Geospatial Data

1
Intersecting Geospatial Information Information
Technology
Geographic Information Sciences
Geospatial Data
Information Technology
Marc P. Armstrong The University of Iowa
2
Intersecting Geospatial Information Information
Technology
Visualization HCI Group Work
Geospatial Data
Information Technology
Marc P. Armstrong The University of Iowa
3
GIS technologies and geospatial information
Play key roles in everyday life, and support
complex scientific investigations and informed
public and private sector decision making.
4
Limitations of GIS

• Despite advances during past two decades,
  significant limitations persist
• GIS software is ill-suited for use by groups
  task forces/committees/panels that are convened
  to address complex policy problems.
• GIS software is complex education lags.
• Two dimensional maps are inadequate for
  visualizing high-dimensional information and
  require users to have map reading skills (big
  assumption).

5
GIS software is designed for single users.
!
?
This can create a bottleneck effect, isolating
decision-makers and requiring the presence of an
expert user to elicit preferences from them.
Source Armstrong, 1998
6
A better approach enables decision-makers to work
both directly and cooperatively to search for
solutions to complex problems.

!
?
?

Source Armstrong, 1998
7
Interaction effect complexity is increased when
multiple stakeholders criteria are involved.
a.
b.
Fails public water supplies criterion
GIS analyses can be complex.
Public water supplies
Buffer (1mi)
Fails private water wells criterion
Private water wells (UTM)
Buffer (1000)
Private water wells
Project
Address Match
Private water wells database
Overlay
Areas that fail screening criteria
Select freq 100
Fails floodplain criterion
Floodplain
Soils
Fails land use criterion

Fails depth to water table criterion
Select water table lt 5ft
Soils w/ depth to water tab.
Join
Soils database
Overlay
Digitize
Fails distance to property line criterion
Buffer (50)
Parcels
Parcels
Fails resid. development criterion
Landuse (polygon)
Select LUResid.
Residential development
Landuse (raster)
Raster? vector
Buffer (500)
Source Bennett and Armstrong, 2001.
8
Large Conceptual Barriers to Use

• Education about geographic fundamentals (e.g.,
  scale) is needed
• Science and technology overlap (GIScience Info
  Sciences) can help erode barriers
• GIS user interfaces, visualization, group work
• Result will be improved ability to cope with
  geospatial data

9
The ideal user interface
10
The ideal user interface
Easy to understand and unambiguous.
11
What do we have instead?

• Complexity
• Original command line (good for careers, but
  akin to JCL (not Java Constraint Language) for
  newbies IEF640I EXCESSIVE NUMBER OF POSITIONAL
  PARAMETERS)
• Move toward GUI, objects drag drop for
  model creation
• New modeling and interface metaphors

12
Spatial metaphors in information search and
retrieval

• Topic Maps (www.topicmaps.org for specs)
• Graphical representations (www.inxight.com)
• Spatial info-navigation metaphors- (M.
  Czerwinski, Microsoft http//antarcti.ca/)

See Martin Dodges An Atlas of Cyberspaces
http//www.cybergeography.org/atlas/atlas.html
13
Spatial metaphors are the focus of several
e-business initiatives
Several examples are at http//www.map.net
14
Recursive Metaphor- using a virtual earth
representation to access geospatial information
http//www.ai.sri.com/digitalearth/
15
Task Specificity of Metaphors

• Previous examples focused on information access
  and retrieval.
• Much of GIS involves analysis of geographic
  information and different metaphors may apply.
• Geographers Desktop from U of Maine is one good
  example

16
Geographers Desktop Map Cabinet Metaphor
Source Max Egenhofer
17
Pictographic Representations of Map Algebra
Source Max Egenhofer, UMaine
18
But are we skating to the puck?

• Guy LaFluer, when asked why he was such a great
  hockey player, paused and then replied
• Other players skate to where the puck is. I
  skate to where the puck will be.

19
GIS, Viz, HCI, CSCW A Scenario

• A train derailment occurs in January on a track
  in rural Iowa that is situated along a bluff that
  overlooks the Mississippi River.
• Two tanker cars threatens to release a highly
  toxic substance.
• Local expertise in handling this type of
  emergency is absent.

20
Uh Oh

• Responders need expert advice to create of an
  emergency plan.
• They turn to a chemical engineer, hydrologist and
  an expert in topographic analysis.
• Digital terrain information (a DEM) is
  downloaded.
• Information about the toxic substance and
  atmospheric conditions are put into a dispersion
  model.

21
Teleimmersion Analysis

• The team uses the DEM and a 3D plume dispersion
  model in a teleimmersive environment.
• The dispersion model indicates that since it is
  January (-10C) the chemical is non-volatile and
  will flow as a liquid (like molasses in Jan).
• The DEM helps the team to visualize that the flow
  will be blocked from entering the Mississippi
  immediately because of a levee.

22
Far Fetched?

• Consider developments in technology and their
  effects on GIS use

23
Modified after MacEachren et al.
24
Single User Workstation
25
CSCW, digital conferencing
26
Teleimmersion
27
Immersion

• Economics of display technologies force 2D
  output.
• This limits insight and creates problems (e.g.,
  project 3D ? 2D).
• Immersive visualization enables users to
  experience realistic environmental
  representations and examine high-dimension
  information.
• Immersive technologies (e.g., ImmersaDesk CAVE)
  can be used by gt one person at a time. Think
  Holodeck

28
3D topographic model of Iowa River valley using
Viz5D running on an ImmersaDesk.
Source Wang and Armstrong, 2000.
29
Teleimmersion

• Synchronous use of immersive visualization by two
  or more people over a network (e.g., Internet2,
  MAGIC).
• Also called collaborative virtual environments
  (CVE).
• Research focus of NSF ITR, and NSF Distributed
  Terascale Facility.

Communication through Network
Sourcehttp//fakespacesystems.com
30
Teleimmersion a research challenge that unites
HCI, Visualization and CSCW.
It also underscores need for HPC/Grid for
near-real-time interaction.
Prototype teleimmersive grid cell GIS with a
slider bar user interface.
Source Wang and Armstrong, 2000.
31
Animation
Source Wang and Armstrong, 2000.
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
Representation of Decision-Makers
Virtual personae avatars. This is a crude
head arm(face and pointer) representation.
Source Wang and Armstrong, 2000.
36
Realism improves usability?
Source Wang and Armstrong, 2000.
37
Teleimmersion Research Issues

• What is an appropriate interface for
  teleimmersive GIS?
• What level of geographical realism is appropriate
  in a teleimmersive GIS? What level of realism is
  appropriate for avatars?
• How is group use handled (computer
  mediated/virtual aggression, dominance and
  submission) and how are mapped results
  synthesized (e.g., areas of agreement and
  disagreement)? What tools are needed? A new group
  map algebra?
• How do we support a simultaneous computational
  need to track users, render spaces and analyze
  data (using NP-hard models)?
• How do we address generic technical problems,
  such as latency and jitter, in data-rich
  teleimmersive GIS?

38
//SYSIN DD DSNAUDIENCE.QUESTIONS