2m temperature Alps

Forum dedicated to older versions of EMS package (WRFEMS v3.2, v3.1 or older). Support is user-to-user based, so please help others if you can.
Antonix
Posts: 256
Joined: Fri Oct 16, 2009 8:53 am

2m temperature Alps

Post by Antonix » Sat Apr 21, 2012 7:40 am

Hello everyone.
is a long time that I try to predict the temperatures in the mountains-
for example the city of Aosta (Italy) is located in the heart of the Alps and I can never predict exactly temperatures at these locations.
you have any advice for me??

I tried to increase the vertical levels, resolution, pbl, LSM ...
but when I extract the data with grads is always a given ... VERY underestimated!
(in the order of 5-10 °)

you have similar problems??


best regards!!!

meteoadriatic
Posts: 1512
Joined: Wed Aug 19, 2009 10:05 am

Re: 2m temperature Alps

Post by meteoadriatic » Sat Apr 21, 2012 8:23 am

That is probably because height in model is much higher than actual height of that location. If location is surrounded with high mountain peaks, model will raise elevation of that valley because of too coarse horizontal resolution.

So you basically have two options:
- either raise horizontal resolution of model (resource hungry), or
- make correction in postprocessing (you will have to use your imagination how :) )

Antonix
Posts: 256
Joined: Fri Oct 16, 2009 8:53 am

Re: 2m temperature Alps

Post by Antonix » Sat Apr 21, 2012 8:40 am

meteoadriatic thanks!
you figured the problem was the topography.
you think you can extract the height (in meters) of the location with grads?!

meteoadriatic
Posts: 1512
Joined: Wed Aug 19, 2009 10:05 am

Re: 2m temperature Alps

Post by meteoadriatic » Sat Apr 21, 2012 8:47 am

Yes there is some variable like "surface height" or something like that, I don't remember exactly. You might need to turn it on in static/wrfpost_cntrl.parm file.

Yes, it is hgtsfc in GrADS.

Antonix
Posts: 256
Joined: Fri Oct 16, 2009 8:53 am

Re: 2m temperature Alps

Post by Antonix » Sat Apr 21, 2012 9:17 am

ok!
in the file. ctl the variable is not present.
in wrfpost.cntrl.parm is present:

(SURFACE HEIGHT ) SCAL=( 6.0)
L=(10000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000)

I must change that!?


(SURFACE HEIGHT ) SCAL=( 6.0)
L=(11111 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000)

meteoadriatic
Posts: 1512
Joined: Wed Aug 19, 2009 10:05 am

Re: 2m temperature Alps

Post by meteoadriatic » Sat Apr 21, 2012 9:19 am

Antonix wrote:(SURFACE HEIGHT ) SCAL=( 6.0)
L=(10000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000)
This is OK. Do you have hgtsfc variable in .ctl file?

Antonix
Posts: 256
Joined: Fri Oct 16, 2009 8:53 am

Re: 2m temperature Alps

Post by Antonix » Sat Apr 21, 2012 9:37 am

yes, but the height of the surface geopotential, is different from the height in meters of a point.
I would know Aosta (lat lon point) (for example) at what height you triva, for the model

meteoadriatic
Posts: 1512
Joined: Wed Aug 19, 2009 10:05 am

Re: 2m temperature Alps

Post by meteoadriatic » Sat Apr 21, 2012 9:46 am

hgtsfc is in geopotential meters which is almost the same as meters. What you get when you do:

in my very coarse (36km) domain I get:

Code: Select all

ga-> open 201204151200_arw_wrfout_d01.ctl 
Scanning description file:  201204151200_arw_wrfout_d01.ctl
Data file 201204151200_arw_wrfout_d01.grb is open as file 1
LON set to -13.03 38.77 
LAT set to 27.873 63.247 
LEV set to 1013 1013 
Time values set: 2012:4:15:12 2012:4:15:12 
E set to 1 1 
ga-> set lat 45.73
LAT set to 45.7165 45.7165 
ga-> set lon 7.31
LON set to 7.29658 7.29658 
ga-> d hgtsfc
Result value = 2180.62 
ga-> 
So Aosta is at 2180 meters above sea level in my 36km ARW domain. In reality it is at 580m. You can imagine how different forecast will be for that point.

What do you get when you issue these commands in GrADS for your finest domain?

Antonix
Posts: 256
Joined: Fri Oct 16, 2009 8:53 am

Re: 2m temperature Alps

Post by Antonix » Sat Apr 21, 2012 10:11 am

Notice: Implied interpolation for file control1.ctl
Interpolation will be performed on any data displayed from this file
ga-> set lat 45.73
LAT set to 45.7213 45.7213
ga-> set lon 7.31
LON set to 7.31262 7.31262
ga-> d hgtsfc
Notice: Automatic Grid Interpolation Taking Place
Result value = 1095.17
ga->

This is my result at a resolution of 3 km!!!!!!
This is my result at a resolution of 3 km!!!!!!
is crazy ... No solution exists?!

I find the point at 500 meters, but then throughout Italy would be very difficult!

meteoadriatic
Posts: 1512
Joined: Wed Aug 19, 2009 10:05 am

Re: 2m temperature Alps

Post by meteoadriatic » Sat Apr 21, 2012 10:16 am

Yes, there are >3000m peaks around Aosta, so that is what it is.

If you managed to find nearby point that has model elevation 500m, then try to use that location for Aosta temperature.

Another question: did you used enough terrain resolution to create your domains? How did you created domains, with dwiz or ems_domain.pl script? And, what wrfems version you used to create domains?

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest