Origin-Destination (OD) matrices are mathematical objects usually integrated inside transport models or treated as inputs for quantitative analysis. Trying to map an OD matrix is challenging due to the density and size of its data, and this often makes graphical representations unintelligible. Traditional approaches rely on partial matrices or aggregate indicators, limiting the spatial and relational insights that can be derived. This paper presents a novel analysis method that retains both the aggregate numerical quantities and the spatial relations between origins and destinations, providing a multidimensional representation of large-scale mobility patterns. The proposed approach is applied to two case studies: the 2011 commuting for work and study matrix of Italy, and the 2021 work commuting matrix of England and Wales. The Italian matrix has also been analyzed diachronically by mapping changes occurred between 1991 and 2011. Each case study demonstrates the method's ability to reveal mobility trends and is used to describe and interpret the spatial structure of the territory and inform policies. The paper also discusses the method's broader applicability and limitations, emphasizing its potential to enhance the interpretability of OD matrices across different contexts and using different data sources.
We introduce ring maps, a new augmented reality (AR) mapping paradigm that supports pedestrian navigation in urban environments while balancing map complexity, environmental occlusion, and spatial awareness. Ring maps position landmark symbols in a circular layout around the viewer based on direction and distance. We situate ring maps alongside other AR map types in a design space that considers cartographic generalization, anchoring, and planimetric accuracy. To understand the trade-offs of ring maps, we created prototype ring maps and validated them in two studies. Our first study, conducted in virtual reality (VR), used navigation tasks to evaluate spatial reasoning along with an attention task to evaluate cognitive load. The second study, conducted using an AR headset in a range of outdoor and indoor environments, examined the experience of using ring maps for real-world navigation. We find that ring maps and three-dimensional maps perform comparably for point-to-point navigation tasks, but also complement one another. Participants commented that three-dimensional maps support planning and decision-making, while ring maps support active navigation with clearer distance and angle estimation. Finally, we identify use cases for ring maps, and the potential for improving and combining ring maps with other augmented reality map types.
Depth contours are essential chart features, portraying seabed morphology and delineating the depth areas used by navigation systems to assess route safety and trigger alarms. Their portrayal must simultaneously satisfy operational safety and visual clarity. For navigational safety, contours must be displaced only toward the adjacent deeper-depth area, ensuring that charted depths never appear deeper than the source bathymetry. Contours must also remain easily interpretable and adhere to Electronic Navigational Chart data constraints, such as the minimum length of line segments. Despite numerous efforts, existing methods suffer from limitations, including parameters difficult for cartographers to interpret, unnatural shapes, uncontrolled contour displacement, and increased vertex density. This paper introduces a side-selective line-simplification method that enforces the safety constraint and controls line displacement, segment length, and bend geometry using parameters expressed in millimeters at the target scale, to produce visually coherent and application-compliant contours. Tests conducted on six testbeds demonstrate a reduction in geometric complexity while preserving the structural form of contours. Comparisons with representative alternatives highlight the method's advantages in reducing data volume and producing scale-appropriate, consistent outputs. The side-selective method forms an essential building block toward automated workflows for the cartographic generalization of depth contours and other relevant chart features.
Geographic masking is a key technique to protect individual privacy when sharing spatial point data. Among probabilistic approaches, donut masking is widely used because of its conceptual simplicity and ease of implementation. Whereas most previous studies have focused on average trends linking masking parameters to privacy and analytical utility, this study highlights the often-overlooked role of execution-level variability: differences in outcomes from repeated masking executions under the same parameter settings. Using synthetic crime data and census-based household distributions from Suginami Ward in Tokyo, we applied donut masking across a range of displacement radii and assessed its effects on both spatial k-anonymity and the preservation of spatial analytical results derived from K-function analysis. To quantify analytical preservation, we introduced two novel metrics based on p-value profiles: the Euclidean distance and consistency rate. Our results showed that while a larger displacement increased privacy, it also introduced substantial variability in the analytical outcomes. However, some executions, even under strong masking, exhibited high fidelity in the analysis results. Comparisons with other masking methods suggest that combining execution-level selection with probabilistic masking can help balance privacy protection and analytical preservation. Based on these findings, we proposed two practical strategies for selecting favorable executions: fixed-count selection and threshold-based iterations. Processing time assessments confirmed the feasibility of this approach in standard computing environments. Rather than viewing geographic masking as a fixed-parameter process, this study reframes it as a flexible selection problem that enables more effective anonymization with minimal analytical compromise.
