NOTE: The nws-changes mailing list is no longer active. The list archives are made available for historical reasons.
This is a preliminary announcement from the Environmental Modeling
Center of a change we propose to make in the AVN/MRF forecast system
(the global forecast system, gfs) in April/May 2002. We are starting a
coordinated evaluation period after preliminary evaluation by our own
scientists. The change we are planning to make involves an increase in
the model resolution. We will increase the model grid size from 75 km to
55 km and increase the number of vertical levels in the model from 42 to
64. A brief description of the change is appended below. Please feel
free to contact us for information and your evaluation. The primary
evaluators will be the forecasters from the NCEP centers (HPC, MPC, AWC,
TPC, and SPC). However, some forecast maps will be available on the EMC
websites for all users to evaluate. We will send a final notice when the
evaluation has been completed and we are ready for implementation.
sincerely,
Hua-Lu Pan
Chief, Global modeling branch, EMC
Notice of Intent to Change the NCEP Global Forecast System
Brief description of change(s):
We plan to increase the resolution of the Global Forecast System
(GFS) to T254L64 for the Global Data Assimilation System (GDAS) and the
MRF model (MRF/AVN forecasts). The first 84 hours of the forecast for
the MRF/AVN will be at the high resolution (approximately 55 km grid
size). We plan to implement a cascade strategy for the resolution
during the rest of the forecast. From 84 hours to 180 hours, we will
run in T170L42 resolution (75 km grid size), which is the current
MRF/AVN resolution. From 180 hours to 384 hours the model resolution
will be T126L28 (105 km grid size). Currently, we use T170L42 for the
0-168 hour forecast and T62L28 for the 168-384 hour forecast. Ensemble
forecasts will be made using a T126 resolution out to forecast hour
180. Currently the ensemble forecasts use T126 out to 84 hours.
Reason for Change(s):
The resolution increase is part of the changes planned for the
computer upgrades that occurred in 2001. With higher resolution, we can
create model initial conditions that are closer to the actual
observations. In addition, many atmospheric features such as jet streaks
and tropical cyclones require very high resolution to be properly
simulated. This is a step toward better modeling of the atmosphere. At
longer forecast hours, model accuracy in the small scale phenomena
quickly decreases due to the chaotic nature of the atmosphere. The
cascade strategy reflects such reality. The higher resolution beyond
180 hours of the forecast will improve the week two threat assessment
via improved modeled physical processes.
Schedule for change:
Final testing completed: to be completed March 2002
CAFTI Approval: will seek approval March 2002
Expected implementation: planned implementation April 2002
Description of testing:
Statistical summaries:
From http://www.emc.ncep.noaa.gov, choose AVN model performance
icon to getto the GFS evaluation pages.
Forecast examples:
From the same web site, select the experiment and parallel run
icon to get to the maps.
Anticipated impact on forecasts:
Since the change is strictly one of resolution, we anticipate only
incremental improvements in the forecast skills. Model biases and
characteristics will remain largely unchanged. The highest resolution
forecast in the first 84 hours is expected to improve model forecast of
the details of fronts, jet streams, QPF, and tropical storms. The 3-7
day forecast will benefit from having both a better initial condition
and a better day 3 forecast due to the highest resolution. The change of
the week two resolution from T62 to T126 is aimed at getting a better
forecast of transients in the longer forecast range in order to improve
predictions of regime changes.
Further documentation available:
From the emc web site (given above), select Parallel MRF runs
home page for more information of the changes.
Human point of contact for further information:
Send e-mail to peter.caplan@xxxxxxxx
Field Evaluation:
This is the start of the field evaluation phase.
nws-changes archives: