GISTAM 2025 Abstracts

Area 1 - Data Acquisition and Processing

Full Papers

Paper Nr:	62
Title:	Person Detection from UAV Based on a Dual Transformer Approach
Authors:	Andrei-Stelian Stan, Dan Popescu and Loretta Ichim
Abstract:	The study introduces a novel object detection system that combines the strengths of two advanced deep learning models, the Detection Transformer (DETR) and the Vision Transformer (ViT), to enhance detection accuracy and robustness in unmanned aerial vehicle (UAV) applications. Both models were independently fine-tuned on the VisDrone dataset and then deployed in parallel, each processing the same input to leverage their advantages. DETR provides precise localization capabilities, particularly effective in crowded urban settings. At the same time, ViT excels at identifying objects at various scales and under partial occlusions, which is crucial for distant object detection. The fusion of their outputs is managed through a dynamic fusion algorithm, which adjusts the confidence scores based on contextual analysis and the characteristics of detected objects, resulting in a combined detection system that outperforms the individual models. The fused model significantly improved overall accuracy, achieving up to 90%, with a mean Average Precision (mAP50) of 85%, and a recall of 80%. These results underline the potential of integrating multiple transformer-based models to handle the complexities of UAV-based detection tasks, offering a robust solution that adapts to diverse operational scenarios and environmental conditions.
Download

Paper Nr:	68
Title:	Analyzing Multitemporal Datasets to Monitor Topographic Changes in Rio Cucco Italy
Authors:	Jad Ghantous, Vincenzo Di Pietra, Elena Belcore and Nives Grasso
Abstract:	Rio Cucco is an Italian catchment located in Malborghetto of Friuli Venezia Giulia. It is considered an area of interest regarding its hydrological and morphological properties. The area has historically been affected by natural hazards such as rockfall and landslides, mainly related to extreme rainfall events like the 2003 storm that affected the Fella river or the Vaia storm of October 2018. These events highlight the importance of understanding the morphological and topographic modification of the area also in relation to the realization of protection and hydraulic works. The changes in Rio Cucco were documented by comparing open-source historical data with ad-hoc UAV surveys focusing the analysis on 3D products like point clouds and the digital terrain models. The source of the recent data was an Aerial LiDAR- based survey conducted by our team in June 2024 while the historical data was taken from the FVG region’s geoportal and referred to 2017. After comparing the different datasets with traditional techniques like nearest neighbour Euclidean distance or DEM of Difference, changes were evident pointing to potential rockfalls between the year 2024 and 2017. A deep learning model was explored and in development for the semantic segmentation of the area.
Download

Short Papers

Paper Nr:	22
Title:	ADAclassifier: Trying to Ascertain Why the Ground Is Moving
Authors:	José A. Navarro, Anna Barra, María Cuevas-González, Pablo Ezquerro, Silvia Bianchini, Marta Béjar-Pizarro, Riccardo Palamà and Oriol Monserrat
Abstract:	The large availability of ground deformation measurements generated using Multi-Temporal Synthetic Aperture Radar (MT-InSAR), further increased by the contribution made by the European Ground Motion Service (EGMS), made displacement maps an increasingly common tool. However, their analysis is a complex task due to the large volume of Measurement Points (MPs) provided. ADAfinder, a tool within the ADAtools suite, allows for a reduction in the volume of information to be analyzed by identifying Active Deformation Areas (ADAs), i.e., areas of the terrain that actually move in a coherent and perceptible manner, including an estimate of the reliability of said identification. From here, the natural next step is identifying the reason why these areas are moving. This work presents ADAclassifier, another tool included in the ADAtools suite, still under development, aimed at evaluating up to five possible causes of ground movement, such as subsidence, landslide, uplift, sinkhole, and constructive settlement.
Download

Area 2 - GIS and Climate Change

Short Papers

Paper Nr:	25
Title:	Unsupervised Image Classification Algorithms Applied to Fire-Prone Area Detection
Authors:	I. Rahimi, L. Duarte and A. C. Teodoro
Abstract:	Remote sensing data has become critical in identifying fire-prone areas, providing essential insights through satellite imagery and various geospatial inputs. These data sources allow for real-time monitoring, mapping fire susceptibility, and assessing factors such as vegetation, fuel moisture, land use, and environmental conditions. Numerous supervised and unsupervised models combined with remote sensing data have shown great potential in predicting fire-prone regions, offering accurate and timely information for early warning systems and resource allocation. This study focuses on applying two unsupervised methods—PCA, and K-means—using inputs like Sentinel-2 imagery, elevation, and the Zagros Grass Index (ZGI) to identify fire-prone areas in the Kurdo-Zagrosian forests, an area increasingly vulnerable to wildfires. Among the two methods evaluated, PCA demonstrated superior performance in predicting fire-susceptible areas, accurately classifying 80% of the burned regions from 2021 to 2023 as moderate to high-risk zones.
Download