This paper is based on the CaGIS Early Career Scholar keynote I gave at the American Association of Geographers (AAG) meeting for Cartography and Mapping Specialty Group (CMSG) in Detroit, MI, in March 2025. In this paper, I interweave my personal journey in cartography from initial enthusiasm to disillusionment to a reignited excitement for maps. I also illustrate how maps are becoming even more center stage in 2025 with their use as powerful political artifacts that shape narratives. While the Trump administration offers the potential for despair, my goal is to illustrate that we can remain hopeful and do the hard (cartographic) work amid disruption and chaos. My past research on how maps are used in the media serves as a framing for the current moment and the use of maps across the political spectrum. I also identify the ways in which scholars at the AAG were doing the hard work of creating and sharing hopeful maps using storytelling and emotions in place-based mapping, and through speaking truth to power in the wake of oligarchy, control, and artificial intelligence through mapping as resistance. I conclude with an outlook on the future of maps and mapping.
Map emotional expression refers to the process of endowing maps with emotional qualities through visual design. Emotional expression consistent with the map context can enhance the user's emotional engagement, while the potential of map emotional expression remains insufficiently explored. Color plays a central role in facilitating emotional communication with map users, appropriate color style design has the potential to improve map's emotional expressiveness. In this study, a map color style design model is proposed to support emotion-oriented color style design and enhance map emotional expression. The model introduces a Map Emotion Dataset for extracting emotional color manifold features, and constructs emotional coloring constraint functions to effectively guide the iterative process to generate map color styles that better align with target emotion. The proposed model offers a framework for enhancing the emotional expressiveness of maps through color style design. The perception evaluation results show that the color styles generated by the proposed model exhibit more stable performance than CycleGAN and manual design in different emotion categories and are generally well accepted by readers. Furthermore, in the evaluation of visual quality, the method demonstrates good performance in scene appropriateness, hierarchical clarity, and color rationality.
This study presents an original methodological contribution by developing and validating a replicable framework for the generation of vehicle trafficability maps through refined spatial modeling techniques, relying on biophysical, infrastructure, and meteorological variables. By integrating remote sensing data, Geographic Object-Based Image Analysis (GEOBIA), Machine Learning classification (Random Forest), and Multicriteria Decision Analysis (MCDA) operationalized via the Analytic Hierarchy Process (AHP), we designed a robust spatial model within the Dinamica EGO platform. This methodological innovation enables the systematic integration of heterogeneous spatial layers (e.g. relief, land use and land cover, soil, wind speed, precipitation, temperature) and non-spatial data (e.g. vehicle tire type), producing a continuous Trafficability Index and its subsequent categorical mapping using Jenks natural breaks approach. The model's validity was demonstrated through comprehensive field surveys and accuracy assessment via statistical metrics, achieving an overall accuracy of 84.7% and a Kappa index of 0.81. This work advances cartographic modeling by formalizing a generalizable procedure that can support decision-making in diverse applications, such as transportation planning, environmental management, and military operations. It contributes as well to the methodological repertoire of GIScience for terrain analysis and cartographic modeling.
Uncertainty in bathymetric data poses a critical challenge for safe marine navigation, yet dedicated systems provide limited support for visualizing and incorporating uncertainty into automated safety checks. The route safety checker in Electronic Chart Display and Information Systems (ECDIS) searches for dangers along a planned track but relies on fixed buffers and charted depths alone, ignoring horizontal and vertical uncertainty. As a result, hazards may be missed or irrelevant features flagged, leading to insufficiently evaluated routes and increased cognitive workload. This study introduces two alternatives, Track and Hazard, to visualize and incorporate uncertainty into safety checks. These are compared to the ECDIS functionality in a controlled experiment with 96 professional mariners. Results show that Track and Hazard reduced error rates and response times, whereas ECDIS produces more errors and higher workload. Statistical analyses confirm significant differences, and post-experiment surveys reveal strong user preference for the proposed methods in both passage planning and monitoring. By embedding uncertainty directly into safety-check algorithms and visualizing hazards efficiently, the proposed approaches support faster and more reliable navigation decisions. Beyond navigation, they demonstrate how uncertainty visualization can enhance operational decision-support, offering a model for systems where safety depends on rapid and context-aware risk evaluation.
