[IDV #NKA-540396]: Possible time-matching snafu

Hi Jim-

The realtime image datasets (RTIMAGES, GINIEAST, GINIWEST, GINICOMP) only save 
the last 96 images.  There will be more Visible images because they are not 
available at night.

If you are going to use multiday images, then I suggest using the IMAGEALL and 
GINIALL datasets (and NEXRCALL for radar) which go back 45 days.  You'll need 
to select an archive date when you first load in the data (just select the 
current day), but when you time match, it will select the appropriate dates.

However, if you are running the nightly build, the image matching should not 
fail, it should just match the times it has images for.  I fixed that for you a 
few weeks ago, so make sure that you use the nightly build.

Also, in tomorrow's nightly build, there is support for time matching of 
surface observations.  Note that these only go back 9 days.

Don Murray  
> I've noticed some strange behavior with the attached script, which
> loads satellite data that is time matched to hourly model data.
> 
> Sometimes it works fine, sometimes it doesn't.  When it gags, the
> IR and WV satellite data won't load and IDV gives me a message that says
> the selected images are not available.  The visible image usually loads
> just fine.
> 
> I've tried to load the IR and WV manually with time matching as
> well and it also fails, even though it appears the data is there for a
> portion of the period covered by the model data.
> 
> I'm guess that this might have something to do with how far back
> the satellite data goes back on the server.  The vis data goes back a
> bit farther than the IR/WV, and perhaps meets the threshold needed for
> the time matching to work?  In the afternoon yesterday, the script ran
> fine, but at that time, there was overlap between the model data and the
> satellite data that was available.
> 
> Jim
> 
> 


Ticket Details
===================
Ticket ID: NKA-540396
Department: Support IDV
Priority: Normal
Status: Open