Re: a problem of arbitrary initialization of LSM

Subject: Re: a problem of arbitrary initialization of LSM
From: Erik Kluzek (
Date: Wed Aug 22 2001 - 11:45:38 MDT

On Wed, 22 Aug 2001, Xuguang Wang wrote:

> Unfortunately, the arbitrary initialization of LSM does not give
> reasonable initial values as the user's guide says. Take TG, the ground
> temperature, as an example, by arbitrary initialization, TG is equal to
> 283K all over the globe. This is obviously unreasonable.
> Have any of you tried the arbitrary initialization for LSM? Could any of
> you give me some suggestions to solve this problem? Thanks.

Xuguang, "arbitrary initialization" is not meant to give "realistic" values --
just "reasonable ones". 283K over the globe is certainly not realistic -- but
it is a reasonable value that represents a value that TG could be. In other
words 170K or 370K would be unreasonable as TG could never reach those values.

Then intent of "arbitrary initialization" is to give
values that will allow you to run the model. If you run with
"arbitrary initialization" it should be understood that you plan on running
for several years to let the land-model spin-up before you start evaluating
your output. At NCAR we typically we typically let the simulation run for five
years or so before we start analyzing simulation data -- when arbitrary initialization
is used. So if we want to analyze a 5-years of simulation run with
"arbitrary initialization" we'll run for 10 ignore the first five years and
only look at the last 5.

As an alturnative you could create your own programs to create more realistic
values to start up LSM. The CCM3 web-page also has LSM initial condition dataset
for T42 that you could use.

Erik Kluzek, (CGD at NCAR)
National Center for Atmospheric Research
Boulder CO, (off) (303)497-1326 (fax) (303)497-1324
--------- Home page and public PGP key---------------


This archive was generated by hypermail 2b27 : Tue Jan 22 2002 - 11:15:58 MST