August 28, 2007

Good afternoon Steve and Bob,

Mike Prince suggested that I forward you the following survey that I sent to RVTEC today. As a member of the newly formed Data Best Practices committee, I thought you may be interested in this RVTEC activity regarding metadata automation.

As you may be aware, I am the chairman of the Shipboard Automated Meteorological and Oceanographic System (SAMOS) initiative (http://samos.coaps.fsu.edu). One of the critical issues for SAMOS is the reliable transfer of metadata updates when instruments are swapped, fail, or are relocated. To date this continues to be a problem both for SAMOS and our data users. Note that currently 3 UNOLS vessel participate in SAMOS and we hope to recruit more next year

Just a bit of background. I look forward to working with the Data Best Practices committee.

Best regards,

Shawn

Dear RVTEC members,

In preparation for the upcoming RVTEC annual meeting, I have been asked by Bill Martin to conduct a survey regarding current practices for metadata associated with underway marine measurements in the UNOLS fleet. The focus of my interest and the subject of my presentation at RVTEC will be "automation of metadata tracking and dissemination from vessels to data users." I suspect that each vessel operator and/or institution has a current means of tracking metadata (either in paper or digital form). What I am concerned about is how do we get this information to users in a reliable way? Can the process be automated in some fashion?

When I talk about metadata, I am primarily thinking of two main groupings, what I call vessel and instrument metadata.

Vessel metadata includes: ship name, call sign, IMO number, vessel dimensions, home institution, technician names, contact addresses, photographs and schematics, etc.

Instrument metadata includes: sensor make and model, parameter name, parameter units, sensor location (height, depth, port, starboard, etc), sampling rate, data precision, calibration dates, etc.

NOTE: My direct experience is with meteorological instrumentation, but this discussion should include other underway measurements (TSG, bathymetry, ADCP, etc.).

So at this point, I am seeking input on this topic to (1) help me better understand current practices within UNOLS and (2) learn about new ideas for automation. To that end, I would appreciate your responses to the following questions by 30 September 2007:

1. How are metadata currently stored for your vessels? Are they hard copy or digital? Do you use text files or some sort of database?

2. What is your current procedure for updating your metadata records? What type of events result in metadata updates (e.g., sensor swap, new calibration)?

3. Are metadata records maintained only on the vessel or are they also stored on shore (e.g., at the vessel home institution)?

3. What mechanisms are being used to transfer metadata from ship to shore? Are these manual transfers or some form of electronic transfer? Do the transfers occur at regular intervals (daily, weekly, end of cruise)?

4. How are metadata disseminated to your data users? Are they burned to CDs given to the science parties? Are they available on-line in any form?

5. Could routine metadata reports be sent from each vessel to interested users on a regular schedule?  (for example, at the start of each cruise).

6. If so, what type of electronic format would you suggest? (I am aware primarily of ascii text and XML formats at present). Also please provide any thoughts on a transfer protocol (e.g., email, ftp, etc.)

7. What type of  software applications are needed to make metadata management easier for techs during cruises? (For example, SAMOS is developing a web interface to our ship profile database that will allow techs to update their SAMOS-specific metadata either dockside or while at sea).

8. Any additional comments on the topic would be welcome.

I appreciate your taking the time to respond to this survey. I look forward to reporting the survey results and having further discussions on the topic at the upcoming RVTEC meeting.

Thank you,

Shawn
----------------------------------------------------------------
Mr. Shawn R. Smith
Research Meteorologist
Director, R/V Data and FSU Winds Center
Center for Ocean Atmospheric Prediction Studies
Florida State University        | (850) 644-6918 (voice)
Tallahassee, FL 32306-2840     USA    | (850) 644-4841 (fax)
| Email: smith@coaps.fsu.edu