wrf top layers

[satjemkes]

hi,
is there some special tricks done in the top layers of wrf to avoid energy leakage? and if yes does this affect the top two layers or the top 5 or ??
and how does this affect the physics there?

 

[Ming Chen]

I am not aware of this. Do you have any specific documents related to this issue?

 

[satjemkes]

Not yet, we are working on this. In summary we are doing some experiments and before and after the experiment we see that top layers react differently than the bottom layers. we can think of - we are not doing our experiments correct, - there is some physics missing in the top layers, - there is some numerics going on in the top layers. We have performed similar (but not identical yet) experiments with two settings of wrf: one 45 levels first level at 50m, top level 5000 (I assume pascal) second experiment 55 levels, 10m and 10 hPa. because the experiments are not identical (so we need to repeat this in the next weeks with an identical setting) we find in the first experiment the different behaviour from say 12 km upward (top is at 19 km) and in the second 17 km (top at 29 km). In both cases it is in the regime were the layer thickness is constant 1000 m.

what i hope is that you could point us to a certain setting in wrf which we could use to test the last hypothesis (numerics) or the second (physics). We run the experiment with either conus alone or combination of conus and tropical. But as indicated we should repeat the experiments with one specific setting to arrive at a baseline (this different setting could also explain why there is a difference in starting layer 12 versus 17 km).

The experiments are not time intensive. Most of the time will be in a careful analysis of results.
if of interest it is appreciated further communications to (or wait for the publication ) :

stephen.tjemkes@verisk.com

thanks
stephen