VINT

[Henrique Barbosa]

Remi,

I guess you are right, grads do consider the portion below ground.
Moreover, it extends the value of the variable below your base level, ie.,
it assumes the value at z=1 to be valid between ps and the first level.

Here are my results to your proposed tests:

I took a random grid point of my model data, where:

z=2    975mb     8.389g/kg
z=1   1000mb    9.356g/kg
ps=1058.55mb

grads says that:

vint(ps,q,300)=17.3869 kg/m2
vint(const(ps,1000),q,300)=11.797 kg/m2

Well, as you said, the results are quite different:

17.3869-11.797=5.5899 kg/m2

But, if we take the second result and add the portion
between 1000-1058.55mb,  we get the first result:

q(1000mb) * 58.55mb * 100 / g = 5.58952 kg/m2

Moreover, in my example, the first two levels are:
1000mb and 975mb. This means that the separation
between these two is 987.5mb. Now look at these results:

vint(const(psmt,1000),uemt,300)=11.797
vint(const(psmt,990),uemt,300)=10.8177
vint(const(psmt,980),uemt,300)=9.96168
vint(const(psmt,970),uemt,300)=8.99535

and so on... Hence, I was wrong in my last email. What
grads really do is consider the full layers above ps plus
the upper fraction of the layer that contains ps. In case
ps is below the lowest layer, it extends this layer until ps,
by keeping constant the value of the integration variable.

I think that it is now settled... But I do agree with you
that the grads manual is not clear at all about this matter.

Sincerely,
Henriique


>  let's consider the case where ps=1030 >1000
>  I ve tried vint(ps,q,300) and vint(const(q,1000),q,300) and it's not the
>  same the result !!
>
>  And i have not tried for the reverse case (ps=915)<1000 but let's say the
>  variable are defined on levels 1000 925 850, i am not sure that :
>  vint((ps=915),q,300)=vint(const(q,850),q,300).
>
>  So for me there is still something not clear in grads documentation.
>  I believe that these questions are especially relevant to regions with
>  high altitude relief.
>  Thanks in advance
>
>  Remi
>
>
>
>
>  > Dear Meynadier,
>  >
>  > From the grads documentation:
>  >
>  > (...)The summation is done for each layer present that is between the
>  > bounds. The layers are determined by the Z levels of the default file.
>  > Each layer is considered to be from the midpoints between the levels
>  > actually present, and is assumed to have the same value throughout the
>  > layer, namely the value of the gridpoint at the middle of the layer.
>  >
>  > If ps<1000, e.g. 915mb, the integration will exclude all levels below
>  > 915mb.
>  >
>  > If ps>1000, e.g., 1030mb, I guess (from the explanation above) that only
>  > levels
>  > starting at 1000mb and above will be used... and the 30mb between ps and
>  > first
>  > level will just be lost.
>  >
>  > Sincerely,
>  > Henrique
>  >
>  > On Thu, Feb 28, 2008 at 8:33 PM, Remi Meynadier
>  > <remi.meynadier at aero.jussieu.fr> wrote:
>  >> Hi folks
>  >>  I am using the vint fonction to integrate specific humidity to obtain
>  >>  precipitable water.
>  >>
>  >>  This is what I ve done : vint(ps,q,100) with ps surface pressure.
>  >>
>  >>  What does the vint fonction do when the first pressure level (ie
>  >> 1000hPa)
>  >>  is above surface pressure ? An extrapolation from the first pressure
>  >> level
>  >>  ?
>  >>
>  >>  and in the same time what if ps is lower than 1000hPa ?
>  >>
>  >>  Thanks in advance
>  >>  Regards
>  >>
>  >>
>  >>  --
>  >>  Rémi Meynadier
>  >>  Institut Pierre Simon Laplace
>  >>  Service d'Aéronomie du CNRS, UMR 7620
>  >>  Université Pierre et Marie Curie
>  >>  Tour 45-46, 3e étage, Case 102
>  >>  4 place Jussieu, 75252 PARIS Cedex 05
>  >>  Tél: 01 44 27 84 45 Fax: 01 44 27 37 76
>  >>
>  >
>
>
>  --
>  Rémi Meynadier
>  Institut Pierre Simon Laplace
>  Service d'Aéronomie du CNRS, UMR 7620
>  Université Pierre et Marie Curie
>  Tour 45-46, 3e étage, Case 102
>  4 place Jussieu, 75252 PARIS Cedex 05
>  Tél: 01 44 27 84 45 Fax: 01 44 27 37 76
>