ISPRS International Journal of Geo-Information

19 pages, 5277 KB

Open AccessArticle

A Machine Learning Approach Using Spatially Explicit K-Nearest Neighbors for House Price Predictions

by Meifang Chen, Changho Lee and Yongwan Chun

ISPRS Int. J. Geo-Inf. 2026, 15(1), 46; https://doi.org/10.3390/ijgi15010046 - 21 Jan 2026

Spatial data has distinctive properties that differentiate it from non-spatial data. One prominent characteristic is spatial autocorrelation (SA). When machine learning techniques are applied for spatial data modeling, they require spatially explicit consideration. If these inherent spatial structures are ignored, models may produce [...] Read more.

Spatial data has distinctive properties that differentiate it from non-spatial data. One prominent characteristic is spatial autocorrelation (SA). When machine learning techniques are applied for spatial data modeling, they require spatially explicit consideration. If these inherent spatial structures are ignored, models may produce biased predictions. However, integrating this property into the model yields additional spatial insight, thereby enhancing learning and improving predictive accuracy. This study examines spatially explicit

K

-nearest neighbors (SE-KNN) by integrating SA as a spatially explicit property,

λ

, into the learning algorithm. The innovation of SE-KNN lies in its alignment with the principle of spatial autocorrelation, as KNN’s core learning assumption—that similar observations tend to have similar outcomes—naturally parallels spatial dependence. The proposed SE-KNN is applied to a house price prediction model using house sales data from Franklin County, Ohio to demonstrate a spatially dependent, data-rich, and real-world problem. The results show that SE-KNN achieved the best prediction accuracy compared to mean of absolute error (MAE) of three other machine learning approaches (i.e., standard KNN, linear regression, and artificial neural networks). The proposed method effectively captures the spatial structures in the housing market and leaves only a trace amount of SA in the residuals. Full article

(This article belongs to the Special Issue Spatial Data Science and Knowledge Discovery)

► Show Figures

Figure 1

29 pages, 15635 KB

Open AccessArticle

Flood Susceptibility and Risk Assessment in Myanmar Using Multi-Source Remote Sensing and Interpretable Ensemble Machine Learning Model

by Zhixiang Lu, Zongshun Tian, Hanwei Zhang, Yuefeng Lu and Xiuchun Chen

ISPRS Int. J. Geo-Inf. 2026, 15(1), 45; https://doi.org/10.3390/ijgi15010045 - 19 Jan 2026

Abstract

This observation-based and explainable approach demonstrates the applicability of multi-source remote sensing for flood assessment in data-scarce regions, offering a robust scientific basis for flood management and spatial planning in monsoon-affected areas. Floods are among the most frequent and devastating natural hazards, particularly [...] Read more.

This observation-based and explainable approach demonstrates the applicability of multi-source remote sensing for flood assessment in data-scarce regions, offering a robust scientific basis for flood management and spatial planning in monsoon-affected areas. Floods are among the most frequent and devastating natural hazards, particularly in developing countries such as Myanmar, where monsoon-driven rainfall and inadequate flood-control infrastructure exacerbate disaster impacts. This study presents a satellite-driven and interpretable framework for high-resolution flood susceptibility and risk assessment by integrating multi-source remote sensing and geospatial data with ensemble machine-learning models—Extreme Gradient Boosting (XGBoost) and Light Gradient Boosting Machine (LightGBM)—implemented on the Google Earth Engine (GEE) platform. Eleven satellite- and GIS-derived predictors were used, including the Digital Elevation Model (DEM), slope, curvature, precipitation frequency, the Normalized Difference Vegetation Index (NDVI), land-use type, and distance to rivers, to develop flood susceptibility models. The Jenks natural breaks method was applied to classify flood susceptibility into five categories across Myanmar. Both models achieved excellent predictive performance, with area under the receiver operating characteristic curve (AUC) values of 0.943 for XGBoost and 0.936 for LightGBM, effectively distinguishing flood-prone from non-prone areas. XGBoost estimated that 26.1% of Myanmar’s territory falls within medium- to high-susceptibility zones, while LightGBM yielded a similar estimate of 25.3%. High-susceptibility regions were concentrated in the Ayeyarwady Delta, Rakhine coastal plains, and the Yangon region. SHapley Additive exPlanations (SHAP) analysis identified precipitation frequency, NDVI, and DEM as dominant factors, highlighting the ability of satellite-observed environmental indicators to capture flood-relevant surface processes. To incorporate exposure, population density and nighttime-light intensity were integrated with the susceptibility results to construct a natural–social flood risk framework. This observation-based and explainable approach demonstrates the applicability of multi-source remote sensing for flood assessment in data-scarce regions, offering a robust scientific basis for flood management and spatial planning in monsoon-affected areas. Full article

(This article belongs to the Topic Machine Learning and Big Data Analytics for Natural Disaster Reduction and Resilience)

► Show Figures

Figure 1

13 pages, 3979 KB

Open AccessArticle

Decomposing Spatial Accessibility into Demand, Supply, and Traffic Speed: Averaging Chain Substitution Method

by Kyusik Kim and Kyusang Kwon

ISPRS Int. J. Geo-Inf. 2026, 15(1), 44; https://doi.org/10.3390/ijgi15010044 - 18 Jan 2026

Abstract

Spatial accessibility to healthcare services is commonly determined by three core components: demand, supply, and traffic speed. Although understanding which factors contribute to accessibility changes can help prioritize interventions to enhance accessibility in underserved areas, limited research has examined the extent of their [...] Read more.

