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

[IDV #CXU-904892]: IDV - gfs data on motherlode

Hi Bill-


> > Anyway, I've sent a note to the person in charge of these
> > things to see if the pqact pattern used for the GFS netCDF
> > data should be changed.
> >
> 
> I suspect this will take care of it.
> 
> Over the weekend, I tried to display just GFS 500mb Z for all times
> between VT 0 and 72 hours. All times were displayed, except one -
> the one I mentioned earlier. So, I don't think it has much to do with
> matching the NAM times; I suspect one 500mb Z time is just missing
> for some unknown reason. It seems to keep happening, so it might be
> the script that creates the dataset ?

It is the pattern that is used to file the data as it arrives from 
the LDM that is problematic.  The missing time in the GFS is the
effect.
 
> > In the mean time, I would urge you to start using the GRIB
> > files instead of the netCDF files.  

> That is good to know. Thanks for the heads up.
> 
> We have grib files saved locally, which I have experimented with
> through the file system. I know I can display them. But, it would
> be easier to serve them, and, from what you have said, it sounds like
> the way to go is to install a THREDDS server ?

Yes, that way you could have your own local catalog.  I also
talked to Nolan while we were at AMS and suggested that he
convert the realtime WRF runs to GRIB and serve them up through
the TDS.
 
> Over the weekend, I redid my bundle file using strides in the
> dataset properties. That works nicely and speeds things up.
> It seems to be using the cached data for the simple difference !
> I am very happy !

Glad to hear it.

> I got used to selecting times and levels in the Field Selector tab,
> but learned the hard way, that this is not the best way. Often, I
> want to limit the amount of data while I develop a display or bundle.
> When it's done, I want to increase the amount of data used; however,
> I've had trouble doing this through the Field Selector tab. I often
> have to recreate displays from the beginning. So, I am going to use
> the dataset properties from now on.

The Data Subset tab in the Field Selector only applies to the
parameter being displayed.  The only thing that is semi-persistent
is the choice of level (for good or bad).  However, given what
you are trying to do (using the same subsets for all params), the
best way is to do this at the dataset level.  Note that the vertical
stride will give different results based on the levels for a parameter.
For example, a stride of 4 will give you different levels for 
geopotential heights than for absolute vorticity because the latter
only has data at 850, 700, 500, and 300 hPa.

Another suggestion would be to load in your GFS dataset, do
the subsetting you want to do and then save it as a Favorite Dataset
by right clicking on the dataset in the Field Selector and choosing
the Save as Quick Link menu.   You can then call it up through the
Quick Links->Saved Data Source tab or through the Displays->Data Sources
menu (the latter will change to Data->Saved Data Sources in the
2.2 release).

Don  

Ticket Details
===================
Ticket ID: CXU-904892
Department: Support IDV
Priority: Normal
Status: Open