Snow parameter in WRF EMS

All other topics about postprocessing model data (GrADS and other software), about other numerical weather prediction software (including WRF-NMM and WRF-ARW discussion unrelated to UEMS/WRF EMS), and general meteorology talk go in this forum.
Post Reply
wettersat
Posts: 32
Joined: Fri Aug 21, 2009 9:30 pm

Snow parameter in WRF EMS

Post by wettersat » Sat Oct 10, 2009 9:36 pm

In NMM v3 I use 'd WEASDsfc' for snow. It seems that this is cumulative snow and not snow rate (e.g. per hours) like for rain

Gerald

wettersat
Posts: 32
Joined: Fri Aug 21, 2009 9:30 pm

Re: snow parameter

Post by wettersat » Sun Oct 11, 2009 10:58 am

Upps, WEASDsfc is accum.snow. Now I use APCPsfc*CSNOWsfc and it seems ok.

Gerald

meteoadriatic
Posts: 1602
Joined: Wed Aug 19, 2009 10:05 am

Re: snow parameter

Post by meteoadriatic » Sun Oct 11, 2009 11:35 am

wettersat wrote:Now I use APCPsfc*CSNOWsfc and it seems ok.
Me too, but be careful. I think that variable CSNOWsfc refers to single one timestep, that is, the time of model output, but nothing in between two model output times as APCPsfc cumulative precipitation is. What that means? If you for example in your model have snow between 13:00 and 13:59, and then at 13:59 snow stops, you will have in 14:00h output CSNOWsfc=0. But, between 13:00 and 14:00 model will accumulate much of precipitation and show that in APCPsfc variable. When you print APCPsfc*CSNOWsfc you wont be able to see any snow on your graphics, because CSNOWsfc=0. In fact, your snow will look like rain on map. This is especially problem if you use 3 or 6hr output step.

How to handle this? No idea. I pulled my hair off last winter because of this. Now I plan to plot WEASDsfc on map and use that variable to distinct will there be (new)snow on ground or not.

meteoadriatic
Posts: 1602
Joined: Wed Aug 19, 2009 10:05 am

Re: snow parameter

Post by meteoadriatic » Tue Oct 13, 2009 8:32 am

I see now that WEASDsfc is not good. The problem is in initial conditions from GFS where there is too broad regions of snow on ground - WRF just takes this as starting point and this looks really bad on fine resolutions.

meteosciez
Posts: 55
Joined: Sat Sep 05, 2009 6:36 am

Re: snow parameter

Post by meteosciez » Wed Dec 22, 2010 10:23 am

meteoadriatic wrote:I see now that WEASDsfc is not good. The problem is in initial conditions from GFS where there is too broad regions of snow on ground - WRF just takes this as starting point and this looks really bad on fine resolutions.
I observed the same problem with the initial conditions when trying to generate snowdepth maps. I have decided therefore to substract the initial condition data and cumulate only the changes in snowdepth between hours (snow fall or snow melt calculated by the model); in order to display only the snow accumulated during the WRF run period.

Code: Select all

runscript = 1 
dist_t = 1
total_fcsthrs=97
freq=1

while(runscript)
'set t ' dis_t  

'set gxout shaded'
'set map 15 1 10'
#Plot snow  - dist_t = time, with dist-t=1 for hour +00  (initial conditions)
if (dis_t="1")
   'define lastsnow = SNODsfc(t='dis_t')-SNODsfc(t='dis_t')'
   'define cumulsnow = lastsnow - SNODsfc(t='dis_t')'
 endif

'define snowdepth = SNODsfc(t='dis_t') + cumulsnow - lastsnow'
'define snowdepth = const(maskout(snowdepth,snowdepth),0.0,-u)'
'define cumulsnow = snowdepth'
'define lastsnow = SNODsfc(t='dis_t')'
 
'run color_snow.gs'
'd snowdepth*100'

'set gxout contour'
'set cint 5'
'set cthick 4'
'set ccolor 0'
'set cmin 0'
'set cmax 100'
'set clab on'
'd snowdepth*100'

'run cbarnskip.gs 1 0.75 0'

if ( dis_t = total_fcsthrs )
 runscript=0
'close 1'
'quit'
endif
dis_t = dis_t+freq
endwhile
You can see my snow maps here : http://www.meteo-sciez.com/wxwrfg.php?map=3&param=snow
meteo-sciez.fr

alfe
Posts: 98
Joined: Thu Nov 25, 2010 8:13 pm

Re: snow parameter

Post by alfe » Sun Jan 30, 2011 5:46 pm

Hi guys,
I also saw this problem, and came to the same conclusion : initialization with GFS field is too coarse. So I do the same. I substract the initial depth to the final depth obtained during the simulation....

beowrf
Posts: 43
Joined: Sat Jul 23, 2011 3:07 pm

Re: Snow parameter in WRF EMS

Post by beowrf » Wed Nov 02, 2011 9:53 am

Thanks, meteosciez!
That was what I was looking for.

Best regards,

Julien

BCHurricane89
Posts: 38
Joined: Thu Apr 28, 2011 4:20 am

Re: Snow parameter in WRF EMS

Post by BCHurricane89 » Fri Dec 06, 2013 2:01 am

Sorry to revive an old thread, but I recently ran into this problem as well.

The issue of using the water equivalent of accumulated snow depth (WEASDsfc) and subtracting the initial conditions is that this field itself (WEASDsfc) accounts for snowfall and snowmelt, so your plots account for positive and negative values of snowfall. This means you are having snowfall values subtracted during the hours, so you not getting the true/total amount of snowfall.

I want to account for JUST the positive snowfall so I can plot the accumulated total snowfall. Unfortunately, I have yet to find away to do this, as the previous method of APCPsfc*CSNOWsfc does not count for snowfall in between output times.

Post Reply