[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #ENB-133694]: NAM and GFS best time series totally different at same time



Greetings Brian,

I see the issue. 

> In two NCEP analysis products over the US?

When you look at data from the GFS 2.5 degree model, you are not
looking at an analysis product (grid) - the GFS 2.5 degree grids that we get
only contain the 192 to 384 hour (8 to 16 day) forecast. That is, the best
time series of the GFS 2.5 degree model will the the 'best' time series that
can be made with 8 to 16 day forecast. See the thredds catalog page for the
GFS 2.5 degree model for more info (found under Documentation -> 
summary):

http://motherlode.ucar.edu/thredds/catalog/fmrc/NCEP/GFS/Global_2p5deg/catalog.html?dataset=fmrc/NCEP/GFS/Global_2p5deg

The NAM 12km, however, contains the analysis and forecasts every 3 hours 
out to 84 hours. So, the best time series of the NAM 12 will be made up from
the analysis grids, as well as forecast grids *much* closer to the analysis time
of each run. Again, here is the summary for the NAM 12km model:

http://motherlode.ucar.edu/thredds/catalog/fmrc/NCEP/NAM/CONUS_12km/catalog.html?dataset=fmrc/NCEP/NAM/CONUS_12km

In the example image you sent, you are seeing the MSLP field of the NAM 12 
analysis grid (valid 2013-03-17 00:00:00 Z) compared to the MSLP field of
the GFS 2.5 degree forecast grid (8 or so days out from the actual initial time 
of the run) valid at 2013-03-017 00:00:00 Z.

Sean

> 20mb differences over Wisconsin?
> Isobar orientations totally orthogonal over half the country?
> In two NCEP analysis products over the US?
> 
> No way, not a chance this is real.
> Data must be mishandled, mislabeled, or mistimed.
> 
> [cid:address@hidden
> 
> 
> 
> On Mar 17, 2013, at 10:33 AM, Unidata IDV Support wrote:
> 
> Greetings Brian,
> 
> I think what you are seeing is the difference between the 12 km grid of the 
> NAM and the
> 2.5 degree (~250 km) grid of the GFS. If you decimate the 12 km NAM to use 
> every 20th
> grid point using the spatial subset, then the difference is not too bad. 
> Also, if you use the
> simple difference formula found on the "Field Selector" tab under 
> "Formulas"->\
> "Miscellaneous", the eye-ball average difference is around 10 mb for the 
> central CONUS),
> but can be as big as ~40 mb near the edge of the grid.
> 
> Cheers,
> 
> Sean
> 
> I was going to look at differences in my mesoscale class.
> But something about these Best time series time levels seems to be totally 
> wrong.
> Complete mismatch.
> ?
> 
> Created with the TDS 42. - 4.3 upgrade plugin in case it matters.
> 
> 
> 
> [cid:address@hidden
> 
> 
> 
> Brian Mapes
> address@hidden<mailto:address@hidden>
> 
> 
> 
> 
> 
> 
> 
> Ticket Details
> ===================
> Ticket ID: ENB-133694
> Department: Support IDV
> Priority: Normal
> Status: Open
> 
> 
> Brian Mapes
> address@hidden<mailto:address@hidden>
> 
> 
> 
> 
> 

Ticket Details
===================
Ticket ID: ENB-133694
Department: Support IDV
Priority: Normal
Status: Open


NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.