This study investigates the impact of two bivariate visualization methods - extrinsic (bar graphs) and intrinsic (Chernoff faces) - and the role of map literacy on user performance in value identification tasks within immersive virtual reality (IVR). A between-subject experiment (n = 126) assessed response correctness and time across two groups with differing presumed map literacy: geographers and laypeople. Supportive, though non-conclusive, eye tracking data (dwell times) were also collected to explore cognitive differences. Building on prior research with 2D stimuli, this study extends the investigation into a 3D IVR setting. Participants from diverse backgrounds (cartography, law, arts, etc.) were presented with 3D bivariate visualizations and tasked with identifying single/combined values of two variables. A PICO Neo3 Pro Eye VR headset with integrated eye tracking was used. Results indicate that Chernoff faces were less effective, yielding lower correctness, longer response times, and greater reliance on the legend, regardless of map literacy. Contrary to expectations, higher map literacy did not improve performance. Task complexity affected the methods differently: response time decreased for Chernoff faces but increased for bar graphs when going from one to two variables. Findings encourage further research with stronger map literacy assessment, varied visualizations, and refined eye tracking.
Mobile map-making on touchscreens remains constrained by interface occlusion and limited input accuracy. This study designs a Mid-Air Gesture interaction for mobile map-making by combining a viewport-first interface scheme (Viewport-Preserving Peripheral Layout, VPPL) with a constrained elicitation-based gesture-set design method (Restricted Gesture Elicitation Study, RGES). A within-subjects comparative evaluation (n = 15) contrasts Mid-Air Gesture interaction with Touchscreen interaction on two dimensions: usability and utility. Usability is measured with the System Usability Scale (SUS) after a map-making task. Utility is measured via point-placement precision (Euclidean spatial error) and map-view visibility (Map-View Visible Area Ratio and UI Occlusion Area). Results show lower point-placement error for Mid-Air Gesture interaction than Touchscreen interaction (9.33 px vs. 16.14 px) and a higher Map-View Visible Area Ratio (92.2% aggregated), whereas SUS scores are higher for Touchscreen interaction (85.33 vs. 70.83). These findings suggest that Mid-Air Gesture interaction is advantageous when precision and viewport visibility are critical, whereas perceived usability currently favors Touchscreen interaction. Future work will broaden scenarios, improve Mid-Air Gesture usability, and assess efficiency and energy use.
Underpinned by computer technology and intelligent devices, the display of geospatial elements has outstripped the constraints of traditional serial scale frameworks, progressing toward continuous multi-scale representation. Buildings are one of the most fundamental geospatial elements. Addressing the issues of feature point mismatching and unreasonable deformation of concave and convex parts of a building contour found in other continuous representation methods, as well as the inability to handle the continuous representation through scale of buildings in complex correspondences, i.e. where the correspondence is of the M:N (M > 1 or N > 1) or M:0 (M > 0) type, this paper proposes a multi-scale continuous transformation method for buildings based on graphical decomposition, and a linked matching method with edge-driven points. First, in the case of 1:1, the method of edge-driven points matching is designed based on distance and angle, for interpolating intermediate graphics. Second, for M:N complex elements, the graphics are decomposed into one-to-one correspondence, and then those are matched and transformed. Finally, this paper simplifies the 1:0 element according to its minimum bounding rectangle and reduces its complexity until it is removed from the dataset. The conducted experiments show that this method can achieve good multi-scale continuous expression of building map.
The preservation of hydrological ecosystems is crucial for sustainable forest management. Thanks to airborne lidar data acquisition programs, it is possible to collect high-resolution geospatial data and produce hydrographic networks from digital terrain models (DTM). However, underground structures such as culverts remain invisible in the DTM, causing deviations in the simulated flows along the paths. Drainage enforcement strategies are needed to solve this problem, but they modify the data. This project aims to integrate the culverts into the linear network, enabling an accurate mapping of the drainage network, including the subsurface component, without drainage enforcement that alters the DTM, or requiring manual intervention. The process relies on mapping the watercourses and drainage divides from the DTM and integrating culverts provided in a database. It involves careful modeling of the culverts, thoughtful design of the hydrographic network data structure, and the redefinition of the algorithm to calculate the drainage network. In addition to addressing underground and surface watercourses as two types of watercourse in the drainage network, it also provides a more accurate delimitation of drainage basins.
Multi-source toponymic matching is the process of consistently discriminating between geometric and attribute information of toponyms, so as to obtain more accurate, complete, and high-quality data. Attribute information of toponyms is expressed by natural language symbols, while geometric information of toponyms is expressed by coordinate values. This difference in expression system and expression mode leads to challenges in interpreting the semantic and spatial features of toponyms. At the present stage, geographic entity matching mainly adopts similarity index calculation method to compare toponymic features. However, these indices rely on subjective judgment thresholds, making it difficult to fully capture and analyze the semantic and spatial features of toponyms. Therefore, we propose a method for deep feature extraction of toponyms, embedding textual and spatial features into a unified high-dimensional vector representation space and using an attention mechanism to capture the intersection features between attributes to obtain the vector representation. Evaluated on GeoNames and OSM, our method achieves 98.20% accuracy and 0.9823 F1-score, significantly outperforming models based on character distances similarity and machine learning classifiers. Results demonstrate effective that this method achieves an effective integration of semantic and spatial features, enabling robust toponymic matching.
