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

[IDV #XNQ-939998]: GFS Best Time Series Problems

> Yuan:
> 
> The reason why I think I'm not geteting 0 hour analyses is that there is
> 3-hour accumulated precipitation at 1800 UTC, 0000 UTC, etc.  That field
> should be zero at an analysis time.  At least that's they way it has
> been historically.  Perhaps it is using a 3-hour forecast rather than a
> 6-hour forecast, but in any event, it's not grabbing the analysis time
> as it should.
> 
> Yes, a proposed analysis dataset collection might help, although if you
> want something every 3-hours from the NAM or GFS, we still need to find
> a way to put the 3-h forecasts in there.  However, I think it will
> probably help quite a bit with the HRRR.
> 
> Jim
> 

Jim,
       The parameter of 3-hour accumulated precipitation is very unique in the 
BST, it is so called time interval variable. 
John believes it is better to use previous run value for the current run 0 
hour. Otherwise, the it is difficult to find the physical
meaning of 0 hour value for the  parameter of 3-hour accumulated precipitation.

Yuan
> 
> On 2/11/15, 10:39 PM, Unidata IDV Support wrote:
> >> Yuan:
> >>
> >> Here's another issue that might be more on the THREDDS Side than IDV.
> >>
> >> In the attached bundle gfs-BTS-problem.xidv, I'm accessing 3-h GFS
> >> Best Time Series data.  Note in the outgoing long-wave radiation how it
> >> is flicking back and forth presumably from differing model grids.  These
> >> are clearly not analysis times or forecast hours from the same model.
> >>
> >> Also, I thought that "Best Time Series" was supposed to grab
> >> analysis grids where possible?  In the attached atmos5010-nam.xidv
> >> bundle, it is clearly grabbing forecast grids when it should be grabbing
> >> analysis grids.
> >>
> >> Let me know if you have further questions.
> >>
> >> Jim
> > Jim,
> >        At this point, we know what is going on with the Best Time Series 
> > (BTS) of GFS. I think the proposed analysis dataset collection
> > will be very helpful for your case.
> >
> >        By the way, I did some tests on the NAM 12 km BTS and individual 
> > runs, it does look correct to me and I wonder how
> > you come to the conclusion of: " First, during the periods prior to the 
> > latest model run, I should be getting 0 hour analyses or
> > 3 hour forecasts from the NAM.  That isn't the case, at least at times when 
> > a 0 hour analysis is available.  Instead,
> > I'm getting a 6 hour forecast (see, for example, 00, 06, 12, or 18 UTC 
> > times early in the loop when satellite and radar
> > imagery are available."
> >
> >       I compared the first few time steps of BTS with each individual runs, 
> > not seeing any 6 hour forecast in the BTS. Do I miss
> > something?
> >
> > Yuan
> >>
> >>
> > Ticket Details
> > ===================
> > Ticket ID: XNQ-939998
> > Department: Support IDV
> > Priority: Normal
> > Status: Open
> >
> 
> 

Ticket Details
===================
Ticket ID: XNQ-939998
Department: Support IDV
Priority: Normal
Status: Open