Spatial accessibility to healthcare services is commonly determined by three core components: demand, supply, and traffic speed. Although understanding which factors contribute to accessibility changes can help prioritize interventions to enhance accessibility in underserved areas, limited research has examined the extent of their individual contributions. To better capture the local dynamics that shape healthcare accessibility, this study decomposes spatial accessibility to primary healthcare services using the chain substitution method (CSM), which quantifies the impact of each component by substituting them one by one. By examining how the order of factor substitution affects the relative impact of each factor on spatial accessibility, we analyzed the importance of substitution order in the CSM. This study found that the order of factor substitution plays a significant role in measuring the relative contribution of each factor. To mitigate the effects of substitution order, we proposed an averaging CSM that uses the average value across all possible substitution combinations. Based on the averaging CSM, our findings offer insight for healthcare policymakers and urban planners by clarifying how demand, supply, and traffic speed interact in determining accessibility, ultimately supporting targeted interventions in underserved areas. Full article

(This article belongs to the Special Issue HealthScape: Intersections of Health, Environment, and GIS&T (2nd Edition))

► Show Figures

Figure 1

26 pages, 14905 KB

Open AccessArticle

Data–Knowledge Collaborative Learning Framework for Cellular Traffic Forecasting via Enhanced Correlation Modeling

by Keyi An, Qiangjun Li, Kaiqi Chen, Min Deng, Yafei Liu, Senzhang Wang and Kaiyuan Lei

ISPRS Int. J. Geo-Inf. 2026, 15(1), 43; https://doi.org/10.3390/ijgi15010043 - 16 Jan 2026

Abstract

Forecasting the spatio-temporal evolutions of cellular traffic is crucial for urban management. However, achieving accurate forecasting is challenging due to “complex correlation modeling” and “model-blindness” issues. Specifically, cellular traffic is generated within complex urban systems characterized by an intricate structure and human mobility. [...] Read more.

Forecasting the spatio-temporal evolutions of cellular traffic is crucial for urban management. However, achieving accurate forecasting is challenging due to “complex correlation modeling” and “model-blindness” issues. Specifically, cellular traffic is generated within complex urban systems characterized by an intricate structure and human mobility. Existing approaches, often based on proximity or attributes, struggle to learn the latent correlation matrix governing traffic evolution, which limits forecasting accuracy. Furthermore, while substantial knowledge about urban systems can supplement the modeling of correlations, existing methods for integrating this knowledge—typically via loss functions or embeddings—overlook the synergistic collaboration between data and knowledge, resulting in weak model robustness. To address these challenges, we develop a data–knowledge collaborative learning framework termed the knowledge-empowered spatio-temporal neural network (KESTNN). This framework first extracts knowledge triplets representing urban structures to construct a knowledge graph. Representation learning is then conducted to learn the correlation matrix. Throughout this process, data and knowledge are integrated collaboratively via backpropagation, contrasting with the forward feature injection methods typical of existing approaches. This mechanism ensures that data and knowledge directly guide the dynamic updating of model parameters through backpropagation, rather than merely serving as a static feature prompt, thereby fundamentally alleviating the “model-blindness” issue. Finally, the optimized matrix is embedded into a forecasting module. Experiments on the Milan dataset demonstrate that the KESTNN exhibits excellent forecast performance, reducing RMSE by up to 23.91%, 16.73%, and 10.40% for 3-, 6-, and 9-step forecasts, respectively, compared to the best baseline. Full article

► Show Figures

Figure 1

28 pages, 3359 KB

Open AccessArticle

Pedestrian Trajectory Prediction Based on Delaunay Triangulation and Density-Adaptive Higher-Order Graph Convolutional Network

by Lei Chen, Jiajia Li, Jun Xiao and Rui Liu

ISPRS Int. J. Geo-Inf. 2026, 15(1), 42; https://doi.org/10.3390/ijgi15010042 - 15 Jan 2026

Abstract

Pedestrian trajectory prediction plays a vital role in autonomous driving and intelligent surveillance systems. Graph neural networks (GNNs) have shown remarkable effectiveness in this task by explicitly modeling social interactions among pedestrians. However, existing methods suffer from two key limitations. First, they face [...] Read more.

