APL%20MURI%20Kickoff

1
APL MURI Kickoff
2
APL MURI TEAM

• Bob Miyamoto
• David Jones
• Jim Pitton
• Keith Kerr
• Mark Krueger

3
Key Strengths of the Team

• Navy/domain experience
• Related research projects
• MHSII
• EVIS
• DRI
• Visualization, Intelligent Agents, Engineering
  Statistics

4
APL Role in MURI Team

• Provide relevance to Navy needs
• Coordinate interaction w/ Navy Orgs.
• Develop task-based visualizations
• Integrate component research efforts
• Facilitate interaction among MURI participants

5
Navy METOC Uncertainty

• Model of the Day?
• Tropical Cyclone forecasts
• EFS at FNMOC
• Ship Routing
• Long Range Temp Forecast

6
Navy METOC Uncertainty

• Based on my experience as the Operations Officer
  at FNMOC
• Forecasters have limited understanding
• Confusing products
• Not designed for forecaster
• Great potential still not fully realized

7
Examples of Navy Uncertainty Products
8
FNMOC Ensemble Temp Mean
9
FNMOC 2m Temp Plume
10
FNMOC Gale Probability
11
Areas to investigate

• How forecasters deal with uncertainty
• Uses of uncertainty information by METOC
  customers
• Easier ways to create uncertainty products
• Better visualization techniques
• Verification

12
FNMOC COAMPS
13
UW Ensemble Mean
14
Next Steps

• Cognitive Task Analysis at
• Whidbey Island
• Norfolk
• Collaborative work with NPMOC, NLMOC, NPMOF,
  FNMOC
• Visualization research gtgtgt

15
Exploratory Software Prototype

• Requirements
• Design
• Prototype Development
• Iterative Refinement
• Implementation Analysis

16
Tentative Requirements

• User-based framework
• Analytic Geospatial visualization tools
• Collaborative, interactive exploration
• Cross-platform availability
• Easy extensibility
• Suitable for broad range of expertise

17
Implementation Tools (1)

• Java Language
• Cross-platform
• Sophisticated network (web) model
• Can wrap models in other languages
• Inference Engine
• Easy to tailor level of expertise
• Can both bound and sequence operations

18
Implementation Tools (2)

• VisAD for Visualization
• Model-View-Controller architecture
• Remote display and collaboration
• Sophisticated data model
• High-level scripting language (Jython)
• Supports specialized toolkit development
• Currently used within meteorological community

19
VisAD Creations (courtesy VMET)Terrain and Wind
Vectors

20
VisAD Creations (Courtesy VMET) Time Series
Wind Fields
21
VisAD Creations (courtesy VMET)MM5 output wind
and temp fields