Christian Tominski, Hans-Jörg Schulz, University of Rostock

Understanding how data evolves in space and time is an essential task in many application domains. Despite the numerous visual methods (e.g., showing the data on a map or plotting a time graph) that have been proposed to facilitate this task, the exploration of data with references to space and time still remains challenging. In this work, we present a novel concept for visualizing spatio-temporal data that refer to 2D geographical space and 1D linear time. The idea is to construct a non-planar slice – called the Great Wall of Space-Time – through the 3D (2D+1D) space-time continuum. Different visual representations can be projected onto the wall in order to display the data. We suggest using the wall’s vertical extent to map the dimension of time and to color-code individual bricks within the wall. Alternatively, a parallel coordinates plot can be shown at the wall. Compared to existing approaches, the wall has the advantage that it shows a closed path through space with no gaps between the information-bearing pixels on the screen. Hence, our novel visualization has the potential to be a useful addition to the user’s toolbox of techniques for exploring the spatial and temporal evolution of data.

Cassandra Lee, Rodolphe Devillers, and Orland Hoeber, Memorial University of Newfoundland, Canada

The visual analysis of spatio-temporal data in a geospatial application can be a challenging task since these systems primarily focus on mapping and exploration of the spatial dimension of data, and may not fully exploit the temporal component. This paper proposes two methods for zooming and panning through temporal data at multiple temporal resolutions within a multi-touch geovisual analytics environment. The first method uses multiple linked time sliders allowing analysts to interactively zoom by selecting data at different levels of temporal resolution. The second method combines a main time slider with a dynamic radial navigation tool for temporal zooming. Both approaches support panning within the coarsest resolution of data without adjusting the selection at the finer temporal resolutions, and the analysts have the option of zooming in to as many levels of resolution as are contained within their datasets. The approaches were implemented in a multi-touch geovisual analytics software prototype.

Caglar Koylu, Diansheng Guo, University of South Carolina, USA

Understanding the structure and evolution of dynamic spatial networks is crucial for various fields such as genealogy, epidemiology and urban planning. The layout of a dynamic spatial network is formed by the geographic coordinates of nodes (individuals) and edges that metaphorically represent relations (e.g., family, friendship). Nodes and edges of such a network evolve (change) over time. Dynamic spatial networks are usually large and multi-relational because there can be different type of relations (e.g., family, friendship, membership in an organization). Visualization can potentially play a key role in analyzing and understanding dynamic networks. To address the challenges in visualizing dynamic and multi-relational spatial networks, we introduce a new visual analytics approach to discover community structures and facilitate the comprehension of structural changes in a dynamically evolving spatial network. We employ an animation approach and visualize a sequence of dynamic layouts. By allowing user interactions and controls, the proposed approach enhances the capability of animation in analysis tasks; facilitates the discovery of community structures and comprehension of structural changes in a dynamically evolving spatial network. We demonstrate our approach by using family tree data which also include information on migration of family members.

Qiuju Zhang, Faculty of Geoinformation Science and Earth Observation (ITC), University of Twente
Jo Wood, Department of Information Science at City University London
Aidan Slingsby, Department of Information Science at City University London
Jason Dykes, Department of Information Science at City University London
Menno-Jan Kraak, Faculty of Geoinformation Science and Earth Observation (ITC), University of Twente
Connie A. Blok, Faculty of Geoinformation Science and Earth Observation (ITC), University of Twente
Rein Ahas, Department of Geography, Tartu University

Movement data about individuals is being collected more frequently, in larger volumes and at higher spatial and temporal resolutions. Visually exploring these data can facilitate analysis of spatio-temporal movement behaviour. Most visualizations focus on representing attributes of movement and very little attention has been dedicated to the context of such movements. We introduce a context sequence view to support comparison tasks in a visual search for patterns in the attributes of moving objects together with their context of movements. Here, land use categories are used to establish context. In the sequence view, data for an individual are depicted as a row of rectangles and multiple individuals are arranged vertically. The rectangle’s positions and sequences of same coloured rectangles represent the time and duration for which an individual stayed in a particular land use category. We focus on design considerations for various views of the data, and use a case study in which patterns are sought in the consumption of urban space by suburban dwellers in Tallinn (Estonia). The context sequence view combines contextrelated colour display, temporally sequencial layout, relative geographic location options, hierarchical sorting strategies, and interaction techniques to provide a visualization of multiple individuals’ movements in a context. The results show temporal regularity in space consumption for each individual and differences of overall pattern between weekdays and weekends. Sorting individuals based on social characteristics enable an analyst to reveal relations between sequences and social factors.

