1. June 18, 2003
Sachin Bhate (PI), Jim Corbin, John Van Der Zwaag
Yee Lau and Surya N Renduchintala
Sponsor: Office of Naval Research
2. “Oceanographers are there, researchers are there, there
are models running in every part of ocean, there is a
buoy floating somewhere, there is some data collected
possibly every minute, and there is a weatherman out
somewhere watching for us. It is very important for us
to keep our eyes and ears open, keep a clear mind and
think, how we can build new bridges between these
modelers and data collectors, improve the
communication between researchers and weathermen,
and create a wave of virtual collaboration in
oceanographic and atmospheric community.”
- Sachin Kumar Bhate
3. Motivation behind DA_MAP
Create an application for a METOC data analyst/modeler
that allows more time to be spent on analysis and less on
finding data and figuring out how to use it. It should:
• Integrate data access interactively with
visualization and analysis
• Provide transparent access to diverse data
• Exploit state-of-the-art software technology
• Be easily extensible and platform independent
• Exploit utility of a desktop/laptop computer
• Be intuitive and user friendly
4. Emphasis areas
• NAVOCEANO: Our initial user
• METOC Data Analysis and Model Assessment.
• Ocean model output and in-situ observations
• Data Registration and Metadata
• SearchNQuery mechanism (Correlation)
• Interactive visualization/analysis environment
5. Our Technical Approach
• Java
– Our main driver for application
– Graphical User Interface
– Visualization (Java2D)
– NetCDF Java Interface (Unidata)
• JWS (Java Web Start)
– An application deployment web technology, gives the
user the power to launch full-featured applications
with a click from a standard web browser
– Just an HTTP server is enough
– Facilitates easy installation and updates for
applications
– Launch DA_MAP on any desktop from anywhere.
6. More…
• XML (eXtensible Markup Language)
– Metadata representation and data registration
– Search and query operation
– Easy to understand and implement
– Less maintenance
– Better future prospects
JWS XML
JAVA
7. DA_MAP Architecture
Front End
(JWS, Java, XML, XPath)
Registration Search&Query Data Analysis
Middle Tier
Metadata
HTTP server, JWS,
(XML)
XML
Back End
Local Remote
Data Data
8. Data Flow between different modules of DA_MAP
SearchNQuery Data Registration/Metadata Creation
Access Store
Metadata Metadata
Provide Data (XML records)
Location
Register Data
Metadata
(Create metadata &
Repository
Data Dictionary)
(XML metadata,
Data dictionary)
Access Metadata
and Data
Dictionary
Import Data
Local/Remote
Data Analysis
Data
9. DA MAP Portal
Integrated environment with
access to:
• Setting workspace
• METOC data registration
• Metadata creation
• METOC Data Search
• METOC Data Analysis
• Interactive Analysis
10. Data Registration Module
• Metadata easily read from self describing NetCDF data files
• User interface for mapping variable names (Data Dictionary)
• Metadata stored in XML format.
• Define datatypes to create mappings for a class of
METOC data variables and dimensions to automate
metadata creation for large number of data files.
• Next…..
Closer look at the metadata records…
14. SearchNQuery Module
• XML Query Engine
• Search on small metadata files rather than large primary
data files.
• Data search based on temporal and spatial domain
• Search based on models and observations
• Keyword search
• Save user-defined region
• Query result display in temporal and spatial domain
• Save query results and export directly to METOC Data
Analysis module
16. METOC Data Analysis and Assessment
• Floating interface
• Easy mixing and matching of METOC datasets.
• Interface and functionalities
– Choice of Horizontal and Vertical view
– Choice of base plot.
– Palette associated with each plot.
– Provide new extent for view
– Model data display (SWAFS, NCOM, etc.)
• Colorfill, contours, vectors, Streaklines
– Observation data display (ADCP, drifters, Buoy etc.)
• Horizontal & vertical plots.
– Overlay of model versus model data and
model versus observation
17. More…
• Model time series
• Sigma to Z-level conversion
• Vertical Stacking of time series plots
• Interactivity within the display
• Import data directly into analysis module
• Difference Maps, Color legend
• Editors for various functions
19. METOC Data Header information Quick View
Visualization View
Selection
20. METOC
Data
Analysis
Display
Window
Depicting
Various
Features
Data Source:
NAVOCEANO
(Public Release
Approved)
21. Model Contours, Colorfills & Currents (SWAFS)
Editors associated
with these views
• Scalar Editor
• Vector Editor
• Palette Editor
• Grid Editor
Model Currents
Sea Temperature
Streak lines
Data Source:
NAVOCEANO (Public Release Approved)
22. Model Contours, Colorfills & Currents (NCOM)
Sea Temperature Sea Temperature & Model currents
Editors associated with these views
Data Source:
• Scalar & Vector Editor Florida State University
• Palette Editor (COAPS)
23. Observations (Shipboard ADCP)
Horizontal Plots
ADCP ‘U’ Component at certain time & depth ADCP ‘U&V’ Component at certain time & depth
Editors associated with these views
• Horizontal View Editor (Scalar & Vector)
• Palette Editor
• Grid Editor
Note: ADCP data synthesized. Tracks &
values associated not true.
24. Observations (Shipboard ADCP)
Time Series (U Component)
Time Series Stack (U, V & W Component)
Stick Plots (U&V Component)
Editors associated
with these views
• Vertical View
Editor (Scalar &
Vector)
• Palette Editor
• Axes Editor
Note: ADCP data synthesized. Time & values associated not
25. Overlays Model & Observations
Stick Plots: Horizontal Plots:
Model current v/s ADCP current Model ‘U’ v/s ADCP ‘U’ Component
Note: ADCP & Model data synthesized. Tracks &
model data values associated not true.
26. Interactive Polygon over METOC Data Display to find
Correlated METOC Data
User can create a polygon over an area of
interest on any DA_MAP display to find
METOC data (model or observations) correlated
in time and space.
Results on next page…..
Note: ADCP data synthesized. Tracks &
values associated not true.
27. Display Correlated Data (overlaid)
Time Series Stack Horizontal overlay
(Observation & model Current) (Observation & model Current)
Note: ADCP & Model data synthesized. Tracks &
model data values associated not true.