Paper Nr:	56
Title:	Advancing Real-Time Land Cover Classification for Biomass Density and Carbon Stocks Estimation in Google Earth Engine
Authors:	Dan Abudu, Lucy Bastin, Katie Chong and Mirjam Röder
Abstract:	Addressing climate change requires timely and accurate biomass and carbon stocks information. Traditional biomass estimation techniques rely on infrequent ground surveys and manual processing, limiting their scalability. This study proposes a novel framework that advances land cover classification to estimate biomass and carbon stocks using machine learning algorithms in Google Earth Engine. By integrating remote sensing data, machine learning algorithms, and allometric models, the framework automates above-ground biomass (ABG) and below-ground biomass (BGB) calculations, facilitating large-scale carbon stock assessments. The methodology leverages Landsat imagery, alongside derived Normalized Difference Vegetation Indices, to classify seven land cover types and estimate biomass. Equations are applied to derive AGB, with BGB calculated as a fraction of AGB. Carbon stock is estimated using a standard conversion factor of 0.47. Real-time processing capabilities of GEE ensure continuous monitoring and updates, enhancing accuracy and scalability. Findings demonstrate the potential for real-time biomass mapping and the identification of carbon-dense regions. The proposed approach is vital for sustainable land practices, carbon accounting, and forest conservation initiatives, to provide policymakers with accurate, real-time data, that supports climate mitigation efforts and contribute to realizing the Sustainable Development Goals 13 and 15.
Download

Paper Nr:	17
Title:	Preservation and Protection of Cultural Heritage in High Tourism Areas Using GIS Technology: A Case Study of the Medieval City of Rhodes
Authors:	Foteini-Pelagia Leventi, Lemonia Ragia and Dorina Moullou
Abstract:	The Medieval City of Rhodes is one of the most famous destinations in Greece, as it attracts more and more tourists, especially during the summer season. Given that this particular site is included in the UNESCO list, its protection and preservation are of utmost importance. Through a literature review, an effort is made to analyse the phenomenon of overtourism in the area. Both quantitative and qualitative data available on the topic were used from various sources in order to accurately frame the issues and implications of the phenomenon. Finally, several proposals are provided using a mix of strategies and mitigation measures relevant to the issue, emphasizing the importance of using QGIS.
Download

Paper Nr:	45
Title:	Hazard Modeling Using Sentinel Data for Risk Assessment and Management of Tsho Rolpa Glacier, Nepal
Authors:	Prasil Poudel, Nikesh Budha Chettri, Nabin Sah and Drabindra Pandit
Abstract:	The accelerated melting of the Himalayan glaciers due to climate change has caused significant issues including Glacier Lake Outburst Floods (GLOFs), to the downstream communities and the infrastructure. This study focuses on the Tsho Rolpa Glacier in Nepal, utilizing Sentinel-1 Synthetic Aperture Radar (SAR) data to measure ice velocity and analyze the impacts of glacier melting on lake dynamics. A GLOF occurs when the volume and surface area of glacier lake water exceeds the capacity of the moraine dam. For hazard modeling, SAGA GIS 8.3.0 was used for the terrain modeling and hydrological analysis, which highlighted the flood-prone areas. ArcGIS 10.5 facilitated the integration of Sentinel-2 imagery with local topographic data to predict the flood scenarios. The integration of these tools enhanced the accuracy of flood paths predictions and provided much needed valuable insights into the impacts on the local infrastructure. The results highlights the growing risks associated with the climate-induced GLOFs, demonstrating the importance on real-time glacier monitoring and predictive hazard modeling. This study tries to be a helpful tool for decision-support framework for mitigating the socio-economic impacts of GLOFs in vulnerable regions such as the Himalayas.
Download