T. von Landesberger, Technische Universität Darmstadt, Germany
S. Bremm, Technische Universität Darmstadt, Germany
T. Schreck, Universität Konstanz, Germany
D. W. Fellner, Technische Universität Darmstadt and Fraunhofer IGD, Darmstadt, Germany

The study of movement data is an important task in a variety of domains such as transportation, biology, or finance. Often, the data objects are grouped (e.g., countries by continents, etc.). Then, the analysis focuses on three main aspects: (a) Behavior of an individual in the context of its group, (b) dynamics of a given group, and (c) comparison of the behavior of multiple groups. Analysis of group movement data can be effectively supported by data analysis and visualization. Feature-based approaches have shown useful to describe and analyze movement data. However, previous approaches were limited as they did not cover a broad range of situations and required manual feature monitoring. We extend the set of movement analysis features and add automatic analysis of the features for filtering for interesting parts in the movement data. Users can easily detect new interesting characteristics such as outliers, trends and task-dependent data patterns even in large sets of data points over long time horizons. We demonstrate usefulness of our system on a real-world data set from financial domain.

Yi Qiang, Department of Geography, Ghent University, Belgium
Steven Logghe, Be-Mobile NV, Belgium
Philippe De Maeyer, Department of Geography, Ghent University, Belgium
Nico Van de Weghe, Department of Geography, Ghent University, Belgium

Many disciplines are faced with the problem of handling linear data, such as time series and traffic speed along a road. This work introduces an innovative visual representation for linear data, the Continuous Triangular Model (CTM). In CTM, values of all subintervals within the linear data are displayed in a two-dimensional continuous field, where every position corresponds to a specific subinterval of the linear data. The value at a position is derived through a certain calculation (e.g., average or summation) of the linear data within the subinterval it represents. CTM thus provides an explicit overview of the linear data in all different scales. This work shows how CTM can bring added value to the visual analysis of different types of linear data. We also show how the coordinate interval space in CTM can support the analysis of multiple linear datasets through the methods of spatial analysis, including map algebra and cartographic modelling. Real-world datasets and scenarios will be employed to demonstrate the usefulness of this representation in analytical tasks.

Seyed Hossein Chavoshi, Bernard De Baets, Tijs Neutens, Guy De Tré, Nico Van de Weghe, Ghent University, Belgium

Increased affordability and deployment of advanced tracking technologies have led many researchers from different domains to dig into the huge spatio-temporal sets of movement data in order to extract informative knowledge. The discovered knowledge may help to better understand the behaviour of moving objects. Basically, two different approaches can be considered in the analysis of moving objects: the quantitative analysis and the qualitative analysis. This research focuses on the latter using the Qualitative Trajectory Calculus (QTC) – a calculus that represents and reasons about qualitative data on moving point objects (MPOs) –, and establishes a framework to analyse movement behaviour of multiple MPOs. This paper introduces a visualisation technique called Sequence Signature (SeSi), a way of mapping QTC movement patterns in a 2D indexed rasterised space, to measure the similarities and differences of movement patterns of multiple MPOs. The results show that the proposed method can be effectively used to analyse interaction of multiple MPOs in many domains. The applicability of the proposed methodology is illustrated by means of a practical example of comparing Samba dance movements during different time intervals.

Gennady Andrienko, Fraunhofer Institute IAIS, Sankt Augustin, Germany
Natalia Andrienko, Fraunhofer Institute IAIS, Sankt Augustin, Germany
Ana-Maria Olteanu Raimond, Orange Labs R&D, Paris, France
Juergen Symanzik, Utah State University, Logan, USA
Cezary Ziemlicki, Orange Labs R&D, Paris, France

