"Application of Morphological Representation to Enhance the Visualization of Dynamic Weather Maps" - Presentation at the 2007 American Association of Geographers Conference.
Application of Morphological Representation to Enhance the Visualization of Dynamic Weather Maps
1. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Topological Modelling
of Atmospheric
Surfaces- A Concept
Sanjay Rana
2. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Contents
• Visualisation and Modelling of
Atmospheric Surfaces
• Topological Modelling of Atmospheric
Surfaces
• Progress so far
• Future Work
3. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces
Two main implementations of Atmospheric
Surfaces:
• In visualisations for general public and
meteorologist for making Weather Charts etc.
• In Forecasting Models using fairly sophisticated
numerical models involving several atmospheric
measures and supercomputers!
4. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces > Visualisation #1
Common visualisation of
weather maps produced
by experts based on an
amalgamation of
temperature, pressure
and wind velocity
surfaces.
www.bom.gov.au/nmoc/MSL/
WeatherMap.shtml (9/Apr/07)
5. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces > Visualisation #2
Sequence of shaded
contour maps showing:
•Time Slices
•Difference Maps to
show the overall
variations in a period
http://weather.unisys.com/surface
(9/Apr/07)
6. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Sequence of shaded contour maps showing temperature
over North America at every hour over 12 hours period
last week
http://weather.unisys.com/surface
(9/Apr/07)
7. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Sequence of shaded contour maps showing
change in temperature over North America every
hour over past 24 hours last week
http://weather.unisys.com/surface
(9/Apr/07)
8. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces > Visualisation #3
Combination of
exploratory visualisation
techniques (linked
views, brushing,
interactivity),
perspective views and
information
visualisation.
http://www.vrvis.at/simvis/Isabel/
(10/Apr/07)
9. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces > Modelling
Evolution of the global
temperature over a
month as simulated
within the Flexible
Modeling Systems (FMS)
framework.
Requires the use of a
CRAY!
http://www.gfdl.noaa.gov/~fms
(10/Apr/07)
10. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Visualisation and Modelling of
Atmospheric Surfaces > Insights
• Visualising the features of the morphological structure
(i.e. lows, highs, saddles, weather fronts) of
atmospheric surfaces is fundamental for weather
interpretations,
• Recording the Spatio-Temporal variations in the
morphological structure of atmospheric surfaces leads
to the development of forecast models,
• However, at present there are no representations that
create a network of these features i.e. no topological
data structures are used to model the surface.
11. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Topological Modelling of Atmospheric
Surfaces > What’s the topology?
• All atmospheric surfaces have morphological features
such as lows (pits), highs (peaks), ridges, channels as
found in terrain and many other surfaces.
• Earlier studies in other kinds of surfaces e.g. terrain
density, socio-economical, medical imaging, fluid flows
etc. have shown that topological representations based
on these morphological features provide a global
representation of the morphological structure.
• This presentation demonstrates some experiments to
model the topology of atmospheric surfaces.
12. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Topological Modelling of Atmospheric
Surfaces > What’s the topology?
• Morphological Features
Peak Pit Saddle Monkey
Saddle
Channel Ridge
13. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Topological Modelling of Atmospheric
Surfaces > Topological Data Structures
• An Example - Surface Network
x (1000)
3000 z1
y1
z1(5000) z2(4000)
2000
1000
y1(2000)
z2
y2(3000)
x 2000
x (1000) 2000
x : Pit
y : Pass y2 2000
z : Peak z3(5000) z3
x (1000)
(a) (b)
(a) A contour representation of an island showing morphological features (elevations
given in parentheses), channels (dashed arrows), ridges (solid arrows); (b)
corresponding Surface Network Graph. Numbers on the edges of the graph in (b) are
the differences in elevation across the edges.
14. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Topological Modelling of Atmospheric
Surfaces > Morphological Features in
Atmospheric Surfaces
Pits
Channels Saddle
Ridge
Peak
15. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Progress so far #1>Visualisation of an
animated sequence of Atmospheric Surfaces
16. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Progress so far #2> Modelling the Spatio-
Temporal variations in Atmospheric Surfaces #1
time
events
length, length direction Merging
direction
length, Scale Space of the surface
bifurcation temperature ravine
17. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Progress so far #3> Simulations of
hypothetical weather scenarios #1
• The structure of a Surface Network is a natural
representation of the morphology e.g. unlike
TIN or Raster
• Topologically-consistent manipulation of a
Surface Network structure will thus yield
theoretically valid alternative states of
atmospheric surfaces
18. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Progress so far #3> Simulations of
hypothetical weather scenarios #2
Simplification Extensification
y1 zo
xo z x xo
x zo
yo
yo
y’
xo z z’ xo
x
x
yo
x – Pit; y – Pass; z - Pit
19. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Simplification Extensification
Extensification
Refinements on longest ridges (indicated by red arrows) of a
Surface Network; blue lines- channels, orange lines- ridges,
Generating artificial changes by simplification,
red discs - peaks, green discs – passes, and black discs - pits.
in this case a large valley
20. Session 1551
2nd Emerging Computational Techniques
and Technologies in GeoComputation
Future Work
• Extend to Volumetric atmospheric surfaces
• Implement Surface Network/Morse-Smale
Complex
• Further information on Topological Data
Structures for Surfaces in Rana (2005).
• Seeking collaborations from researchers in
• Computer Vision
• Meteorology
• Weather Data Providers
• GeoVisualization