Area 3 - Interaction with Spatial-Temporal Information

Full Papers

Paper Nr:	26
Title:	Are Londoners Getting Healthier?
Authors:	Yijing Li, Sijie Tan, Xiangbo Chang and Xiaohui Chen
Abstract:	It utilised urban data from multiple sources, to map out the physical health and mental health patterns in London over space and time. On basis of recognising the spatial patterns trajectory changes, obesity among adults and children in London areas had been investigated with selected demographic, socio-economic and environmental factors, to identify the most influential factors in all, and for local community; similarly, workflow had been again designed to investigate the influential factors for mental health prevalence as well. Upon comparing the selected models, models considering neighbourhood spillover effect has been found to be the optimal, to identify significantly influential factors on urban health, such as age group, green space access, household deprivation, income deprivation and air quality. The findings underscore the necessity for targeted, location-specific public health interventions to effectively combat obesity; highlight the importance of spatial heterogeneity, offering detailed insights into regional variations; and suggest tailored strategies for public health policies. This work fills a critical gap and demonstrates the need for geographically informed public health strategies.
Download

Paper Nr:	38
Title:	Navigating Points of Interest: The Dog-Walker Pathfinding Algorithm
Authors:	Natsuki Tsutsui and Shohei Yokoyama
Abstract:	Navigating complex environments that encompass both road networks and points of interest (POIs) demands innovative pathfinding solutions. Traditional algorithms, such as Dijkstra’s, primarily focus on finding the shortest path between nodes on a graph and are not designed to handle the additional complexity of scattered POIs. This paper introduces the dog-walker pathfinding algorithm, a novel approach that integrates the discovery of waypoints and greedy routing into a single process. By simulating the dynamics between two agents—a dog motivated by POIs and a walker navigating a graph—the algorithm dynamically identifies routes that connect key locations and pass through areas rich in POIs. This method leverages road networks and POIs to provide routes from the start point to the end point tailored to user preferences. Owing to the minimal data requirements for POIs, this method can be easily integrated with various data sources, including X( formerly Twitter), Google Maps, and Instagram. In this study, we develop an agent-based online algorithm inspired by the dogwalker. We verified the algorithm using POI data obtained from Flickr and Google Maps, demonstrating its application in real-world scenarios such as recommending tourist routes. Demonstrations show that our algorithm effectively generates routes that align with user preferences, as evidenced by qualitative and quantitative assessments. Additionally, we demonstrated that our method operates more rapidly than methods utilizing Dijkstra’s algorithm.
Download

Short Papers

Paper Nr:	49
Title:	An Innovative Model Based on Carvalho Rodrigues's Entropy to Assess Governance in Africa: A Guinea-Bissau Case Study
Authors:	João Serras, Paulo Morgado and Jorge Malheiros
Abstract:	Governance is an abstract, intersubjective, fluid concept, meaning that it is built upon shared beliefs, norms, and practices that are collectively agreed upon by a society, community, or group. It means different things to different cultural environments and scopes. Our current research addresses the challenge of obtaining insight into Governance's underlying mechanisms by using the concept of Cohesion derived by Carvalho Rodrigues from applying Shannon's Entropy. Our study is based on empirical data obtained from field observations across Guinea-Bissau. This paper presents our cohesion model and a first outlook for using it in the available data sets. It shows the impact of several potential Governance Determinants over a set of specific Governance Dimensions, demonstrating that Ethnicity Variation, local Community Morphology and the distance of Central Government facilities are the most impacting determinants for better cohesions.
Download

Paper Nr:	57
Title:	Low-Cost GNSS Receivers Reliability Using Centipede RTK Network for Land Surveying
Authors:	Muhammad Ali Sammuneh, Mojtaba Eslahi, Rani El Meouche and Elham Farazdaghi
Abstract:	The question of using low-cost Multiband Global Navigation Satellite System GNSS receivers and antennas in land surveying is real and important. In France, mainly a collaborative Real Time Kinematic (RTK) network called Centipede covers the country providing the corrections open access in real time to the users. Furthermore, the low-cost interface application and software called SW Maps connect the Low-cost GNSS receiver to the Centipede-RTK network using a smartphone. The cost of surveying projects using all these elements is certainly economical. The main question here is the reliability of this package to perform continuous, stable, and reliable RTK land surveying. We test this capability by examining the differences in the RTK position of known control points with a series of measurements over different mount points. The results show that we can use this package for land surveying only with necessary validation and control by experimental users, as the indicators of Centipede RTK accuracy via the SW Maps interface are not representative.
Download