Data reflecting movements of people, such as GPS or GSM tracks, can be a source of information about mobility behaviors and activities of people. Such information is required for various kinds of spatial planning in the public and business sectors. Movement data by themselves are semantically poor. Meaningful information can be derived by means of interactive visual analysis performed by a human expert; however, this is only possible for data about a small number of people. We suggest an approach that allows scaling to large datasets reflecting movements of numerous people. It includes extracting stops, clustering them for identifying personal places of interest (POIs), and creating temporal signatures of the POIs characterizing the temporal distribution of the stops with respect to the daily and weekly time cycles and the time line. The analyst can give meanings to selected POIs based on their temporal signatures (i.e., classify them as home, work, etc.), and then POIs with similar signatures can be classified automatically. We demonstrate the possibilities for interactive visual semantic analysis by example of GSM and GPS data. GPS data allow inferring richer semantic information but only temporal signatures may be insufficient for interpreting short stops. We plan to develop an intelligent system that learns how to classify personal places and trips while a human analyst visually analyzes and semantically annotates selected portions of movement data.

Olufemi A. Omitaomu, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Christopher S. Maness, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Ian S. Kramer, Department of Geography, University of South Carolina, Columbia, SC 29208, USA
Jeffrey B. Kodysh, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Budhendra L. Bhaduri, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Chad A. Steed, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Rajasekar Karthik, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Philip J. Nugent, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Aaron T. Myers, Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA

We present an integrated geovisual analytics framework for utility consumers to interactively analyze and benchmark their energy consumption. The framework uses energy and property data already available with the utility companies and county governments respectively. The motivation for the developed framework is the need for citizens to go beyond the conventional utility bills in understanding the patterns in their energy consumption. There is also a need for citizens to go beyond one-time improvements that are often not monitored and measured over time. Some of the features of the framework include the ability for citizens to visualize their historical energy consumption data along with weather data in their location. The quantity of historical energy data available is significantly more than what is available from utility bills. An overlay of the weather data provides users with a visual correlation between weather patterns and their energy consumption patterns. Another feature of the framework is the ability for citizens to compare their consumption on an aggregated basis to that of their peers – other citizens living in houses of similar size and age – and within the same or different geographical boundaries, such as subdivision, zip code, or county. The users could also compare their consumption to others based on the size of their family and other attributes. This feature could help citizens determine if they are among the “best in class”. The framework can also be used by the utility companies to better understand their customers, plan their services, and manage their operational constraints. To make the framework easily accessible, it is developed to be compatible with mobile consumer electronics devices.

Patrick Köthur, Mike Sips, Andrea Unger, Julian Kuhlmann, Doris Dransch, Helmholtz Centre Potsdam – GFZ German Research Centre for Geosciences

A broad range of measurement devices and computer simulations produce geospatial time series that describe a wide variety of processes of System Earth. A major challenge in the analysis of such data is the complexity of the described processes which requires a simultaneous assessment of the data's spatial and temporal dimensions. To address this task, geoscientists often use automated analysis methods to compute a compact description of the data that, ideally, comprises the characteristic spatial states of the process under study and their occurrence over time.The results of such automated analyses depend on the parameterization of the applied methods, especially the number of spatial states extracted from the data. A particular number of spatial states, however, may only reflect certain spatial or temporal aspects. We introduce a visual analytics approach that uses hierarchical clustering to compute interactive visual summaries that let users vary the number of spatial situations extracted from the data. A visual summary depicts a particular number of spatial situations, and their occurrence over time. We provide four visualization components that facilitate progressive, interactive exploration of visual summaries to detect characteristic spatiotemporal patterns in geospatial time series. Our approach is the result of a close collaboration with geoscientists and a comprehensive task and requirement analysis. In an exemplary analysis of ocean observational data, we show that our approach can help geoscientists to gain understanding of geospatial time series.

Stefan Buschmann, Matthias Trapp, Jürgen Döllner, Hasso-Plattner-Institut, Universität Potsdam

The visualization of trajectories and their attributes represents an essential functionality for spatio-temporal data visualization and analysis. Many visualization methods, however, focus mainly on sparse 2D movements or consider only the 2D components of movements. This paper is concerned with true 3D movement data, i.e., movements that take place in the three-dimensional space and which characteristics significantly depend an all dimensions. In this case, spatio-temporal visualization approaches need to map all three spatial dimensions together with required mappings for associated attributes. We describe visualization approaches for true 3D movement data and evaluate their application within 3D geovirtual environments. We also identify challenges and propose approaches for the interactive visualization of 3D movement data using 3D geovirtual environments as scenery.