Pedestrian trajectory prediction plays a vital role in autonomous driving and intelligent surveillance systems. Graph neural networks (GNNs) have shown remarkable effectiveness in this task by explicitly modeling social interactions among pedestrians. However, existing methods suffer from two key limitations. First, they face difficulty in balancing the reduction in redundant connections with the preservation of critical interaction relationships in spatial graph construction. Second, higher-order graph convolution methods lack adaptability to varying crowd densities. To address these limitations, we propose a pedestrian trajectory prediction method based on Delaunay triangulation and density-adaptive higher-order graph convolution. First, we leverage Delaunay triangulation to construct a sparse, geometrically principled adjacency structure for spatial interaction graphs, which effectively eliminates redundant connections while preserving essential proximity relationships. Second, we design a density-adaptive order selection mechanism that dynamically adjusts the graph convolution order according to pedestrian density. Experiments on the ETH/UCY datasets show that our method achieves 5.6% and 9.4% reductions in average displacement error (ADE) and final displacement error (FDE), respectively, compared with the recent graph convolution-based method DSTIGCN, demonstrating the effectiveness of the proposed approach. Full article

► Show Figures

Figure 1

22 pages, 4621 KB

Open AccessArticle

Spatiotemporal Imbalances in Dockless Bike-Sharing Usage: Evidence from Shanghai

by Ke Song, Keyu Lin and Mi Diao

ISPRS Int. J. Geo-Inf. 2026, 15(1), 41; https://doi.org/10.3390/ijgi15010041 - 14 Jan 2026

Abstract

Rebalancing shared bikes poses a significant challenge for dockless bike-sharing (DLBS) operators, as inevitable spatiotemporal mismatches between demand and supply lead to high redistribution costs. Despite its operational significance, empirical research on the spatiotemporal imbalance of DLBS usage and its underlying drivers remain [...] Read more.

Rebalancing shared bikes poses a significant challenge for dockless bike-sharing (DLBS) operators, as inevitable spatiotemporal mismatches between demand and supply lead to high redistribution costs. Despite its operational significance, empirical research on the spatiotemporal imbalance of DLBS usage and its underlying drivers remain limited. Utilizing one month’s extensive trajectories of shared bikes in Shanghai, China, this study quantifies DLBS net flows at fine-grained grid level by hour to capture demand–supply imbalances across both spatial and temporal dimensions. To uncover dominant patterns in DLBS imbalance, we employ non-negative matrix factorization, a matrix decomposition technique, to extract latent structure of DLBS net flows. Four distinct patterns are identified: self-sustained balance, morning peak outflow, morning peak inflow, and metro-driven imbalance. We further apply multinomial logit models (MNL) to examine how these patterns are associated with different built environment characteristics. The results show that higher population density, greater diversity of points of interest, and proximity to city centers promote more balanced DLBS flows, whereas high road network density and concentrations of subway stations, residential communities, and firms intensify imbalances. These findings provide valuable insights for enhancing the operational efficiency of DLBS systems and supporting informed transportation management and urban planning practices. Full article

► Show Figures

Figure 1

24 pages, 12254 KB

Open AccessArticle

Dynamic Measurement and Equity Analysis of Walking Accessibility in Primary Healthcare Institutions Under Diverse Supply–Demand Scenarios: Evidence from Shenyang

by Yang Li, Enxu Wang, Shasha Li, Qiao Cui and Hao Xie

ISPRS Int. J. Geo-Inf. 2026, 15(1), 40; https://doi.org/10.3390/ijgi15010040 - 13 Jan 2026

Abstract

The walking accessibility of primary healthcare institutions (PHCIs) is a pivotal determinant of health equity. However, prior studies often lack a comprehensive assessment that integrates the spatiotemporal dynamics of both multi-faceted supply and multi-scenario demand. To bridge this gap, this study develops an [...] Read more.

The walking accessibility of primary healthcare institutions (PHCIs) is a pivotal determinant of health equity. However, prior studies often lack a comprehensive assessment that integrates the spatiotemporal dynamics of both multi-faceted supply and multi-scenario demand. To bridge this gap, this study develops an enhanced two-step floating catchment area method (2SFCA-MSD) that concurrently incorporates multiple types of service supply and multiple temporal demand scenarios to quantify PHCI walking accessibility, with equity evaluated using the Gini coefficient and Lorenz curve. The results indicate that: (1) Both supply and demand exhibit pronounced spatiotemporal inequalities. (2) Walking accessibility varies substantially across scenarios; Health services for vulnerable groups (Service B) exhibit the highest walking accessibility across all three supply scenarios, while the morning work scenario demonstrates the best walking accessibility among the four demand scenarios. (3) Gini coefficients exceeding 0.5 across all scenarios reveal severe resource allocation inequity. By establishing a dynamic supply–demand integration framework, this research advances methodological precision in accessibility evaluation, uncovers critical spatiotemporal mismatch patterns, and provides actionable insights for optimizing PHCI planning to promote spatial justice in urban health. Full article