Paper Nr:	70
Title:	Tracking the Progression of Burned Areas in Tropical Peat Swamp Forests by Integrating Sentinel Optical and SAR Imagery: A Case Study of Binsuluk Forest Reserve in Sabah, Malaysia
Authors:	Nurul Aina Abdul Aziz, McKreddy Yaban, Muhamad Zulfazli Zakaria and Siti Atikah Mohamed Hashim
Abstract:	Climate change and rising global temperatures are driving forest fires to become more intense and frequent worldwide, particularly in peat swamp forest. Since the predominant burning mechanism in peatland forest is smouldering combustion, it causes widespread air pollution and emits massive amounts of carbon due to prolonged episodes of fire events. Therefore, the development of a unique approach to monitor forest fire progressions through burned area mapping mainly in persistent cloud cover is vital for the estimation of fire extent, location, and land cover affected. Thus, this research aims to evaluate the capabilities of Sentinel-1 SAR and Sentinel-2 optical time series in boosting the frequency and accuracy of burn area progression mapping in peatland areas. Results from the forest fire series in Binsuluk Forest Reserve, which occurred from February to April 2024, indicated a reduction in the backscatter value of the cross-polarized (VH) signal in the burned area for Sentinel-1 SAR C band. Despite the cloud cover challenge, Sentinel-2 continues to deliver essential data on the positioning of active fires and smoke plumes, with burn area detection being more precise when utilizing the Normalized Difference Moisture Index (NDMI) compared to the Normalized Burn Ratio (NBR). The integration of Sentinel optical and SAR imagery has effectively facilitated an increased tracking frequency and precision for the evolution of burned areas.
Download

Area 4 - Managing Spatial Data

Short Papers

Paper Nr:	30
Title:	Navigating Boundary Discrepancies in SAD69-Based Delimitation: A Case Study and Practical Guidelines
Authors:	L. R. Costa, H. S. Delabary and R. Z. Araujo
Abstract:	The delimitation of the 84,130-hectare Serra do Tabuleiro State Park, a fully legally protected Conservation Unit in Santa Catarina, Brazil, relies on shapefiles provided by the managing institution and based on the SAD69 Coordinate Reference System. However, user-defined parameters when handling these shapefiles may result in up to three slightly different polygon representations, each affecting the perception of boundaries shared with adjacent territories. This study investigates these polygon discrepancies and assesses which representation most accurately reflects the intended delimitation. Although a definitive solution is not reached, the authors provide valuable recommendations for public authorities and GIS users to standardize interpretations and improve boundary accuracy.
Download

Area 5 - Spatial Data Mining

Short Papers

Paper Nr:	47
Title:	Planning Delivery Services: Depot Clustering Based on Socio-Economic Indicators and Geospatial Metrics
Authors:	Iñaki Cejudo, Laura Rabadán, Eider Irigoyen and Harbil Arregui
Abstract:	People’s lifestyles have evolved in recent years, making home deliveries a necessity for various types of services. Moreover, with the growth of big data and Artificial Intelligence, predicting the performance and customer demand of new businesses is a key aspect of logistics and last-mile delivery planning. By using examples and predictions as a foundation for goods delivery services, initial over-sizing costs can be signifi-cantly reduced. In this paper, we analyze and compare operational zone similarities for food and parcel delivery services in Spain, considering socio-economic indicators and urban network features. The study leverages motorbike delivery metrics to complement the analysis. The results demonstrate how similar depots can be clustered, providing a foundational performance scenario for decision-making when planning the launch of a new service.
Download

Paper Nr:	60
Title:	The Impact of Data Science on Geography: A Review with Optimization Algorithms
Authors:	Roberto de Oliveira Machado
Abstract:	We conducted a systematic review using the PRISMA methodology, analyzing 2,996 studies and synthesizing 41 to explore the evolution of data science and its integration into geography. Optimization algorithms were employed to enhance the efficiency and precision of literature selection. Our findings reveal that data science has evolved over five decades, facing challenges such as the integration of diverse spatial data and the increasing demand for advanced computational skills. In the field of geography, data science emphasizes interdisciplinary collaboration and methodological innovation. Techniques like large-scale spatial data analysis and predictive algorithms hold promise for applications in natural disaster management and transportation optimization, enabling faster and more effective responses. These advancements highlight data science’s pivotal role in solving complex spatial problems. This study contributes to the application of optimization algorithms in systematic reviews and underscores the necessity for deeper integration of data science into geography. Key contributions include identifying challenges in managing heterogeneous spatial data and promoting advanced analytical capabilities. The intersection of data science and geography leads to significant improvements in disaster management and transportation efficiency, fostering more sustainable and impactful environmental solutions.
Download

