GSWP Meeting Report
From Mail Dated: Mon, 3 Jun 1996 12:11:36
Dear All:
On Thursday and Friday of last week (May 30-31) we held a
workshop of the Global Soil Wetness Project, immediately after
the conclusion of the ISLSCP Science Panel meeting at NASA/Goddard
in Greenbelt Maryland. There were a number of interested parties
at the ISP, several of whom stayed on to attend the GSWP meeting.
There is a growing amount of interest and support for our efforts
in the community, and we are being encouraged by scientists in many
fields.
Details and other information about the status and guidelines
for the GSWP are being updated at the web site.
Over the next couple of days all the current information will be
posted. This includes SEVERAL CHANGES which are also described below.
Please check the web site for the latest information.
OVERVIEW
The status of all the participating production groups
was compared.
There were 13 groups in at one time or another - only one has dropped
out completely. Two others will probably not finish before the
30 June deadline for the control runs, but will have results later
in the year and should still be participating in the sensitivity study
aspect of the project. Two groups had finished their control integrations
by the time of the meeting and had preliminary results to share (JMA-CCSR
and NASA/GSFC/CliRad).
The validation team is prepared to examine the point data. Comparisons
are planned with in situ, remote sensed, and inverse/balance methods. Also,
Robock & Vinnikov will also compare the 2-D gridded fields with their
gridded observations for Russia and China. The Intercomparison Center is
ready and waiting to accept data - results of comparisons between SiB and
modified bucket were shown. Also, comparisons with runoff data were
shown, indicating the need to use a river routing scheme to validate
the seasonal cycle of runoff from the land surface schemes (Taikan Oki
will examine this problem).
The final meeting of this GSWP (which, we expect, will be reincarnated
after the ISLSCP Initiative II data are released in early 1998) has been
set for February 1997 in conjunction with the Second Conference on Hydrology
at the AMS
Annual Meeting in Long Beach, California. There will be a
special session on the Global Soil Wetness Project at the meeting, and
there will also be a one-day final workshop on Friday, 7 February at the
AMS venue. Everyone is encouraged to attend, and also to submit abstracts
for the GSWP special session. The AMS deadline for abstracts is 1 July 1996.
Production groups are encouraged to present talks on their sensitivity
studies, validation participants on their work.
SPECIFIC CHANGES AND ISSUES OF NOTE TO PRODUCTION TEAM:
- There have been some additions to the output list for the production
teams. Two fields have been added to the SNAPSHOT VARIABLE list:
Available water in the top meter of soil, and a Normalized Available
Water Index. Also, Ground heat flux has been added to the Decad
mean fields list. Please see the
final list of output variables for details and
clarifications of the definitions of the various soil moisture
variables.
- Some new grid points have been added to the list of mean diurnal
cycle points (now up to 16 points). Also, the lats & lons have been
corrected to lie on the .5s instead of the .0s.
- NEW SPIN-UP CRITERION. It has been found that the requirement to
converge the soil moisture to within 5%/5mm (whichever is less
stringent) for the entire soil column is causing very long spin-up
times for some models at some points. It has been decided that the
converngence should be for soil moisture down to ROOTING DEPTH ONLY!
That is the same as the second variable in the list of snapshot
variables. THIS MAY NOT HAVE BEEN CLEAR AT THE MEETING as we also
talked about using the top meter. I also recommend that you not
proceed beyond 10 years for spinup. If you still have points that
do not converge after 10 years with the new criterion, please note
the number of points and report that to the ICC when you submit
your data.
- The shortwave code of Jean Morrill has no provision for a leap-day
(29 Feb 1988 in this experiment). It was decided to repeat 28 Feb
SW radiation factors for 29 Feb. Jean will post a fix this week on
her web site, or you may write your own fix.
- There is a BAD DATA POINT on the CD - at 67.5N, 65.5W (I=115, J=23).
- D:\data\vegetatn\veg_map\veg_clss.vgc says type 10 (tundra)
- D:\data\hydr_sol\soils\domtex.sol says type 6 (ice)
Please change domtex to type 1 (coarse) for this experiment.
- Other oddities on the CD: There are some unrealistic variations in
soil depth at some points (very deep soils at isolated grid points).
Also, there appears to be an unrealistic phase lag in the downward
longwave radiation (6-hourly values) causing it to peak near or
after sunset at most points. Thirdly, there may be some incongruities
in the monthly LAI fields. Each of these problems is beyond the
ability of a quick fix, so for this experiment please use these data
AS IS. These problems are being reported to ISLSCP Initiative I
people so that they can be corrected in Initiative I and/or avoided
for Initiative II.
- Information on submission of data to the ICC, including naming of the
files on the 8mm tape, and requested ancillary information, has been
established.
Thanks to all of you who attended or sent your opinions and findings
to me before the meeting! I think things are running remarkably well
for what is essentially a pilot project. The GSWP is truly benefiting
from your efforts, discussions and queries, and soon the community at
large will also benefit. Thanks again!
Paul A. Dirmeyer dirmeyer@cola.iges.org
Center for Ocean-Land-Atmosphere Studies http://grads.iges.org/
4041 Powder Mill Road, Suite 302
Calverton, Maryland 20705-3106 Phone: (301) 902-1254
USA Fax: (301) 595-9793