(This article belongs to the Topic Sustainable Development and Coordinated Governance of Urban and Rural Areas Under the Guidance of Ecological Wisdom—2nd Edition)

► Show Figures

Figure 1

35 pages, 17539 KB

Open AccessArticle

Conceptual Neighborhood Graphs of Discrete Time Intervals

by Matthew P. Dube and Brendan P. Hall

ISPRS Int. J. Geo-Inf. 2026, 15(1), 39; https://doi.org/10.3390/ijgi15010039 - 12 Jan 2026

Abstract

Temporal reasoning is an important part of the field of time geography and spatio-temporal data science. Recent advances in qualitative temporal reasoning have developed a set of 74 relations that apply between discretized time intervals of at least two pixels each. While the [...] Read more.

Temporal reasoning is an important part of the field of time geography and spatio-temporal data science. Recent advances in qualitative temporal reasoning have developed a set of 74 relations that apply between discretized time intervals of at least two pixels each. While the identification of specific relations is important, the field of qualitative spatial and temporal reasoning relies on conceptual neighborhood graphs to address relational similarity. This similarity is paramount for generating essential decision support structures, notably reasonable aggregations of concepts into single terms and the determination of nearest neighbor queries. In this paper, conceptual neighborhood graphs of qualitative topological changes, with discretized temporal interval relations in the form of translation, isotropic scaling, and anisotropic scaling, are identified using data generated through a simulation protocol. The outputs of this protocol are compared to the extant literature regarding conceptual neighborhood graphs of the Allen interval algebra, demonstrating the theoretical accuracy of the work. This work supports the development of robust spatio-temporal artificial intelligence as well as the future development of spatio-temporal query systems upon the spatio-temporal stack data architecture. Full article

(This article belongs to the Special Issue Spatial Data Science and Knowledge Discovery)

► Show Figures

Figure 1

30 pages, 22514 KB

Open AccessArticle

Spatiotemporal Heterogeneity Analysis of Net Primary Productivity in Nanjing’s Urban Green Spaces Based on the DLCC–NPP Model: A Long-Term and Multi-Scenario Approach

by Yuhao Fang, Yuyang Liu, Yuan Wang, Yilun Cao and Yuning Cheng

ISPRS Int. J. Geo-Inf. 2026, 15(1), 38; https://doi.org/10.3390/ijgi15010038 - 12 Jan 2026

Abstract

In the context of the “Dual Carbon” goals, accurately predicting the spatiotemporal evolution of urban Net Primary Productivity (NPP) is crucial for resilient urban planning. While recent studies have coupled land use models with ecosystem models to project NPP dynamics, they often face [...] Read more.

In the context of the “Dual Carbon” goals, accurately predicting the spatiotemporal evolution of urban Net Primary Productivity (NPP) is crucial for resilient urban planning. While recent studies have coupled land use models with ecosystem models to project NPP dynamics, they often face challenges in acquiring high-resolution future vegetation parameters and typically overlook the stability of NPP under changing climates. To address these gaps, this study focuses on Nanjing and develops a long-term, multi-scenario analysis framework based on the Dynamic Land Cover–Climate Model (DLCC–NPP). This framework innovatively integrates the PLUS model with a Random Forest (RF) algorithm. By establishing a direct statistical mapping between macro-climate/micro-land cover and NPP, the RF model functions as a statistical downscaling tool. This approach bypasses the uncertainty accumulation associated with simulating future vegetation indices, enabling precise spatiotemporal NPP prediction at a 30 m resolution. Using this approach, we systematically analyzed the NPP dynamics from 2004 to 2044 under three SSP scenarios. The results revealed that Nanjing’s NPP exhibited a fluctuating upward trend, with urban forests contributing the highest productivity (mean NPP ~266.15 gC/m²). Crucially, the volatility analysis highlighted divergent response characteristics: forests demonstrated the highest stability and “buffering effect,” whereas grasslands and croplands showed high volatility and sensitivity to climate fluctuations. Spatially, a distinct “stable high-NPP core, decreasing periphery” pattern was identified, driven by the interaction of urban expansion and ecological conservation policies. In conclusion, the DLCC–NPP framework effectively overcomes the data scarcity bottleneck in future simulations and characterizes the spatiotemporal heterogeneity of vegetation carbon fixation in urban ecosystems, providing scientific support for optimizing green space patterns and enhancing urban ecological resilience in high-density cities. Full article

(This article belongs to the Topic Advances in Multi-Scale Geographic Environmental Monitoring: Ecosystem Differences and Multi-Scale Comparisons)

► Show Figures

Figure 1

23 pages, 2493 KB

Open AccessArticle

Rule-Based Scenario Classification Using Vehicle Trajectories

by Sungmo Ku and Jinho Lee

ISPRS Int. J. Geo-Inf. 2026, 15(1), 37; https://doi.org/10.3390/ijgi15010037 - 11 Jan 2026

Abstract

Ensuring the safety of autonomous driving systems (ADS) requires scenario-based testing that reflects the complexity and variability of real-world driving conditions. However, the nondeterministic nature of actual traffic environments makes physical testing costly and limited in scope, particularly for rare and safety-critical scenarios. [...] Read more.