The COVID-19 pandemic has precipitated profound socioeconomic and public health ramifications globally. Wastewater-based epidemiology has been increasingly utilized for the surveillance and mitigation of COVID-19 outbreaks and transmission. The implementation of wastewater surveillance methodologies at a small scale has demonstrated considerable cost-effectiveness as an alternative to individual clinical testing, particularly in high-density environments such as academic institutions. Wastewater surveillance necessitates the acquisition and analysis of complex spatiotemporal data, requiring sophisticated interpretation and integration with complementary epidemiological parameters to effectively inform intervention strategies. The systematic management and analysis of these multivariate datasets present formidable logistical and computational challenges for timely decision making. This investigation advances geospatial science through a novel web-based spatial decision support system (SDSS) framework to resolve these challenges. This study encompasses the main campus of the University of North Carolina at Charlotte, where we implemented a spatiotemporal data model that revolutionizes management of multidimensional space-time data of wastewater surveillance. We employ advanced spatiotemporal analysis incorporating innovative cluster detection algorithms and uncertainty quantification to elucidate latent spatio-temporal patterns of SARS-CoV-2 virus abundance, which conventional approaches consistently fail to detect. This investigation conclusively demonstrates the superiority of integrated space-time cluster pattern analysis from both wastewater surveillance and clinical test results, quantifying their differential robustness when subjected to spatiotemporal uncertainties. The SDSS framework developed here represents a breakthrough in automated management, analytics, and dissemination of spatiotemporal wastewater epidemiological data. Our framework delivers transformative capabilities for informing evidence-based prevention strategies and establishing targeted intervention protocols for COVID-19 outbreaks within complex institutional environments.
This study proposes Geospatial Simultaneous Localization and Mapping for Outdoor Mobile Augmented Reality (GSOMAR), a framework that integrates geospatial data with perceptual computing to enable real-time modeling and interaction in complex urban environments. Unlike traditional mobile augmented reality (MAR) systems based on isolated overlays, GSOMAR ensures spatial continuity and semantic integration of heterogeneous elements - such as buildings, roads, points of interest (POIs), and pedestrians - within a coherent and responsive MAR environment. The framework fuses real-time kinematic global navigation satellite system (RTK-GNSS) measurements with an inertial measurement unit (IMU) to achieve accurate pose estimation, calibrates global scale and orientation, and supports real-time semantic modeling for interactive MAR scenes. Experimental results show that GSOMAR consistently outperforms existing RTK-GNSS + visual - inertial odometry (VIO) and ORB-SLAM-based methods across diverse outdoor scenarios. It achieves high-precision initialization over varied terrain, reducing average positional and gravity-vector errors by up to threefold compared with baseline methods. Extended tracking tests over 1400 m confirm sub-decimeter accuracy and improved stability with reduced drift. virtual-real alignment experiments further demonstrate robust spatial registration under occlusion and extended-range conditions. Overall, GSOMAR enables accurate, interactive, and semantically aware outdoor MAR, supporting applications such as urban modeling, navigation, and infrastructure inspection.
Point of Interest (POI) data is essential for accurate spatial analysis of pressing healthcare issues such as Drug and Substance Abuse (DSA). However, retrieving accurate healthcare POI information remains complicated. POI conflation integrates data from multiple sources to enhance spatial attribute quality and coverage. This study proposes a multi-step framework for healthcare POI conflation, including POI data collection from Location-Based Services (LBS), geographic and spatial attributes collection, calculating similarity across datasets, POI matching, spatial attributes enrichment, manual labeling, quality control, and deployment. We tested this framework on a DSA use case in California, USA. Our automated approach was able to detect 11,936 unique POIs related to this healthcare tag. Of the locations used for use case validation, 33% were common with the commercial benchmark POI dataset used by Safegraph. Out of the 11,936 total POIs, we were only able to find 4535 (38%) that were not in the commercial benchmark dataset and relevant to the use case. Moreover, our results indicate that spatial attributes fill rates are 32% at the geometry building level and 98% at the Census block group (CBG) level. We conclude that using LBS can provide POIs with relevance and spatial attributes similar to commercial datasets.
