Global Parallel Experiment Log
- Experiment Name
- T170 L42 MRF Forecast
- Parallel Slot
- PRX
- Control Slot
- MRF
- Start date of parallel experiment
- 1999/06/30/00
- End date of parallel experiment
- 2000/01/24/12
- Global Modeling Branch scientists
- Mark Iredell
- Abstract (including Motivation, Hypothesis and Method)
-
The T170 L42 "Final" Global Data Assimilation System (FNL GDAS)
and daily Medium Range Forecast (MRF) are run on the new IBM SP.
The data assimilation system started on July 22; until then the T170 L42
forecast was initialized from the operational T126 L28 analyses.
The T170 L42 analysis fits the observations much better than the 1998
implementation; this analysis is essentially the same version as the
T170 L42 parallel tests run on the Cray in late 1998 and early 1999.
The T170 L42 model is the same model that was run at this resolution
in 1998 (and in parallel in early 1999).
A digital filter initialization half-period of 3 hours is used.
The forecast remains at T170 L42 resolution for 7 days.
Then the 6.5 day forecast is truncated to T62 L28 and run out to 7 days
again with a 3-hour digital filter initialization half-period.
This new 7-day forecast is used to run the MRF extension at T62 L28
out to 16 days. This procedure at day 7 is designed to reduce the
convective precipitation spinup noted when the resolution is reduced.
This is the first global parallel on the new IBM SP.
- Itemized source or script changes
- No scientific changes to the model, but
- All new source and scripts are used on the IBM SP.
- Itemized product or file changes
- The order of the fields in pressure GRIB file has been shifted slightly.
- New fields in the pressure GRIB file
- CAUTION:
New forecast time units used in the pressure GRIB file
- Sample inventory for pressure GRIB file
- Sample inventory for pressure GRIB file for forecast hour 264
- All fixed files are copied to or regenerated on the IBM SP.
- Experiment changes log
- Evaluation of parallel results
-
Daily forecast maps by Pete Caplan
-
D+8 and week 2 maps by Jae Schemm (CPC)
-
D+8 and week 2 scores by Jae Schemm (CPC)
-
Anomaly correlation score comparisons by Yuejian Zhu
-
February 1999 T170 vs T126 fits to obs by Suranjana Saha
-
February 1999 T170 vs T126 fits to obs by Yuejian Zhu
-
February 1999 T170 vs T126 model diagnostics by Glenn White
-
August 1998 T170 vs T126 fits to obs by Suranjana Saha
-
August 1998 T170 vs T126 fits to obs by Yuejian Zhu
-
August 1998 T170 vs T126 model diagnostics by Glenn White
-
August 1998 T170 vs T126 more diagnostics by Glenn White
- Conclusion
-
Draft Technical Procedures Bulletin (TPB) by Caplan and Pan
- Operational implementation on 2000/01/24/12.
Global Parallel Home Page
Disclaimer
Page Last Modified: January 24, 2000