Report on THORPEX workshop November 14-15,
2002
by Sim Aberson (HRD)
The THORPEX workshop began with welcomes from Mel Shapiro and Rolf
Langland, and introductions by all the participants. Mel Shapiro and
Michel Beland then gave the perspectives of the International Science
and Core Steering Committees, respectively, and Mel gave an overview
of THORPEX so far. This is all background information that is also
available at the
THORPEX website, so I will not go into these presentations here.
Alan Thorpe then gave the charge to the workshop participants, to come
up with a science plan for the program. As has already been decided,
THORPEX is to address improvements to numerical weather prediction
models to the edge of deterministic predictability, specifically for
high-impact events. THORPEX is about global to regional influences on
the prediction of high-impact weather. The details of the high-impact
weather are beyond the scope of THORPEX.
The remainder of the meeting was filled with discussion of the four
sub-programmes by the committee leaders and discussion of the
sub-programme plans. I have distributed the sub-programme plans to
various people here, and if you are interested, I can send out the
most recent iterations of these documents. These discussions were
punctuated with discussion of short-term projects during the next
year and opportunities for research in the near-term.
- Sub-programme I
Observing System Development and Evaluation, was
presented by Dave Carlson. My main comment on this sub-programme
was that the oceans should not be ignored since they are important
for prediction of high-impact weather, especially to the two-week
time scale. Andrew Lorenc stated that he believes THORPEX is an
atmospheric program and needs to focus on that. No further
discussion ensued.
- Sub-programme II
Predictability and Dynamical/Physical Processes was
presented by Prashant Sardeshmukh and Michael Morgan. Discussions
did center on Tropical-extratropical interactions, and the Tropics
received some interest from the participants. How this new interest
changes the THORPEX plans is still unknown. It is important to note
that Michael Morgan works in tropical and tropical cyclone research.
- Subprogramme III
Data assimilation and observing strategies, was
presented by Florence Rabier, Chris Snyder, Lars-Peter Riishojgaard.
This included lengthy discussion concerning data versus information,
that THORPEX should not be focussed on gathering more data, but on
gathering data that can be assimilated into the models to improve
forecasts. This part of the data was called information.
Discussion of whether THORPEX should be concerned with continuing
gradual improvement to numerical guidance or to eliminating busted
forecasts of high-impact events also occurred.
- Subprogramme IV
Social Economic Impact Assesssment, was presented by
Rebecca Morss and Lenny Smith, neither of whom have much experience
in this field. They suggested that forecasts themselves have no
value, that the value comes from reaction to the forecasts. They
discussed some simple tests, such as the Weather Roulette, that can
show whether action should be taken in certain events or not, and
that results pertinent to one area may not be pertinent to others.
They discussed the recent oil spill off the coast of Spain, how the
ship was going through an area which was supposed to experience a
storm, but the captain continued through to save money. Insurance
and reinsurance companies are beginning to void policies when
high-risk activities are pursued instead of potentially more costly
alternatives. One goal of this group is to present potential
economic benefits to industry groups so that these groups can help
in the funding of THORPEX. They discussed the need to entrain
those with experience into the field.
- TOST I: North Pacific THORPEX Observing System Test
This is really just WSR. I think it is an effort to have a field
program with the word THORPEX in it. There may be some satellite work
in conjunction with the G-IV work, but that has yet to be determined.
- TOST II: North Atlantic THORPEX Observing System Test
The second TOST is much more applicable to our own work. It will
take place October and November, 2003 in the North Atlantic basin. It
will involve special satellite rapid scanning, an ASAP 15 ship
rawinsonde network, AMDAR on 550 commercial aircraft to get flight level
data across the ocean, European Union Upper Air and Surface data special
obs, European BA146 Research Aircraft, 10-20 new drifting buoys. The
Aerosonde and driftsonde may also be available for this program.
Sarah Jones and I mentioned that this might be an opportunity for a
cross-basin extra-tropical transition experiment in which HRD could use
2004 flight hours, and we should try to get the Canadians involved as
well.
- Earth Simulator and AFES (AGCM for the Earth Simulator) -
Wataru Ohfuchi
The Earth Simulator is now the fastest computer in the world and is
available for simulation studies. The computer is 40 Tflops, and has
10 Tbytes of memory. An integration of a T1279L96 simulation, based on
the CCSR/NIES Japanese university model. Kurihara is involved in this
effort.
- Multi-scale features of the Baiu (Meiyu) frontal zone simulated
using AFES - Enamoto Takeshi
A simulation of the Baiu front was shown.
- Breakout sessions and further discussions:
-
Possible THORPEX workshop for satellite people. Thinking
especially of the tropics.
- New observations, especially in the tropics, needed to
diagnose and reduce model errors. Data assimilation in the tropics is
not to the point where extra data may be effective.
- Extend extratropical techniques to the tropics past the limits
of linearity (diabatic singular vectors).
- Get others involved. Industry interested, will go to Congress
for money.
- There will be a call for THORPEX proposals in next few months.
- THORPEX/Vaisala Cooperative Fellowship programme - two to three
for 2-3 years, postdocs to work on THORPEX issues. Host lab pays 50%.
Contact Walter Debberdt of Vaisala. This will be out around April.
- Government spending is an "accounting problem akin to ENRON."
-- Dave Rogers, on funding
- OSSE versus OSE. OSE better, but nobody besides Aberson has
done this to show that targeting really works. OSSE is too difficult
because of model and sensor changes, too difficult to properly set up.
Can be done in simplified system.
- Definitions - TOST is for technology assessment, regional
experiment is more ambitious and scientific.
- THORPEX must do what will not already happen at National
Centres. Research and operations must work together, and this is a very
difficult problem. More must be done than have researchers visit
operational centres.
- Joint Data Assimilation Announcement of Opportunity soon.
- THORPEX Science Symposium for 2003.
- International Science Plan will be ready in next few weeks.
Sim Aberson
AOML/Hurricane Research Division
Miami, FL
December 13, 2002
|
|