Recent Changes to AMPS MM5 Configuration

Recent Changes to AMPS MM5 Configuration

2004-08-05/12: 5-day (120-hour) forecasts on the 90/30 grids. The two extra days adds roughly an hour to the processing time.
2004-05-27/12: 3D Variational Data Assimilation (3DVAR) is now the default analysis technique for AMPS MM5 initialization.
2003-09-19/12: Added a 10-km grid over the Antarctic Peninsula. This adds about an hour to the total time it takes to run AMPS MM5, and brings us to a total of 6 two-way interacting grids.
2003-07-30: We no longer wait for a full set of GFS files out to 72 hours (for boundary conditions) before starting AMPS MM5 preprocessing. We now start preprocessing GFS files as they arrive, and when we have processed out to 18 hours, there is enough information to start MM5. We continue to process further GFS files as they come in. This allows us to start an hour or more earlier than we would have.
2003-05-20/12:
  • Implemented a top boundary condition devised by OSU/BPRC researchers. It nudges T and PP in the top eight model layers toward AVN forecast fields.

  • Extended the model top from 100 mb to 50 mb, and added two sigma levels to accommodate the new top boundary condition. Repositioned all sigma levels to account for higher lid and new levels.
  • 2002-12-04/12: Shut off cumulus parameterization on grid 5 (3.3 Ross-Island grid). Prior to this forecast, the Grell cumulus parameterization had been active on grid 5.
    2002-11-04/12: Grid 2, the 30-km grid, has been expanded, so that the boundaries of the grid could be a little farther away from the Antarctic coast. The current dimensions (X x Y) of grid 2 are 220 x 208. Prior to this forecast, the dimensions had been 178 x 187.
    2002-07-03/00: Grid 1, the 90-km grid, has been expanded to include the ocean between Antarctica and South Africa. The current dimensions (X x Y) of grid 1 are 110 x 145. Prior to this forecast, the dimensions had been 110 x 115.
    2001-12-13/00: Grid 5, the 3.3-km Ross Island grid, has been implemented for hours 06-36.
    2001-11-28/00: Grid 4, the 10-km South Pole grid, has been implemented for hours 06-36.
    2001-10-31/00: Grids 1 and 2 have been extended from 48 to 72 hours. Grid 3, previously running from 06 to 30 hours, has been extended to hour 36.
    2001-09-07/00: Fractional sea-ice coverage implemented, from a satellite-derived sea-ice data set. Prior to this forecast, sea ice as a binary (ice/no-ice) field was specified as a function of sea-surface temperature.


    Recent Changes to AMPS Products

    2004-08-05/12: All products from the 90/30 grids are now available out to 120 hours.
    2003-09-15/12: Added a Pole to Rothera Cross section.
    2003-09-01: Additional time series sites.
    2003-08-19/12: Textual time-series reports, similar to FOUS reports, are now posted.
    2003-07-22/12: Added a time series for Ice Runway.
    2003-07-17/12: Added time series for Pegasus South and Williams Field.
    2003-07-15/12: We now post products as the model progresses. Previously we waited until the model had completed before posting graphics. Now, the early hours are posted early on, and later forecast hours are posted as the model integrates over the course of about 4 hours.
    2003-07-09/12: Updated some hard-wired constants in the time series plotting program, which had been left at old values. This results in pressure traces on the order of 15 mb lower.
    2003-07-02/12: Added a time series and sounding at Halley, at the request of BAS researcher.
    2003-06-05: Added a time series at Carvajal.
    2003-06-04: Added two new sounding sites, Eduardo Frei and Carvajal, to 30-km window and GLOBEC window, at the request of Chilean forecasters.
    2003-03-06: We've made the final shift. The AMPS page has been updated to reflect new directory structures. Should be mostly invisible to users.
    2003-03-03: Additional time series in 30-km grid (and GLOBEC window) of station Eduardo Frei, at the request of Chilean forecasters.
    2002-11
    through
    2002-12:
    In response to the Workshop on Antarctic Numerical Weather Forecasting for Operations, we've implemented:

  • Additional time series at Neumayer, Kohnen, Rothera, Sky Blu, San Martin, Novolazarevskaya, Sanae, Troll, Syowa, Casey, Davis, Mawson.

  • Additional soundings at Neumayer, Rothera, Novolazarevskaya, Troll, Syowa, Sanae, Casey.

  • Additional plotting window of the 30-km grid for the Queen Maud Land coast.

  • Cross-section from Neumayer to Kohnen.

  • Moved the GLOBEC plotting window more over the Peninsula (possible because of the expanded 30-km grid).
  • 2002-10-22: Additional plotting window of the 30-km grid for the Pine Island Bay.
    2002-10-22: Additional plotting window of the 30-km grid for the Marie Byrd Land area.

    e-mail: Kevin Manning
    Last modified: Wed Oct 08 14:15:00 MDT 2003