in-house 1s 33 cat land use/30 s 19 cat soil type working in WPS 4.2 but not 4.5 or 4.6

[wcheng]

I have an in-house 1 sec 33 category land use (compatible with the LANDUSE.TBL - USGS) and 30 sec 19 category soil type (compatible with the
SOILPARM.TBL - STATS). This dataset has been working fine up to WPS 4.2. The usually steps are taken in modifying GEOGRID.TBL. I'm not
sure what has changed since 4.2 but with 4.5 and 4.6, process_tile_module.F is different.

WPS 4.5 and 4.6 produced weird results in the soil type (water is 1e20) when the 1 sec land use is used. This is irrespective of the 30 sec 19 category soil
type or the default 30 sec 16 category soil type.

Attached are the results of WPS 4.5 and 4.6. Does anyone know what the problem is?

Here are the changes is GEOGRID.TBL:

landmask_water = landuse_1s:16 # Calculate a landmask from this field
interp_option = landuse_1s:nearest_neighbor
rel_path = landuse_1s:landuse_1s/

 

[kwerner]

Hi,
Unless you have a specific reason for needing to update WPS to a newer version, you can continue to use the older version (V4.1) to do your preprocessing. Even when updating WRF, you do not necessarily have to do the same for WPS. That may be the easiest solution for this issue.

 

[wcheng]

Hi,
Unless you have a specific reason for needing to update WPS to a newer version, you can continue to use the older version (V4.1) to do your preprocessing. Even when updating WRF, you do not necessarily have to do the same for WPS. That may be the easiest solution for this issue.

Thanks for the response. Actually, if you do this, you need to change the header in the geogrid file so that the version matches the metgrid.exe you are using. There is a check for the version in the geogrid and the metgrid.exe. If they are different, I've seen the soil type messed up. I am actually using 4.2. There seems to be additional variables in 4.5 and 4.6. I'll try and see if the results come out reasonble.