Paper Nr:	66
Title:	Machine Learning for Identifying Potential Photovoltaic Installations on Parking Areas
Authors:	Frederick Kistner and Sina Keller
Abstract:	Integrating renewable energy systems into urban areas is crucial for sustainable development. This study assesses the potential for installing photovoltaic (PV) systems in parking areas, focusing on a case study region in Hesse, Germany. A machine learning approach is developed to classify parking lots larger than 900 m2 into suitable and unsuitable categories. The input data includes OpenStreetMap (OSM), the Authoritative Topographic-Cartographic Information System (ATKIS), and high-resolution geospatial datasets. A reference dataset for the two classification categories is created. Multiple input features are generated, and their significance for the classification task is evaluated. Additionally, several shallow machine learning models are implemented and assessed. The XGBoost model demonstrates the highest accuracy at 99 % and is used to classify 10,894 parking areas throughout Hesse. Key suitability features include the Normalized Difference Vegetation Index (NDVI), surface sealing ratios, and vegetation height. The results indicate that approximately 21.8 km2 of the parking area is suitable for PV installations, requiring minimal ecological intervention. The methodological approach is scalable for application in other regions, and validation in Frankfurt am Main confirms a strong correlation with solar radiation levels. This study provides a data-driven framework for optimizing urban energy systems and supporting sustainability initiatives.
Download

Area 6 - Remote Sensing

Full Papers

Paper Nr:	40
Title:	Satellite Images and Spectral Vegetation Indices as Auxiliary Tools to Monitoring Fuel Availability in Areas Prone to Wildfire: Study Case in the Northern Region of Portugal
Authors:	Bárbara Pavani-Biju, José G. Borges, Susete Marques and Ana C. Teodoro
Abstract:	Remote sensing data has led to the development of spectral indices for monitoring ecosystems, land surface changes, and water quality. These indices are used in various applications, including agricultural and wildfire monitoring, to understand vegetation cycles and disturbances. Wildfire research focuses on the effects of extreme occurrences, and understanding forest ecology after severe events is crucial for evaluating forest health. Vegetation Indices (VIs) are frequently used in forest and wildfire monitoring studies to account for plant biophysical, biochemical, and physiological characteristics. Normalized Difference Vegetation Index (NDVI), Normalized Burn Ratio (NBR), Normalized Difference Infrared Index (NDII), and Plant Senescence Reflectance Index (PSRI) are indices used to assess vegetation conditions. VIs are valuable resources for monitoring post-wildfire occurrences, as they measure biophysical changes and provide comprehensive monitoring of the affected area, playing a crucial role in assessing the health of forests. Pre-wildfire vegetation conditions monitoring is also important for implementing preventative measures in critical regions to increase wildfire defense and identifying wildland fuels is crucial for improving fuel management actions. This research aims to demonstrate the effectiveness of chosen VIs and fuel models as tools to assess pre-fire conditions, enabling decision-makers to increase wildfire surveillance and landscape resilience in Vale do Sousa, Portugal's northern area. Despite limitations, this approach is valuable, especially in terms of financial or logistical constraints. Moreover, combining VIs with fuel hazard models can improve fuel reduction efforts.
Download

Short Papers

Paper Nr:	16
Title:	Differential Interferometric Synthetic Aperture Radar-Based Landslide Monitoring: A Case Study of Wayanad, India
Authors:	Manvi Kanwar and Samsung Lim
Abstract:	On July 30th, 2024, the Wayanad region of Kerala, India, experienced a devastating landslide triggered by intense monsoon rainfall. This event highlighted the region's need for effective landslide monitoring and early warning systems. Differential Interferometric Synthetic Aperture Radar (DInSAR) is a powerful and efficient remote sensing technique for detecting ground deformation and monitoring landslides. This study utilizes Sentinel-1 SAR images to monitor the Wayanad landslide by processing and analyzing SAR data using the DInSAR technique to identify ground displacement patterns. The results demonstrate the effectiveness of DInSAR in capturing pre- and post-event deformations, offering valuable insights into the landslide dynamics. This study underscores the potential of SAR-based real-time landslide monitoring and risk mitigation in landslide-prone regions.
Download

