Re: big difference of snow height between ccm3 and NCEP/NCAR reanalysis

Subject: Re: big difference of snow height between ccm3 and NCEP/NCAR reanalysis
From: Peter Paul Smolka (
Date: Fri Aug 24 2001 - 09:28:07 MDT

On Thu, 23 Aug 2001, Xuguang Wang wrote:

> Date: Thu, 23 Aug 2001 17:15:33 -0400 (EDT)
> From: Xuguang Wang <>
> To: ccm-users@UCAR.EDU
> Subject: big difference of snow height between ccm3 and NCEP/NCAR
> reanalysis
> Dear ccm3 users:
> I found there is a huge snow height difference between ccm3 and
> NCEP/NCAR reanalysis.
> For example, around south pole, in September,2000, the snow height given
> by NCEP/NCAR reanalysis is around 50m. While looking at file
> (climatological equilibrium state in September) from ccm3
> package, the snow height at the same region is around 0.15m.
> I feel this is not the problem of units. Could any of you tell me why
> there is such a big difference?

Dear Xuguang,

your e-mails touch two different topics, one of them (the second)
was also discussed as future opportunity on various congresses
(for example the last International Geological Congress in Rio, 2000).

1) The numbers themselves:

If you scrutinize all data thoroughly you will find many differences.
For example in the old version of the initial conditions the ocean
is, due to the spectral nature of the model (see on this the handbook
of Trenberth) below zero near the north pole etc. (if I remember

The good thing is that even though there are such things the results
are remarkably usable although of course ongoing progress is the major
part of our work (better SSTs, better parametrizations, better
processes etc.)

2) The second part is of more interest, maybe also for you (as this can
only be a distributed initiative with knowledge from various sides):

The models we have are factually weather models, although we apply them
for climatic questions (such as environments in the Miocene).

All calculations, even with a coupled model, are factually equilibrium
runs, as long times (for example several glacial/interglacial
cycles)are still out of question.

In some oceanographic centers the need to reassess some of the algorithms
is seen, especially as severalequations are (time-consuming) solved

On the other hand quite usable box-models are developed at various places.

One thing could thus be:

a) To develop / couple various good box models but with data for each
grid point (models that simulate for example situations 12 times a year)
(= the detail of the box model, run at all gridpoints and with an
eye on long-term, non-iterative solutions). I am aware of what
I write (also the difficulties involved).

b) using initial data with average fluxes for representative
situations (summer,
spring, winter etc.), see representative CO2 fluxes etc. depending
on the state of the vegetation, precipitation,ice builtup (factually
the glacial/interglacials are more a classical self-organizing reservoir
problem than a weather problem) (examples for the successful simulation
of for example salinar cycles with self organizing models can be given,
including concepts on "how to tackle the problem").

c) from time to time run by the environment model a classical weather
model such as ccm3 to obtain new
precipiations, winds etc etc. (orography, vegetation etc. as result of

d) continue then with (b).

I am, depending on my time, anyway planning such work. Especially
if right from the start algorithms are designed with an eye on
long times quite complex processes might be modelled at high resolutions
as well.

This should NOT be a criticismof existing works (without them we
couldn t achieve anything). It is only a potential concept for future
progress in the light of current knowledge and technology.

If you or parts of your group are thinking this way as well, we might
the mid range (maybe together with others thinking this way) arrive
at a climate model (not a weather model) that utilizes for
example ccm3 for the weather for respective situations.
(= a work that you or others can only do in addition to
ongoing work as for example a PhD depending on this would be
"one of the last adventures life can offer").

Note that I am aware of the difficulties, changing orographies etc.
by the program, feeding them in a suitable way to ccm3 etc etc.

The good thing is that we do not need a Cray for it (= the main point
are the good ideas).

If you (or others) are in the mid-range interested in such an
initiative, we might discuss respective ideas off-list.
> Thanks,
> Xuguang

Best regards, Peter

Dr. Peter P. Smolka
University Muenster
Geological Institute
Corrensstr. 24
D-48149 Muenster

Tel.: +49/251/833-3989 +49/2533/4401
Fax: +49/251/833-3989 +49/2533/4401

This archive was generated by hypermail 2b27 : Tue Jan 22 2002 - 11:16:00 MST