Hi
I’m trying to flatten some orography over part of my domain. I realise I can’t just use xancil as then my other orography ancillary fields won’t match. I’ve tried using ANTS but then realised I don’t generate the silhouette geographic roughness or half peak to trough (stash req 17 & 18 in the below namelist) via ANTS that’s done via CAP only
ancilfilename=‘$UM_ANCIL_OROG_DIR/$UM_ANCIL_OROG_FILE’
domain=1
!!interval=1
l_ignore_ancil_grid_check=.false.
!!netcdf_varname=9*‘unset’
!!period=1
source=2
stash_req=5,6,7,17,18,34,35,36,37
update_anc=.false.
!!user_prog_ancil_stash_req=
!!user_prog_rconst=0.0
What I am missing? I’m guess this is done fairly often so some one must know how to do this, any pointers would be useful or a name of someone to ask!
Cheers
Helen
Helen
We have a suite that runs the CAP - I guess you would need to provide an orography source file with the flattening, then let the CAP use that to create ancils for your domain. Do you have a global orography with the required flattening?
(there may be cleverer ways to do this, but nothing else has come to mind)
Grenville
Hi Grenville
I tried that with a RAS ID u-cw696 but it generated blank files. I took the globe 30 data set and modified it to a new netcdf4 /work/n02/n02/hburns/Terramaris/TerraMaris/ANCILS/big_grid/globe.nc , but CAP couldn’t read it I tried using xancil to make an orog ancillary but this only worked with longitude being between 0-360 where as globe data set on values are 0 -36000. I made an ancil with the lon values converted to 0-360 /work/n02/n02/hburns/Terramaris/TerraMaris/ANCILS/big_grid/qrparm.orog_big_360 and put that as the source file but I just got a blank orography file.
The suite dir on archer is /work/n02/n02/hburns/cylc-run/TM_RAS/
Any advice on if I’m passing the source file in incorrectly or if I need to have the longitude values in 0-36000 and the file format to be an ancil file how to make the file would be much appreciated.
After some trial and error and then some discussion with met office scientists I’ve got some experiments working and looking sensible.
In case anyone stumbles across this looking for how this works: The other orography fields in qrparm.orog are all the subgrid fields used in stuff like the gravity wave drag scheme, so they need to be calculated from the high-res datasets and for some fields, these are even pre-calculated, and CAP simply regrids the precalculated fields (you see CAP referring to a .ash file). That’s why ANTs just uses CAP output to create some of the fields and if you just edited the global source data and not the precalculated fields to be regridded, you would also end up with inconsistent fields.
If you want to remove orography, e.g. over an island, you can just set the other fields to a constant value, for a pan flat plain this could be zero. If you want to apply a filter you can apply the same filter across the fields to keep approximately consistent. Then, use Xancil to create your ancillaries from all the modified fields.
If you’re adding in an idealised orography, you can take typical values from the region you’re replicating and apply them as constant values to your idealised mountain or create some variation with height based on averages across a region of interest. This would replicate your jagged mountainsides.