Grantee Research Project Results
1998 Progress Report: An Integrated Approach for Effective Representation and Analysis of Space/Time Environmental Data
EPA Grant Number: R825195Title: An Integrated Approach for Effective Representation and Analysis of Space/Time Environmental Data
Investigators: Peuquet, Donna J. , MacEachren, Alan
Institution: Pennsylvania State University
EPA Project Officer: Aja, Hayley
Project Period: October 10, 1996 through October 9, 1999 (Extended to December 30, 2000)
Project Period Covered by this Report: October 10, 1997 through October 9, 1998
Project Amount: $585,086
RFA: High Performance Computing (1996) RFA Text | Recipients Lists
Research Category: Human Health , Aquatic Ecosystems , Environmental Statistics
Objective:
The overall goal of this research is to develop the fundamental capabilities needed as the basis for an integrated Temporal GIS/Visualization environment that facilitates science and decision support for assessment of environmental problems and potential solutions. Toward this overall goal, the objective of this research is to design and implement a prototype Temporal GIS with integrated multivariate spatiotemporal capabilities. The prototype will include: (1a) a fully-integrated database model which incorporates object-based, location-based and time-based representations of environmental data in a tightly-coupled manner, (1b) the use of parallelism for highly efficient storage, retrieval and manipulation of data, (2) a flexible database query protocol linked to a graphical user interface (GUI) which facilitates multi-dimensional and multi-scale query and analysis, (3) a visualization capability which facilitates exploratory and traditional analysis of time sequences of three-dimensional data, and (4) a data mining capability closely coupled with the visualization methods.Progress Summary:
We have just completed the second year of a three year effort. The timetable for the second year of the project called for the following steps to be accomplished or begun: Database issues (1a and 1b above); (a) performance testing and database model refinement, Query and interface issues (2 above); (a) continuing GUI development, (b) testing of GUI data model links, (c) testing of GUI GVis links, (d) integration of temporal GIS/GVis, Visualization (3 and 4 above); (a) coupling of visual representation methods with custom user interface controls. Progress has been made on each component. Toward attainment of our overall goal of providing an integrated environment for exploring large, complex space-time environmental data, we have also begun work on the development of Knowledge Discovery in Databases (e.g., data mining) techniques that will be integrated with the database and visualization capabilities. We focus below on the specific accomplishments in each of these areas.
Database issues:
Detailed performance statistics have been obtained
on the tree-based parallelization strategy that has been implemented on a
distributed network. Results of this performance testing has been documented in
[1]. The design of the database model has been extended from a triad of
representations into a more comprehensive and unified Pyramid design. This
Pyramid design incorporates observational and simulation data relating to time,
location and attributes of observed features, but also higher-level information
relating to characteristic traits and behaviors of the phenomena being observed.
This latter is derived information that represents stored knowledge that in turn
can be tested with further observation and simulation, and can be used to
facilitate database search.
Query and interface issues:
In the environment envisioned, the
graphic user interface is a single, integrated facility through which the user
controls operations applied to the database query, visualization display, and
analytical operations in a uniform manner. During the first eighteen months of
this project, GVis methods and tools were developed somewhat independently and
in parallel for application to a variety of environmental data sets in order to
evaluate various potential approaches. We have begun the process of integrating
these efforts into a single interface. The interface will allow domain experts
to take advantage of our innovative methods for both database and visual
representation, a powerful and flexible query protocol, complex supervised and
unsupervised classification techniques, and intuitive but powerful interface
controls designed to facilitate structured data exploration and analysis. At
this stage, we have designed and begun implementation of a preliminary visual
query interface and a corresponding set of 'interaction forms' that can be used
to control how retrieved information is visually represented and how
representations (visual and digital) are linked. The visual query interface used
a 'visual query composer' designed to allow various functions and data
representations to be graphically combined and complex analytical models
incrementally constructed, using the computer display as an interactive
'canvas.' The graphical interaction forms include 'brushing' and 'focusing,'
dynamic temporal legends, and a manipulable parallel coordinate plot that
facilitates multivariate and multiscale analysis. A key component of our
interface effort is development of a minimal set of operators for all database,
analysis/modeling, knowledge discovery and visualization functionalities. The
goal here is a robust set of operators that can easily incorporate future
developments and capabilities.
Visualization and data mining:
The major Apoala GVis extension over
the past year has been work to combine GVis and Knowledge Discovery in Database
(KDD) methods in the context of spatiotemporal environmental data analysis. GVis
methods and KDD facilitate the identification and interpretation of
spatiotemporal object classes. The environment uses a set of linked visual
representation forms that allow for simultaneous exploration of geographic and
attribute spaces (at multiple times and scales). The components of data
displayed in each representation form are controlled through multiple
interaction forms (discussed above). Exploratory analysis and visualization of
complex, multi-dimensional space-time data demands new and innovative tools.
Over the past three months, we have extended beyond our previous
workstation-based GVis methods to explore the potential of immersive and
collaborative visualization methods (using ImmersaDesk VR technology and high
speed network connections). Our tests thus far suggest that immersive display
environments offer tremendous potential to extend the power of visualization to
facilitate understanding of spatiotemporal environmental data (particularly when
linked to make collaborative visualization possible). The test, however, has
also revealed the difficulties involved in integrating these heretofore
independently-developed environments.
Other accomplishments:
Increasing the visibility of the Apoala
project in order to disseminate results and encourage feedback was a significant
element of our work during this phase of the project. A completely redesigned
web site was launched.
Future Activities:
In the third year of this effort, the extended Pyramid database design will be implemented and integrated into the Apoala system. We also plan to complete implementation of the integrated graphical user interface, and to develop an implementation of combined visualization/data mining tools in a parallel processing environment that allows real-time user response to visual displays generated to reflect the iterative data mining process. A focus will also be upon testing the integrated system in real-world problem environments, and documentation of our results.Journal Articles on this Report : 3 Displayed | Download in RIS Format
| Other project views: | All 59 publications | 12 publications in selected types | All 9 journal articles |
|---|
| Type | Citation | ||
|---|---|---|---|
|
|
Edsall RM, Harrower M, Mennis JL. Tools for visualizing properties of spatial and temporal periodicity in geographic data. Computers and Geosciences, 2000, Volume: 26, Number: 1 (FEB), Page: 109-118. |
R825195 (1998) |
not available |
|
|
Kraak MJ, Edsall RM, MacEachren AM. Kartographische animaties en legenda's voor temporele kaarten: exploratie en/of interactie. Kartografisch tijdschrift 1997;23(4):13-18. |
R825195 (1998) R825195 (Final) |
Exit |
|
|
MacEachren AM, Wachowicz M, Edsall R, Haug D, Masters R. Constructing knowledge from multivariate spatiotemporal data: integrating geographic visualization with knowledge discovery in database methods. International Journal of Geographical Information Science 1999;13(4):311-334. |
R825195 (1998) R825195 (Final) |
Exit Exit |
Supplemental Keywords:
Space-time GIS, space-time data, information interfaces, animation, parallel processing, very large databases, RFA, Ecosystem Protection/Environmental Exposure & Risk, computing technology, geospatial data, data sharing, environmental decision making, space-time environmental data, computer science, geographical information systems, integrated visualization, data management, TGIS/GVIS, information technology, multivariate spatiotemporal visualizationRelevant Websites:
http://www.geovista.psu.edu/apoala/index.htm
http://www.geog.psu.edu/tempest
http://www.geovista.psu.edu/index.html
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.