unit problem (MAUP). This study uses big data of individual vehicle trips in Tampa, Florida, to define the precise population and employment distribution locations, and then aggregates them with uniform areal units such as squares, triangles, and hexagons to examine and mitigate the scale and zonal effects. Both monocentric and polycentric models are employed in the analysis of urban population density functions. The results suggest that the exponential density function remains the best fitting monocentric function in most areal units including census units and designed uniform units. The polycentric model reveals two centres (downtown and University of South Florida) exerting influences on the areawide population density pattern. The zonal effect is not significant in the designed uniform units, but the scale effect remains evident in all areal units.

ABSTRACT The Arctic region is undergoing significant changes in maritime accessibility. This study investigates observed ship trajectories from 2013 to 2020 to demonstrate the recent trends of Arctic traffic. A notable surge in maritime activities has been observed, particularly during summer months, driven by economic interests and the increasing popularity of existing routes. Unique patterns in the northern Barents Sea have been observed where ships favour different routes based on seasonal ice conditions. Another contribution from this work is the validation of the Arctic Traffic Accessibility Model (ATAM) using the observed ship traffic data. Results show that the ATAM model underestimates the accessibility and vessel travel speed. This is largely due to outdated model parameters. The predefined ice multipliers and calculation of ice numerals used in the ATAM may not accurately reflect real-world conditions.

ABSTRACT The Andes region has a rich history of environmental and human interactions that has shaped the landscape for millennia. Our study quantified the land cover changes in the districts of Huanta and Luricocha after the human population abandoned the region due to the armed conflict with Peruvian authorities (1980 – 1990) and their progressive return (1990 – 2020). We analysed satellite images of various resolutions, including Landsat-5 for the years 1986, 1990, 2000 and 2010; Landsat-8 for the year 2020, Sentinel-2 for the years 2017 and 2020; PeruSat-1 for the years 2017 and 2020, to assess the temporal change of land cover and landscape metrics. The classification was based on the pixel approach for Landsat images and Sentinel-2 with an average accuracy of 88.29% and 92.13% respectively, and an object-based (GEOBIA) approach for Perusat-1 images, with an average accuracy of 87.75%. The dominant coverages are high Andean grasslands, scrublands and croplands. According to the obtained data, high Andean grasslands and scrublands have decreased, which is associated with an increase in population over time. Likewise, there has been an increase in the area of lagoons and the expansion of nearby wetlands due to the construction of dams, driven by an increase in demand for water for human consumption and agriculture. However, areas with degraded soils, rocky outcrops, and forest plantations have expanded. The results suggest that abandoned agricultural lands did not recover quickly enough and more research and action is needed for restoration of the high-mountain ecosystem. Landscape metrics such as Edge Density (ED), Euclidean Distance to Nearest Neighbor (ENN) and Aggregation Index (AI) were calculated for different land cover (LC/LU) to characterize composition and structure, showing their variation at different spatial resolutions. This indicates that increasing spatial resolution makes it possible to better define the landscape characteristics.