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

[IDV #ALX-750694]: reading gridded netcdf time-series output in idv



Greetings Larry!

I've taken a look at your file, and you are correct that the issue is that it 
is not
CF compliant. I've attached an NcML file to this email, and if you place this in
the same directory as your netCDF file and load it (the NcML file) into the IDV,
things will work as you expect. 

This fix will allow you to use the IDV with your model out. