Geographic information science and the analysis of place and health

The representation of place is a key theoretical advancement that geographic information science can offer to improve the understanding of environmental determinants of health, but developing robust computational representations of place requires a substantial departure from conventional notions of geographic representation in GIS. Unlike conventional GIS representations based on either objects or locations, we suggest that place representation should incorporate dynamic subjective, experiential, and relational aspects of place, as the influence of place on health behavior concerns not only the features that can be objectively observed at a particular location but also the environmental perceptions of the individual, as molded by biological, social, and experiential characteristics. In addition, assessments of environmental exposures on health outcomes should focus on individuals' time–activity patterns and microenvironment profiles, which form a potentially unique personalized exposure environment for each individual. Addressing these representational challenges via collaborative research has the potential to advance both geographic information science and health research.     

Modeling Urban Land Use Changes Using Support Vector Machines

Support Vector Machines (SVM) is a machine learning (ML) algorithm commonly applied to the classification of remotely sensing data and more recently for modeling land use changes. However, in most geospatial applications the current literature does not elaborate on specifications of the SVM method with respect to data sampling, attribute selection and optimal parameters choices. Therefore the main objective of this study is to present and investigate the SVM technique for modeling urban land use change. The SVM model building procedure is presented together with the detailed evaluation of the output results with respect to the choice of datasets, attributes and the change of SVM parameters. Geospatial datasets containing nine land use classes and spatial attributes for the Municipality of Zemun, Republic of Serbia were used for years 2001, 2003, 2007 and 2011. The Correlation‐based Feature Subset method, kappa coefficient, Area Under Receiver Operating Characteristic Curve (AUC) and kappa simulation were used to perform the model evaluation and compare the model outputs with the real land use datasets. The obtained results indicate that the SVM‐based models perform better when implementing balanced data sampling, reduced data sets to informative subsets of attributes and properly identify the optimal learning parameters.     

A hybrid approach to modeling territorial control in violent armed conflicts

Territorial control is central to the understanding of violent armed conflicts, yet reliable and valid measures of this concept do not exist. We argue that geospatial analysis provides an important perspective to measure the concept. In particular, measuring territorial control can be seen as an application of calculating service areas around points of control. The modeling challenge is acute for areas with limited road infrastructure, where no complete network is available to perform the analysis, and movements largely occur off road. We present a new geospatial approach that applies network analysis on a hybrid transportation network with both actual road data and hexagon‐fishnet‐based artificial road data representing on‐road and off‐road movements, respectively. Movement speed or restriction can be readily adjusted using various input data. Simulating off‐road movement with hexagon‐fishnet‐based artificial road data has a number of advantages including scalability to small or large study areas and flexibility to allow all‐directional travel. We apply this method to measuring territorial control of armed groups in Sub‐Saharan Africa where inferior transport infrastructure is the norm. Based on the Uppsala Conflict Data Program's (UCDP) Georeferenced Event Data (GED) as well as spatial data on terrain, population locations, and limited transportation networks, we enhance the delineation of the specific areas directly controlled by each warring party during civil wars within a given travel time.     

Single frequency processing of Ørsted GPS radio occultation measurements

The Global Positioning System (GPS) radio occultation measurements obtained using the TurboRogue GPS receiver on the Danish satellite Ørsted have been processed using the single frequency method. Atmospheric profiles of refractivity and temperature are derived and validated against numerical weather prediction data from the European Centre for Medium-Range Weather Forecast (ECMWF). Results from the Ørsted GPS measurement campaign in February 2000 indicate that the single frequency method can provide retrievals with accuracy comparable to that of using two frequencies. From comparisons between measured dry temperature profiles and corresponding dry temperature profiles derived from ECMWF analysis fields, we find a mean difference of less than 0.5 K and a standard deviation of 2–4 K between 500 and 30 hPa in height. Above 30 hPa the impact of the ionosphere becomes more dominant and more difficult to eliminate using the single frequency method, and the results show degraded accuracy when compared to previous analysis results of occultation data from other missions using the dual frequency method. At latitudes less than 40° (denoted low latitudes), the standard deviation is generally smaller than at latitudes higher than 40° (denoted high latitudes). A small temperature bias is observed centered at 200 hPa for low latitudes and at 300 hPa for high latitudes. This indicates that the ECMWF analyses do not adequately resolve the tropopause temperature minimum. In the lowest part of the troposphere an observed warm bias is thought to be due to erroneous tracking of the GPS signal in cases of atmospheric multipath propagation.     

Space-Time Intelligence Systems: Technology, applications and methods

Geographic Information System (GIS) software is constrained, to a greater or lesser extent, by a static world view that is not well-suited to the representation of time (Goodchild 2000). Space Time Intelligence System (STIS) software holds the promise of relaxing some of the technological constraints of spatial only GIS, making possible visualization approaches and analysis methods that are appropriate for temporally dynamic geospatial data. This special issue of the Journal of Geographical Systems describes some recent advances in STIS technology and methods, with an emphasis on applications in public health and spatial epidemiology.The STIS expert workshops were funded in part by grants R01CA092669 and R01CA096002 from the National Cancer Institute, and by grants R43-ES010220 and R44-ES010220 from the National Institute of Environmental Health Sciences. Gillian AvRuskin provided cheerful editorial assistance. We thank the participants at the workshops for providing invaluable expertise and critical insights.     

