Global Parallel Experiment Log

Experiment Name

GSI hybrid real-time

Parallel Slot

PRX

Control Slot

GFS

Start date of parallel experiment

2007/03/01/00

End date of parallel experiment

ongoing

Implementation date if any

May 1, 2007

Environmental Modeling Center scientists

whole GlobMob

Abstract (including Motivation, Hypothesis and Method)

Analysis changes (Russ Treadon, Daryl Kleist, John Derber)
--Replace Spectral Statistical Interpolation (SSI) analysis system with Gridpoint Statistical Interpolation (GSI) analysis system.
--Use of increased observations including:
----Full resolution AIRS (every field of view form of AIRS rather than only center spot AIRS (used in oeprational GFS)
----COSMIC GPSRO
----A problem in COSMIC GPSRO was corrected on March 5, 2007.
Model changes
--Modify sigma coordinate system to a hybrid sigma-pressure coordinate system
--Updated GFS radiation module (Yu-Tai Hou)
----overall restructuring of radiation related programs to help future development and upgrade
----In RRTM1-LW, minor upgrade of emissivity coefficient. Rare gas absorption effect turned on
----In aerosol calculation, vertical sigma based structure changed to pressure based structure
----Minor bug correction in cloud related calculation
----Performance of new radiation very close to operational model
Product changes
--Add 30 records to surface flux fields
--Change output of GDAS from 3-hourly to hourly
--Model to native output file will change from a spectral binary file to a hybrid spectral binary file
Questions? Contact Glenn.White@noaa.gov or Mark.Iredell@noaa.gov

Experiment changes log

Implementation changes list

Product changes list

Background links

Evaluation of parallel results

Skill evaluation of real-time parallel Apr 12 by Glenn White

Skill evaluation of real-time parallel Mar 29 by Glenn White

Daily forecast maps by FangLin Yang

Daily forecast maps of 500 hPa height for many models by FangLin Yang

Daily forecast statistics by Pete Caplan

Daily forecast statistics by FangLin Yang

Verification against sondes and aircraft by Suru Saha

GFS PRX case study April 1 by Richard Pasch

Biases of operational and parallel GDAS and Guess against observations by Brad Ballish

Verification of precipitation over continental United States 000UT by Y. Zhu

Verification of precipitation over continental United States 1200UT by Y. Zhu

Clouds and radiation verification and comparison for March 2007 by Ken Campana

Clouds and radiation verification and comparison for Feb 2007 by Ken Campana

Conclusion

parallel skill in extratropics very close to operational
parallel slightly more skill in precipitation forecasts over continental United States
parallel improves tropical wind forecasts below 300 hPa, worse wind forecasts above 300 hPa
parallel much less negative bias in Asian jet
parallel short-range forecasts less rms error against analyses
parallel short-range height forecasts significantly closer to radiosondes
parallel winds closer to radiosondes
parallel forecast winds fit ACARS, aircraft winds better between 200 and 150 hPa
parallel extratropical temperatures closer to radiosondes
parallel analysed moisture closer to radiosondes

Detailed comments

Total cloud slightly less in the parallel and may be due to less stratus.
High cloud is a bit more in the parallel.
Radiation changes driven by cloud changes: day 1, SW absorption larger in parallel
which may help force warmer land surface in parallel (sfc LW up in parallel
Cloud correlation and bias differences are small, but low cloud in parallel
seems to have less correlation with USAF clouds and has negative bias.
These are small differences. Day 1 differences seem larger than day 5 differences. (Ken Campana)

Global Parallel Home Page

Disclaimer