Ensuring the safety of autonomous driving systems (ADS) requires scenario-based testing that reflects the complexity and variability of real-world driving conditions. However, the nondeterministic nature of actual traffic environments makes physical testing costly and limited in scope, particularly for rare and safety-critical scenarios. To address this, simulation has become a core component in validation by providing scalable, controllable, and repeatable testing environments. This study proposes a trajectory-based scenario classification framework that emphasizes both generality and interpretability. Specifically, we define a set of rule-based maneuver classification criteria using lateral acceleration patterns and apply them to simulated urban driving scenarios modeled with OpenSCENARIO. To address overlapping maneuver characteristics, a priority ordering of classification rules is introduced to resolve ambiguities. The proposed method was evaluated on a dataset comprising 7 types of maneuvers, including straight driving, lane changes, turns, roundabouts, and U-turns. Experimental results demonstrate the effectiveness of rule-driven classification based on vehicle trajectory dynamics and highlight the potential of this approach for structured scenario definition and validation in ADS simulation environments. Full article

(This article belongs to the Topic The Geography of Digital Twin: Concepts, Architectures, Modeling, AI and Applications)

► Show Figures

Figure 1

22 pages, 5754 KB

Open AccessArticle

Low-Cost Deep Learning for Building Detection with Application to Informal Urban Planning

by Lucas González, Jamal Toutouh and Sergio Nesmachnow

ISPRS Int. J. Geo-Inf. 2026, 15(1), 36; https://doi.org/10.3390/ijgi15010036 - 9 Jan 2026

Abstract

This article studies the application of deep neural networks for automatic building detection in aerial RGB images. Special focus is put on accuracy robustness in both well-structured and poorly planned urban scenarios, which pose significant challenges due to occlusions, irregular building layouts, and [...] Read more.

This article studies the application of deep neural networks for automatic building detection in aerial RGB images. Special focus is put on accuracy robustness in both well-structured and poorly planned urban scenarios, which pose significant challenges due to occlusions, irregular building layouts, and limited contextual cues. The applied methodology considers several CNNs using only RBG images as input, and both validation and transfer capabilities are studied. U-Net-based models achieve the highest single-model accuracy, with an Intersection over Union (

I o U

) of 0.9101. A soft-voting ensemble of the best U-Net models further increases performance, reaching a best ensemble

I o U

of 0.9665, improving over state-of-the-art building detection methods on standard benchmarks. The approach demonstrates strong generalization using only RGB imagery, supporting scalable, low-cost applications in urban planning and geospatial analysis. Full article

(This article belongs to the Special Issue Testing the Quality of GeoAI-Generated Data for VGI Mapping)

► Show Figures

Figure 1

30 pages, 3799 KB

Open AccessArticle

Geospatial Knowledge-Base Question Answering Using Multi-Agent Systems

by Jonghyeon Yang and Jiyoung Kim

ISPRS Int. J. Geo-Inf. 2026, 15(1), 35; https://doi.org/10.3390/ijgi15010035 - 8 Jan 2026

Abstract

Large language models (LLMs) have advanced geospatial artificial intelligence; however, geospatial knowledge-base question answering (GeoKBQA) remains underdeveloped. Prior systems have relied on handcrafted rules and have omitted the splitting of datasets into training, validation, and test sets, thereby hindering fair evaluation. To address [...] Read more.

Large language models (LLMs) have advanced geospatial artificial intelligence; however, geospatial knowledge-base question answering (GeoKBQA) remains underdeveloped. Prior systems have relied on handcrafted rules and have omitted the splitting of datasets into training, validation, and test sets, thereby hindering fair evaluation. To address these gaps, we propose a prompt-based multi-agent LLM framework (based on GPT-4o) that translates natural-language questions into executable GeoSPARQL. The architecture comprises an intent analyzer, multi-grained retrievers that ground concepts and properties in the OSM tagging schema and map geospatial relations to the GeoSPARQL/OGC operator inventory, an operator-aware intermediate representation aligned with SPARQL/GeoSPARQL 1.1, and a query generator. Our approach was evaluated on the GeoKBQA test set using 20 few-shot exemplars per agent. It achieved 85.49 EM (GPT-4o) with less supervision than fine-tuned baselines trained on 3574 instances and substantially outperformed a single-agent GPT-4o prompt. Additionally, we evaluated GPT-4o-mini, which achieved 66.74 EM in a multi-agent configuration versus 47.10 EM with a single agent. The observations showed that the multi-agent gain was higher for the larger model. Our results indicate that, beyond scale, the framework’s structure is important; thus, principled agentic decomposition yields a sample-efficient, execution-faithful path beyond template-centric GeoKBQA under a fair, hold-out evaluation protocol. Full article

