[THREDDS #WWR-473152]: strange CSV result from thredds.ucar.edu

Hi Daryl,

So the problem is that the GFS output, for whatever reason, includes 3 hour 
totals out to 381 hours, 6 hour totals out to 384 hours, and run totals from 
0-3 hours to 0-384 hours. And of course in GRIB land, these aren't encoded as 
different parameters, but as identical parameters (Total precipitation) with 
different reference time periods.

Trying to programmatically map the space of valid GRIB encodings to valid 
netCDF mappings means either mapping the identical parameters to a single 
netCDF variable, and then encoding the proper time bounds (what you see 
currently), or trying to map each unique period length to its own variable. Due 
to the varying run total periods, this would result in about ~130 different 
variables, with 128 of them having only a single value. So we've gone with 
option (1).

Regardless, we probably should look at improving the output for grid-as-point, 
since none of the formats include the needed time bounds.

Cheers!

Ryan

> Greetings,
> 
> Thanks for your continued interest.  Indeed, that would not be very useful.  
> What I desire is that this:
> 
> "Total precipitation (Mixed_intervals Accumulation)"
> 
> become
> 
> "Total precipitation (3_Hour Accumulation)"
> "Total precipitation (12_Hour Accumulation)"
> "Total precipitation (18_Hour Accumulation)"
> 
> So that I can query just that 3 hour variable and only get those.  
> Conversely, if the start and end timestamps could be included in the CSV, 
> that would suffice as well, since I could compute the intervals.


Ticket Details
===================
Ticket ID: WWR-473152
Department: Support THREDDS
Priority: Normal
Status: Closed
===================
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.