Paper Nr:	44
Title:	Analysis of Surface Thermal Behavior in Different Local Climate Zones (LCZ): A Case Study in Bragança (Portugal) (2013-2024)
Authors:	Cátia Rodrigues de Almeida, Artur Gonçalves and Ana Cláudia Teodoro
Abstract:	The Urban Heat Island (UHI) effect occurs when temperatures in urban areas are higher than surrounding vegetated areas, especially during the sunset and sunrise. UHI impacts include effects on public health and well-being, changes to the local microclimate, and influence on the local biome. This study evaluates the Land Surface Temperature (LST) and the corresponding Surface Urban Heat Island Intensity (SUHIint) across different Local Climate Zones (LCZs) in Bragança (Portugal) from 2013 to 2024, using images from Landsat 8 and 9 data collected with a portable thermal camera on different surfaces to assess thermal behavior across different scales. The results confirm the existence of the UHI effect in Bragança, where vegetated areas exhibit milder temperatures compared to built areas, especially in summer afternoons. Satellite-derived LST data indicate that the lowest temperature was recorded in an LCZ with vegetation, reaching (-7ºC), while the highest minimum temperature was observed in an LCZ with higher density of anthropogenic elements (-3ºC). Thermal camera measurements showed that surfaces such as asphalt and exposed soil reached 80 ºC in the morning and remained above 60ºC in the afternoon. These findings underscore the importance of considering mitigation measures, such as increasing vegetation in urbanized areas or replacing impervious surfaces.
Download

Paper Nr:	48
Title:	Advancing Airport Land Subsidence Monitoring Through Time-Series InSAR Technology
Authors:	Anuphao Aobpaet, Intira Thanomsin, Chanikan Yodya and Suchanpong Obnam
Abstract:	The airport is a pivotal infrastructure project, serving as a hub for parking, transporting, and maintaining aircraft carrying passengers, freight, and cargo. However, the substantial usage of the airport leads to the challenge of land subsidence, necessitating ongoing monitoring and assessment. This study focuses on monitoring land subsidence at Suvarnabhumi Airport, Thailand's premier international airport catering to passengers and aircraft. In a lowland area with soft soil, applying advanced technology becomes imperative for continuous monitoring and analysis of subsidence over time. Employing InSAR Time-Series technology, researchers processed data from Sentinel-1 satellites spanning October B.E. 2017 to December 2023 to analyse the evolving conditions at Suvarnabhumi Airport. Results reveal that the most significant subsidence occurs in the Runway and Taxiway areas, with values ranging between -9.1 and 5.1 millimeters. per year. This subsidence is likely attributed to the constant heavy air traffic on these surfaces. Continuous monitoring and evaluation are crucial to planning effective maintenance. InSAR technology is valuable for monitoring land subsidence or displacement, alleviating data constraints and streamlining operational processes.
Download

Paper Nr:	64
Title:	Assessment of Fine-Tuned Canopy Height Maps from Satellite Imagery: A Case Study in the Czech Republic
Authors:	Leonidas Alagialoglou, Ioannis Manakos, Olga Brovkina, Jan Novotný and Anastasios Delopoulos
Abstract:	This study evaluates the performance of a lightweight convolutional Long Short-Term Memory (ConvLSTM)based deep learning model for estimating canopy height across three test areas in the Czech Republic using Sentinel-2 time series data. The model, initially trained on forest data from Germany and Switzerland, incorporate uncertainty quantification techniques and was fine-tuned and evaluated using dense airborne laser scanning (ALS) data collected between 2022 and 2024. Results show that fine-tuning reduced mean absolute error (MAE) from 4.26 m to 2.74 m in the primary test area, with similar improvements across other regions. Species-specific uncertainties were also analyzed, highlighting performance variations between deciduous and coniferous forests.
Download

Paper Nr:	69
Title:	Remote Sensing-Based Temporal Analysis of Aletsch Glacier Retreat (1990–2020)
Authors:	Andrija Krtalić, Ana Kuveždić Divjak and Kristina Zeman Šteković
Abstract:	Glaciers represent an important component of the cryosphere and are among the most sensitive indicators of climate change and global warming. Over recent decades, climate change has significantly accelerated glacier retreat, prompting the development of various monitoring methods, such as the Glacier Monitoring in Switzerland (GLAMOS) program. Key parameters for understanding glacier dynamics include changes in mass balance, length, surface area, and snow accumulation, all of which are closely tied to climatic variations. These changes manifest as alterations in glacier morphology and mass, resulting in notable retreat compared to previous decades when such trends were less pronounced. This study focuses on analysing changes in the Great Aletsch Glacier, the largest glacier in the Alps, over the period 1990–2020 using remote sensing data/techniques. Automated glacier detection and mapping methods were applied, utilizing optical satellite data from Sentinel-2 and Landsat-5 missions. Glacier extents were delineated and analysed over the 30-year period, integrating satellite-derived estimates with official GLAMOS data and climatological records from the MeteoSwiss agency. The results reveal a reduction of approximately 5.2% in the surface area of the Great Aletsch Glacier, providing valuable insights into the glacier’s response to ongoing climate change.
Download