(This article belongs to the Special Issue LLM4GIS: Large Language Models for GIS)

► Show Figures

Figure 1

39 pages, 20112 KB

Open AccessArticle

High-Spatiotemporal-Resolution Population Distribution Estimation Based on the Strong and Weak Perception of Population Activity Patterns

by Rui Li, Guangyu Liu, Hongyan Li and Jing Xia

ISPRS Int. J. Geo-Inf. 2026, 15(1), 34; https://doi.org/10.3390/ijgi15010034 - 8 Jan 2026

Abstract

Population activity drives urban development, and high-spatiotemporal-resolution population distribution provides critical insights for refined urban management and social services. However, mixed population activity patterns and spatial heterogeneity make simultaneous high-temporal- and -spatial-resolution estimation difficult. Therefore, we propose the High-Spatiotemporal-Resolution Population Distribution Estimation Based [...] Read more.

Population activity drives urban development, and high-spatiotemporal-resolution population distribution provides critical insights for refined urban management and social services. However, mixed population activity patterns and spatial heterogeneity make simultaneous high-temporal- and -spatial-resolution estimation difficult. Therefore, we propose the High-Spatiotemporal-Resolution Population Distribution Estimation Based on the Strong and Weak Perception of Population Activity Patterns (SWPP-HSTPE) method to estimate hourly population distribution at the building scale. During the weak-perception period, we construct a Modified Dual-Environment Feature Fusion model using building features within small-scale grids to estimate stable nighttime populations. During the strong-perception period, we incorporate activity characteristics of weakly perceived activity populations (minors and older people). Then, the Self-Organizing Map algorithm and spatial environment function purity are used to decompose mixed patterns of strongly perceived activity populations (young and middle-aged) and to extract fundamental patterns, combined with building types, for population calculation. Results demonstrated that the SWPP-HSTPE method achieved high-spatiotemporal-resolution population distribution estimation. During the weak-perception period, the estimated population correlated strongly with actual household counts (r = 0.72) and outperformed WorldPop and GHS-POP by 0.157 and 0.133, respectively. During the strong-perception period, the SWPP-HSTPE model achieves a correlation with hourly population estimates that is approximately 4% higher than that of the baseline model, while reducing estimation errors by nearly 2%. By jointly accounting for temporal dynamics and population activity patterns, this study provides valuable data support and methodological insights for fine-grained urban management. Full article

► Show Figures

Figure 1

24 pages, 33749 KB

Open AccessArticle

Ultra-Wideband System for Museum Visitors Tracking: Towards the Integration of the Positioning System with the Vision Sensors

by Angeliki Makellaraki, Vincenzo Di Pietra, Paolo Dabove and Milad Bagheri

ISPRS Int. J. Geo-Inf. 2026, 15(1), 33; https://doi.org/10.3390/ijgi15010033 - 8 Jan 2026

Abstract

Indoor positioning systems (IPSs) are increasingly applied in indoor settings where satellite-based GNSS signals are unavailable, including museums and other cultural heritage spaces. Within the META-MUSEUM project, we present a pilot study integrating an Ultra-Wideband (UWB) positioning system and an eye-tracking device to [...] Read more.

Indoor positioning systems (IPSs) are increasingly applied in indoor settings where satellite-based GNSS signals are unavailable, including museums and other cultural heritage spaces. Within the META-MUSEUM project, we present a pilot study integrating an Ultra-Wideband (UWB) positioning system and an eye-tracking device to monitor and quantify visitor behavior in a real museum environment. The absence of common timestamps between the two systems, and the presence of UWB signal noise, have been the main challenges to address. A cross-correlation–based synchronization method was developed to align the two independent UWB and eye-tracking datasets. Data were collected from 100 visitors, of whom 7 different clusters were considered based on the characteristics of the visitors. The results demonstrate the system’s feasibility and provide two complementary metrics, Normalized Engagement and Collective Engagement, which are used to quantify the duration and spatial distribution of visitor engagement at specific exhibits. This work establishes a scalable multi-sensor foundation by addressing practical deployment challenges under real-world conditions. These findings form the basis for the project’s broader goal of linking spatial visitor behavior with neurophysiological responses, opening new possibilities for improving visitor engagement and supporting interactive cultural heritage experiences. Full article

► Show Figures

Figure 1

16 pages, 9338 KB

Open AccessArticle

Integrated Revealing GIS Models to Monitor, Understand and Foresee the Spread of Diseases and Support Emergency Response

by Cristiano Pesaresi and Davide Pavia

ISPRS Int. J. Geo-Inf. 2026, 15(1), 32; https://doi.org/10.3390/ijgi15010032 - 8 Jan 2026

Abstract

The importance of GIS models to monitor the spread of infectious diseases and support emergency response has been underlined by a large body of literature and strengthened by the COVID-19 pandemic to identify possible solutions able to recognise spatio-temporal clusters and patterns, evaluate [...] Read more.