Andrew Chung-Da Lee, Claus Rinner, Ryerson University, Canada

Neighbourhoods in the City of Toronto are continuously changing. The objective of this study was to visualize socioeconomic status based on 1996, 2001, and 2006 census data, and to identify characteristic changes at the Census tract level. Self-Organizing Maps (SOM) were used to represent the spatiotemporal patterns of change. The similarities and differences between neighbourhood trajectories were explored, and trajectory length was mapped as an indicator of the magnitude of demographic change. The results suggest that the city was divided into three areas by 2006. Greater socioeconomic change occurred in eastern Toronto between 1996 and 2001 and in western Toronto between 2001 and 2006. Additionally, the City appears to have become less socioeconomically diverse during the study period. The paper also discusses technical issues with the census data and limitations of the SOM tool within GeoVista Studio.

Rene Rosenbaum, Hans-Jörg Schulz, Steffen Hadlak, Heidrun Schumann, University of Rostock, Germany

For multilevel data, levels are not the result of a hierarchical aggregation, but contain independently produced data. While first visualization techniques for this kind of data exist, suitable interactive exploration techniques have rarely been investigated so far. In this paper, we introduce means for the analysis, representation, and exploration of spatiotemporal multilevel data as first solutions to fill this gap. Particularly, these are analysis strategies serving to order the data appropriately, a novel data display providing an overview to the data as well as level- and path-based exploration strategies supporting systematic data browsing. Appropriate level and path filtering techniques are also provided. Results from an application of this technology to election data indicate their usefulness for conducting visual analysis tasks in spatiotemporal multilevel data.

Xiaojing Wu, Raul Zurita-Milla, Menno-Jan Kraak, Department of Geo-Information Processing, University of Twente, The Netherlands

The changes of phenological phenomena closely depend on climate change. Synchronization of weather data at different scales would be identified to help solve phenological problems. This extended abstract visually explore and address temporal scale issues because they have much less been solved compared with spatial ones. We propose a geovisual analytic workflow that combines clustering and cycle plot for discovery of synchronization in weather data at daily, weekly and monthly scale. Clustering method is employed to group calendar-based time series to cluster-based data which can prevent possible false patterns recognized based on calendar recorded data. Kohonen self-organizing maps (SOM) algorithm is used for clustering. With station-sorted and year-sorted of weather data as input variable separately, the results of SOM reveal totally different temporal patterns. Cycle plot is re-looked from a particular perspective as cluster-based time series are seen as basic unit instead of calendar date. Finally synchronization and trends of weather data will reveal itself when cycle plot is combined with clustering. We also aggregate daily weather data to weekly and monthly and synchronization results are compared at multiple temporal granularities.

Katerina Vrotsou, Anders Ynnerman, Matthew Cooper, Linköping University, Sweden

The study of human activity in space and time is an inherent part of human geography. In order to perform such studies data on the time use of individuals, in terms of sequence and timing of performed activities, are collected and analysed. A common assumption when analysing individuals time use is that groups that exhibit similar background and demographic characteristics also display similarities in how they use their time to structure their daily lives. In this paper we set out to investigate the correctness of such assumptions. We propose a visual analytics process, based on sequence similarity measures tailored for event-based data such as performed activity sequences. The process allows an analyst to retrieve similarly behaving records according to user-selected similarity preferences, and interactively explore aspects of this similarity in a multiple linked-view environment.

Laura G. Tateosian, NCSU
Helena Mitasova, NCSU
Sidharth Thakur, RENCI
Emily Russ, NCSU
Bruce Blundell, US Army Topographic Engineering Center, Alexandria, VA

In the Outer Banks of North Carolina, water, wind, gravitation, vegetation, and human activity continuously alter landscape surfaces over time. Scientific visualizations provide an important tool for understanding these processes to guide land use and management decisions. In this region, repeated high-resolution terrain scans have been conducted to generate time series of point clouds to represent landscape surfaces at snapshots in time. Visualizing the combination of spatial and temporal components along with semantic attributes poses a complex challenge. Common techniques rely on showing pairwise time step elevation differences or the evolution of volume change. Scientists have also started using isosurfaces to trace the movement of a fixed elevation through a series of time steps. This paper focuses on the space-time cube as it applies to land surface elevation. We illustrate the concept with a case study of Outer Banks, NC barrier island topography where wind and waves are the main change agents.