Area 7 - Modeling, Representation and Visualization

Full Papers

Paper Nr:	39
Title:	Characterizing Machine Guidance in Geospatial Analysis
Authors:	Yue Hao and Guoray Cai
Abstract:	Geospatial analysis poses challenges for individuals with limited expertise in Geographic Information Science (GIScience) methods and tools, requiring complex decision-making and spatial reasoning, often leading to difficulties or even failures. This research explores machine guidance as a novel approach to support domain analysts throughout the geo-analytical process. This approach can be implemented as an intelligent interface agent that is capable of recognizing user’s analytical difficulties and providing proactive assistance. We propose a conceptual framework that characterizes the cooperative role of machine guidance in geospatial analysis. We focus on answering two key research questions: (1) when to guide? and (2) how to guide?. The framework provides a foundation for future research on machine-guided geospatial analysis, informing the development of other computer-aided systems that enhance usability and analytical effectiveness in GIScience.
Download

Paper Nr:	54
Title:	A Luminance-Based Lane Marking Candidate Quality Assessment for Autonomous Driving in GIS Contexts
Authors:	Pathum Rathnayaka, Young Hun Kim, In Gu Choi, Gi Chang Kim and Duk Jung Kim
Abstract:	The detection of lane marking candidates is crucial for autonomous vehicles and advanced driver-assistance systems ADAS as they deliver essential information for accurate lane following, localization, and route planning. Detecting these candidates becomes difficult when road conditions deteriorate or the marking paint is faded, sometimes making identification nearly impossible. While deep/machine learning (DL/ML) methods perform well in reliable detection, they often come with the need for extensive labeled datasets, substantial computational power, and intricate parameter adjustments. In this research, we present a method purely based on digital image processing to identify and evaluate lane marking candidates, thus avoiding the use of specialized reflectivity equipment (such as luminance meters) and bypassing complex DL/ML methodologies. Our pipeline initially identifies lane marking regions using a collection of image-processing techniques. It then subsequently evaluates their quality using two conditional metrics: a luminance-based contrast ratio and a white-to-black pixel ratio. Each candidate is categorized as good, bad, or ambiguous by comparing these metrics to empirically determined thresholds. Evaluations conducted on large sets of road images from conventional and urban highways in South Korea confirm the effectiveness of our proposed method. The system significantly reduces the dependence on time-consuming labor-intensive annotation, high-end hardware, and DL/ML expertise. We thus claim that our lightweight, deployable method effectively addresses a significant gap in luminance-centric lane candidate quality evaluation and can serve either as an independent solution or as a supplementary option to more sophisticated DL/ML systems in GIS and ADAS applications.
Download

Short Papers

Paper Nr:	27
Title:	Urban Growth in Metropolitan Regions Using Dynamic Modeling by Cellular Automata: A Comparative Analysis Between Brazil and Portugal
Authors:	Elizabeth Maria Feitosa da Rocha de Souza, Antonio Alberto Teixeira Gomes and Vandre Soares Viegas
Abstract:	This study synthesizes the outcomes of land use changes obtained through the implementation of dynamic modeling by cellular automata across two metropolitan regions in Portugal and Brazil. The purpose is to analyze the primary findings acquired, considering the particularities of each nation and evaluate the potentialities of the used data. The study examined the metropolitan regions of MRRJ (Rio de Janeiro, Brazil) and AMP (Porto, Portugal). Modifications were implemented in the DinamicaEgo software to the fundamental data representing static and dynamic variables for each context. The findings revealed a substantial increase of urban areas in the MRRJ, and the modeling demonstrated its applicability across the two contexts, considering the requisite modifications for the data accessible in each country.
Download

Paper Nr:	61
Title:	Surface Current Visualization in Waterway Based on Mike 21 Model and S-100 Standards
Authors:	Zixuan Wang, Mingyang Pan, Shaoxi Li, Chao Li and Zongying Liu
Abstract:	The S-100 standard, proposed by the International Hydrographic Organization (IHO), aims to address the limitations of the S-57 standard in terms of data application and interoperability. This study focuses on the Fujiangsha water area, using the Mike 21 hydrodynamic model to simulate the flow dynamics in the region. The standardized processing of this data generates S-111 surface current data that complies with the S-100 specifications. The visualization of surface current data is realized using Cesium, which displays the flow field characteristics of the Fujiangsha water area. By integrating the standardization of S-111 data with visualization technology, this paper seeks to provide new technical support and application demonstrations for maritime management and navigation safety, while exploring the potential for further development of the S-100 standard in practical applications.
Download