The importance of GIS models to monitor the spread of infectious diseases and support emergency response has been underlined by a large body of literature and strengthened by the COVID-19 pandemic to identify possible solutions able to recognise spatio-temporal clusters and patterns, evaluate the presence of acceleration factors and define specific actions. In the field of applied research on health geography and geography of safety, this work briefly displays the main aims of the project “Integrated revealing GIS models to monitor, understand and foresee the spread of diseases and support emergency response” and shows some illustrative applications. The basic assumption of the project is to test revealing models regarding key objectives of social utility, and one of its main aims is to elaborate GIS applications able to understand the spread of COVID-19, relating the geocalisations of the cases with specific variables. In order to provide targeted evidence able to better highlight local differences, a number of elaborations derived from (Arc)GIS models and based on data regarding COVID-19 according to sex, age and healthcare facilities in the Rome municipality (Italy) are presented and contextualised as examples, also replicable for precision preparedness. Full article

(This article belongs to the Special Issue HealthScape: Intersections of Health, Environment, and GIS&T (2nd Edition))

► Show Figures

Figure 1

19 pages, 2460 KB

Open AccessArticle

GeoAI in Temperature Correction for Rice Heat Stress Monitoring with Geostationary Meteorological Satellites

by Han Luo, Binyang Yang, Lei He, Yuxia Li, Dan Tang and Huanping Wu

ISPRS Int. J. Geo-Inf. 2026, 15(1), 31; https://doi.org/10.3390/ijgi15010031 - 8 Jan 2026

Abstract

To address the challenge of obtaining high-spatiotemporal-resolution and high-precision temperature grids for agricultural meteorological monitoring, this research focuses on rice heat stress monitoring with the China Meteorological Administration Land Data Assimilation System (CLDAS) and develops a temperature correction model that synergizes physical mechanisms [...] Read more.

To address the challenge of obtaining high-spatiotemporal-resolution and high-precision temperature grids for agricultural meteorological monitoring, this research focuses on rice heat stress monitoring with the China Meteorological Administration Land Data Assimilation System (CLDAS) and develops a temperature correction model that synergizes physical mechanisms with a data-driven strategy by introducing a GeoAI framework. Ensemble learning methods (XGBoost, LightGBM, and Random Forest) were utilized to process a comprehensive set of predictors, integrating dynamic surface features derived from FY-4 satellite’s high-frequency observation data. The data comprised surface thermal regime metrics, specifically the daily maximum land surface temperature (LST_max) and its diurnal range (LST_{max_min}), along with vegetation indices including the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI). Further, topographic attributes derived from a digital elevation model (DEM) were incorporated, such as slope, aspect, the terrain ruggedness index (TRI), and the topographic position index (TPI). The approach uniquely capitalized on the temporal resolution of geostationary data to capture the diurnal land surface dynamics crucial for bias correction. The proposed models not only enhanced temperature data quality but also achieved impressive accuracy. Across China, the root mean square error (RMSE) was reduced to 1.04 °C, mean absolute error (MAE) to 0.53 °C, and accuracy (ACC) to 0.97. Additionally, the most notable improvement was that the RMSE decreased by nearly 50% (from 2.17 °C to 1.11 °C), MAE dropped from 1.48 °C to 0.80 °C, and ACC increased from 0.72 to 0.96 in the southwestern region of China. The corrected rice heat stress data (2020–2023) indicated that significant negative correlations exist between yield loss and various heat stress metrics in the severely affected middle and lower Yangtze River region. The research confirms that embedding geostationary meteorological satellites within a GeoAI framework can effectively enhance the precision of agricultural weather monitoring and related impact assessments. Full article

(This article belongs to the Topic Geospatial AI: Systems, Model, Methods, and Applications)

► Show Figures

Figure 1

23 pages, 2540 KB

Open AccessArticle

Sensing Envelopes: Urban Envelopes in the Smart City Ontology Framework

by Andrej Žižek, Peter Šenk and Kaja Pogačar

ISPRS Int. J. Geo-Inf. 2026, 15(1), 30; https://doi.org/10.3390/ijgi15010030 - 8 Jan 2026

Abstract

The paper examines the phenomenon of urban envelopes, a conceptual parallel to building envelopes, which is considered an emerging theme in studies of the built environment. The term ‘envelope’ refers to various physical and non-physical occurrences in the built environment that delimit, enclose, [...] Read more.

