Ocean data management, access, and analysis

1
Ocean data management, access, and analysis
2

• Data Management
• Ocean data is managed in combination with
  meteorological data, by a staff educated in the
  same fields
• Each dataset is maintained in a standardized
  system
• Metadata, identification of files, software, etc
  are secure under a change control system
• Each dataset has a web page that is automatically
  generated, and is always current.
• User support
• Staff specialist for each dataset
• User guides to different categories of data
• Of course, web search tools

3
Example of ocean data users guide
4

• Data Access
• NCAR facility users
• All data are on the SCD MSS and directly
  accessible to all computing platforms
• File location, metadata, documentation, software,
  and some data are available online through a web
  interface
• Remote users
• Data discovery is through a web interface
• Data packages provided upon request
• Staged to disk for network transfer, e.g. FTP
• A variety of magnetic tapes and CD-ROMS
• All users
• Receive consulting from staff specialist as needed

5
Example of remote access to ocean model data

• MICOM, Miami Isopycnic Coordinate Ocean Model
• Web order interface
• Automatic processes to stage the data from the
  MSS and create variable, temporal, and spatial
  subsets
• Data selected is staged to a server for FTP pickup

6
Six year mean 5m T
7

• Data Analysis
• Large Unix systems directly connected to the MSS
  and dedicated to data analysis, e.g. dataproc
• Software, NCAR Command Language with NCAR
  Graphics
• Ingests many standard formats (netCDF, HDF, GrIB)
  and any others by user design
• Many built in computation features, e.g.
  interpolations, divergences, vorticity, etc
• Many grid transformations, e.g. gaussian to
  lat./lon, etc
• FEATURE We control the design and do so to
  support the scientific needs
• Many software libraries to support optimized
  computations
• Top-of-the-line visualization facility and
  support staff

8

• Future Access and Analysis
• Community Data Portal
• Purpose
• Web-based access for MSS collections
• Data sharing for multi-institution cooperative
  studies
• Components
• User interface, Live Access Server (LAS)
• Middleware, Ferret, NCL, GrADS
• File service, local, or DODS
• Status
• Working on pilot project datasets and testing the
  middleware

LAS and Ferret developed by NOAA GrADS developed
by Center for Ocean-Land-Atmosphere Studies
(COLA) DODS (Distributed Oceanographic Data
System) developed by URI
9

• CDP, continued
• To be determined
• Will these systems scale to meet the requirements
  of large research datasets? Response,
  computational analysis, and data delivery must be
  satisfactory for the scientist!
• Interface and Reanalysis 2 example

10
Interface and Reanalysis 2 exampleSea Level
Pressure