Paper Nr:	52
Title:	LiDAR-Based 3D Reconstruction for Robotic Pipelines Inspection
Authors:	Monika Sara Kawka, Lazaros Grammatikopoulos, Ilias Kalisperakis and Christos Stentoumis
Abstract:	Robotic platforms have transformed pipe inspection from routine checks into an automatic data-driven process. Such robotic systems often integrate computer vision technology to collect and analyze inspection data in an automated and efficient way and offer additional capabilities such as 3D reconstruction of pipes and precise measurement of deformations (e.g., dents, buckling). This work presents an initial case study of a robotic inspection system equipped with LiDAR and camera sensors capable of performing automatic pipeline inspections. This proof-of-concept study is dedicated to the 3D reconstruction of the pipeline using LiDAR data collected during inspections. Reconstruction accuracy is evaluated by computing the RMSE for pipe surface reconstruction and the deviation from the reference diameter of a single pipe in a controlled laboratory setting. Reconstruction results reach an accuracy higher than 2cm based on computed RMSE and a precision higher than 0.5cm in pipe diameter estimation. The current implementation is limited to the inspection of matte and non-reflective pipes. Still, it offers a straightforward and scalable solution for various industrial sectors. Future work will incorporate camera data to integrate color mapping into the 3D reconstruction model and detect potential defects and deformations in a pipe.
Download

Area 8 - Domain Applications

Full Papers

Paper Nr:	41
Title:	Modelling the Road Network as an Expression of Historical Spatial System Change in the Klaipeda Region (Lithuania)
Authors:	Thomas Gloaguen, Kęstutis Zaleckis and Sébastien Gadal
Abstract:	Since the 20th century, the Klaipėda region (Lithuania) has undergone significant political, economic, social, and cultural changes related to the disappearance of Prussia (1923), the Soviet occupation (1944) and the restoration of independence (1991). This has led to radical transformations in the spatial system, of which the road network is a key component. Networks could be analysed using space syntax methods by assimilating them to a mathematical graph model, for which quantitative indexes based on topology can be applied to assess their spatial configurations. The ‘Generic City’ approach, by simulating foreground and background urban networks, can be modified innovatively on a territorial scale. Based on historical cartographies and open geographic databases, indexes characterising both networks are derived to describe the length, angularity or accessibility of the pre-Soviet, Soviet and post-Soviet networks. The analysis identified four main spatial structures associated with the network by combining the indexes in a K-means machine learning algorithm. They highlight the spatial impacts of collectivisation, industrialisation, and tertiarisation of the economy, post-World Wars and post-Cold War geopolitical events, and their consequences as drivers of the territorial organisation and dynamics such as the metropolisation or peri-urbanisation.
Download

Paper Nr:	53
Title:	Developing a Geospatial Framework for Calculating a 15-Minute City Index (FMCI): The Case of Quezon City
Authors:	Carlo Angelo R. Mañago, Marielle G. Nasalita, Cesar V. Saveron, Ynah Andrea D. Sunga and Alexis Richard C. Claridades
Abstract:	The 15-minute city concept is a measure of the quality of urban life based on proximity, sustainability, and sociality. This study proposes a geospatial framework for calculating the 15-minute city index (FMCI) aimed to measure the accessibility of its residents to six social functions, which include living, working, supplying, caring, learning, and enjoying. Quezon City, Philippines, was chosen for its urban characteristics that aligned with this vision and served as the study area. To account for pedestrian needs, age-based weights were assigned to the social functions, and service areas were mapped using a uniform walking speed. FMCI values were calculated based on weighted social functions and barangay population distribution by age group. Results revealed that 39% of Quezon City’s barangays achieved a perfect FMCI score of one, indicating access to all six functions within a 15-minute walk. Positive spatial autocorrelation indicated the clustering of barangays with similar FMCI values, with hot spots in the southern and cold spots in the northern parts of the city. These findings offer insights for policymakers in improving urban life quality. The adaptable FMCI framework can be applied to other urban areas to assess service accessibility, considering residents' needs.
Download