The paper examines the phenomenon of urban envelopes, a conceptual parallel to building envelopes, which is considered an emerging theme in studies of the built environment. The term ‘envelope’ refers to various physical and non-physical occurrences in the built environment that delimit, enclose, or demarcate spatial configurations. In the first part of the paper, six distinct types of urban envelopes are identified: physical, programmatic, technological, ecological, environmental, and representational. These are defined based on a systematic literature review to clarify their form, role, and meaning in the context of contemporary cities. All six urban envelope types are formalised using ontology-building methods in Protégé and visualised through WebVOWL, producing domain-agnostic RDF/OWL models that support semantic interoperability. The results provide a concise definition of urban envelopes, which are becoming increasingly relevant in their non-physical representations, such as spaces of control (surveillance of public urban spaces), dynamic environmental and ecological phenomena (pollution, heat islands, and more), temporal or dynamic definitions of space use, and many others in the context of contemporary smart city development. The analysis of possible alignment with existing smart city-related ontologies is presented. By providing the methodology for linking urbanistic principles with data-driven smart city frameworks, the paper provides a unified methodological foundation for incorporating such emerging spatial phenomena into formal urban models. Full article

► Show Figures

Figure 1

39 pages, 2885 KB

Open AccessArticle

Usability Assessment Framework for Crowdsensing Data and the Implicit Spatiotemporal Information

by Ying Chen, He Zhang, Jixian Zhang, Jing Shen and Yahang Li

ISPRS Int. J. Geo-Inf. 2026, 15(1), 29; https://doi.org/10.3390/ijgi15010029 - 7 Jan 2026

Abstract

Crowdsensing data serves as a crucial resource for supporting spatiotemporal applications and services. However, its inherent heterogeneity and quality uncertainty present significant challenges for data usability assessment: the evaluation methods are difficult to standardize due to the diverse types of data; assessment dimensions [...] Read more.

Crowdsensing data serves as a crucial resource for supporting spatiotemporal applications and services. However, its inherent heterogeneity and quality uncertainty present significant challenges for data usability assessment: the evaluation methods are difficult to standardize due to the diverse types of data; assessment dimensions are predominantly confined to internal quality attributes; and a comprehensive framework for data usability evaluation remains lacking. To address these challenges, this study proposes an innovative, multi-layered usability assessment framework applicable to six major categories of crowdsensing data: specialized spatial data, Internet of Things (IoT) sensing data, trajectory data, geographic semantic web, scientific literature, and web texts. Building upon a systematic review of existing research on data quality and usability, our framework conducts a comprehensive evaluation of data efficiency, effectiveness, and satisfaction from dual perspectives—data sources and content. We present a complete system comprising primary and secondary indicators and elaborate on their computation and aggregation methods. Indicator weights are determined through the Analytic Hierarchy Process (AHP) and expert consultations, with sensitivity analysis performed to validate the robustness of the framework. The practical applicability of the framework is demonstrated through a case study of constructing a spatiotemporal knowledge graph, where we assess all six types of data. The results indicate that the framework generates distinguishable usability scores and provides actionable insights for improvement. This framework offers a universal standard for selecting high-quality data in complex decision-making scenarios and facilitates the development of reliable spatiotemporal knowledge services. Full article

► Show Figures

Figure 1

21 pages, 4559 KB

Open AccessArticle

Language-Guided Spatio-Temporal Context Learning for Next POI Recommendation

by Chunyang Liu and Chuxiao Fu

ISPRS Int. J. Geo-Inf. 2026, 15(1), 28; https://doi.org/10.3390/ijgi15010028 - 6 Jan 2026

Abstract

With the proliferation of mobile internet and location-based services, location-based social networks (LBSNs) have accumulated extensive user check-in data, driving the advancement of next Point-of-Interest (POI) recommendation systems. Although existing approaches can model sequential dependencies and spatio-temporal patterns, they often fail to fully [...] Read more.

With the proliferation of mobile internet and location-based services, location-based social networks (LBSNs) have accumulated extensive user check-in data, driving the advancement of next Point-of-Interest (POI) recommendation systems. Although existing approaches can model sequential dependencies and spatio-temporal patterns, they often fail to fully capture users’ dynamic preferences under varying spatio-temporal contexts and lack effective integration of fine-grained semantic information. To address these limitations, this paper proposes Language-Guided Spatio-Temporal Context Learning for Next POI Recommendation (LSCNP). It employs a pre-trained BERT model to encode multi-dimensional spatio-temporal context—including geographic coordinates, visiting hours, and surrounding POI categories—into structured textual sequences for semantic understanding; constructs dual-graph structures to model spatial constraints and user transition patterns; and introduces a contrastive learning module to align spatio-temporal context with POI features, enhancing the discriminability of representations. A Transformer-based sequential encoder is adopted to capture long-range dependencies, while a neural matrix factorization decoder generates final recommendations. Experiments on three real-world LBSN datasets demonstrate that LSCNP consistently outperforms state-of-the-art baselines. Ablation studies and hyperparameter analyses further validate the contribution of each component to the overall performance. Full article

► Show Figures

Figure 1

Journal Description

ISPRS International Journal of Geo-Information

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Further Information

Guidelines

MDPI Initiatives

Follow MDPI