FAQs: Frequently Asked Questions
Last Update: Oct. 22, 2002
This page contains an anonymous listing of questions and responses (by us)
to some frequently asked questions about the code distribution (DRIVER, ISBA, TEB).
This page was begun in October, 2002, and will be updated as needed.
Key:
- S: Subject of question (eg. which option, etc...)
- Q: question posed by an anonymous USER
- A: response by ISBA or TEB "team"
S: (ISBA: YISBA=DIF = Diffusion Option)
Q: INLVLD is set to 5. Is this a suggested minimum, or just a reasonable "average"?
A: This number (INLVLD = number of soil layers) is in-line with what one might consider as the maximum
"justifiable" number for operational numerical weather prediction
models...although one could argue for more on the basis of a better
"numerical" representation, especially for Richards Eq.
THAT being said, you can use whatever you feel is justifiable. For
example,
for a study on soil ice, 7 layers which were
defined by observed soil temperatures (Boone et al. 2000, J Appl. Met).
For another case (MUREX, 1995), 7 layers (Boone, 2000, PhD
thesis)
for which the thicknesses were dictated by the soil texture profile
(wherever a sharp
discontinuity occurred, a new soil layer was defined).
Habets et al. 2002, Glob. Plan. Change, has also
done some testing
along these lines.
Also, the uppermost layer thickness should, in general, range
between 1 and 5 cm or so...
it should be as small as possible to resolve the associated time
scales, but should be large
enough to have a physical meaning (can you really define a "soil
layer"/soil properties
which is, for eg., 0.001 m thick?). We typically use 3 cm, although
1 cm might be more
appropriate. Thicknesses can increase with increasing soil depth.
Summary: Thickness and number of layers depend on your application:
computing
constraints vs numerical solution, stratification of observations
within the soil, profile of roots or soil
properties within the soil: these are most important points to keep
in mind.
S: (ISBA: YISBA=DIF = Diffusion Option)
Q: Could you explain to me what "soil excess fraction profile" means?
A: This is the vertical distribution of a lateral soil moisture source
(eg. from a local
perched water table, variable ZWXS(t)). This variable was added in
order to run the MUREX case
study: for all other applications we have done, ZWXS = 0. So,
you need not worry about specification of the soil water excess fraction profile
if the forcing variable ZWXS=0 (the values for the excess
fraction profile is not used).
S: (ISBA: YISBA=DIF = Diffusion Option)
Q: Is the root fraction(layer x) equal to (root volume in layer x)
divided by (total root volume in soil column) or (root volume in
layer x) divided by (soil volume in layer x)? Something else?
A: The fraction defined on input, ZROOTFRAC(:,:) represents
the distribution (the total over the root zone should = 1, unless there are no
roots in the soil, for which case the sum = 0). This is then normalized by the soil layer thickness
within the code for application to the soil water stress term
of the stomatal resistance parameterization.