Ken Pollak
1. Any observational data sources that have potential use in data assimilation schemes must be routinely available in a timely fashion for operational use. (i.e., ice thickness, velocity, drift, etc.)
2. PIPS3.0 software needs to comply with FNMOC software standards. This is moslty a 6.4 issue, but if software guidelines are followed in the 6.2 area, this could ease the transition to 6.4.
3. PIPS3.0 must have a "restart" and "catch-up" capability to maintain model history in case of missed runs, operational computer problems, etc.
4. Surface wind stresses and heat fluxes from NOGAPS (vice from some other model, such as the NCEP or ECMWF global NWP models) must become the atmospheric forcing of choice early on and throughout the PIPS 3.0 development process. This may lead to identification of problems with the atmospheric forcing and/or the ice model formulation that might not otherwise become apparent until PIPS 3.0 is delivered for operational testing at FNMOC.
5. Strongly encourage modularity of software,and ability of PIPS3.0 to couple to different ocean models (MOM, POP, NRL's NCOM, etc.) and with atmospheric models such as FNMOC's NOGAPS and COAMPS.
6. The old paradigm of configuring a model for specific areas with "hard-wired" data bases is no longer acceptable. Would like to configure PIPS3.0 to run in different regions with ability to couple (one and/or two ways) with different ocean and atmospheric models. This will fit into FNMOC's concept of a "plug and play" approach to a unified air/ocean model suite with compatible model components. Possible configurations are:
- 1. A globally coupled NOGAPS/Wave/PIPS3/Ocean model.
- 2. A COAMPS/Wave/PIPS3/Ocean model
- 3. Arctic/Antarctic PIPS3.0/Ocean model driven with interpolated NOGAPS fluxes