Similarity measurement on human mobility data with spatially weighted structural similarity index (SpSSIM)

Understanding diverse characteristics of human mobility provides profound knowledge of urban dynamics and complexity. Human movements are recorded in a variety of data sources and each describes unique mobility characteristics. Revealing similarity and difference in mobility data sources facilitates grasping comprehensive human mobility patterns. This study introduces a new method to measure similarities on two origin–destination (OD) matrices by spatially extending an image‐assessment tool, the structural similarity index (SSIM). The new measurement, spatially weighted SSIM (SpSSIM), utilizes weight matrices to overcome the SSIM sensitivity issue due to the ordering of OD pairs by explicitly defining spatial adjacency. To evaluate SpSSIM, we compared performances between SSIM and SpSSIM with resampling the orders of OD pairs and conducted bootstrapping to test the statistical significance of SpSSIM. As a case study, we compared OD matrices generated from three data sources in San Diego County, CA: U.S. Census‐based Longitudinal Employer–Household Dynamics Origin–Destination employment statistics, Twitter, and Instagram. The case study demonstrated that SpSSIM was able to capture similarities of mobility patterns between datasets that varied by distance. Some regions showed local dissimilarity while the global index indicated they were similar. The results enhance the understanding of complex mobility patterns from various datasets, including social media.     

Improved Priority‐Flood method for depression filling by redundant calculation optimization in local micro‐relief areas

Depression filling is a critical step in distributed hydrological modeling using digital elevation models (DEMs). The traditional Priority‐Flood (PF) approach is widely used due to its relatively high efficiency when dealing with a small‐sized DEM. However, it seems inadequate and inefficient when dealing with large high‐resolution DEMs. In this work, we examined the relationship between the PF algorithm calculation process and the topographical characteristics of depressions, and found significant redundant calculations in the local micro‐relief areas in the conventional PF algorithm. As such calculations require more time when dealing with large DEMs, we thus propose a new variant of the PF algorithm, wherein redundant points and calculations are recognized and eliminated based on the local micro‐relief water‐flow characteristics of the depression‐filling process. In addition, depressions and flatlands were optimally processed by a quick queue to improve the efficiency of the process. The proposed method was applied and validated in eight case areas using the Shuttle Radar Topography Mission digital elevation model (SRTM‐DEM) with 1 arc‐second resolution. These selected areas have different data sizes. A comparative analysis among the proposed method, the Wang and Liu‐based PF, the improved Barnes‐based PF, the improved Zhou‐based PF, and the Planchon and Darboux (P&D) algorithms was conducted to evaluate the accuracy and efficiency of the proposed algorithm. The results showed that the proposed algorithm is 43.2% (maximum) faster than Wang and Liu's variant of the PF method, with an average of 31.8%. In addition, the proposed algorithm achieved similar performance to the improved Zhou‐based PF algorithm, though our algorithm has the advantage of being simpler. The optimal strategies using the proposed algorithm can be employed in various landforms with high efficiency. The proposed method can also achieve good depression filling, even with large amounts of DEM data.     

Robust abnormality detection methods for spatial search of radioactive materials

Radiological dirty bombs and improvised nuclear devices pose a significant threat to both public health and national security. Growing networks of radiation sensors have been deployed by a number of offices within the U.S. and international agencies. Detecting such threats while minimizing false alarm rates presents a considerable challenge to homeland security and public health. This research aims to achieve a higher probability of detection with a lower probability of false alarms. It focuses on the local spatial instability of radiation levels in order to detect radioactive materials based on robust outlier detection methods. Our approach includes a three‐step abnormality detection method consisting of one‐dimensional robust outlier detection for all gamma‐ray counts, a density‐based clustering analysis, and a two‐dimensional robust outlier detection method using a bagplot, based on spatial associations. The effectiveness of the method proposed is demonstrated through a case study, wherein radioactive materials are detected in urban environments, and its performance is compared with alternative methods employing a k‐sigma approach, local Getis–Ord () statistic, and the goodness of fit of the Poisson distribution.     

Self-organised clustering for road extraction in classified imagery

The extraction of road networks from digital imagery is a fundamental image analysis operation. Common problems encountered in automated road extraction include high sensitivity to typical scene clutter in high-resolution imagery, and inefficiency to meaningfully exploit multispectral imagery (MSI). With a ground sample distance (GSD) of less than 2 m per pixel, roads can be broadly described as elongated regions. We propose an approach of elongated region-based analysis for 2D road extraction from high-resolution imagery, which is suitable for MSI, and is insensitive to conventional edge definition. A self-organising road map (SORM) algorithm is presented, inspired from a specialised variation of Kohonen's self-organising map (SOM) neural network algorithm. A spectrally classified high-resolution image is assumed to be the input for our analysis. Our approach proceeds by performing spatial cluster analysis as a mid-level processing technique. This allows us to improve tolerance to road clutter in high-resolution images, and to minimise the effect on road extraction of common classification errors. This approach is designed in consideration of the emerging trend towards high-resolution multispectral sensors. Preliminary results demonstrate robust road extraction ability due to the non-local approach, when presented with noisy input.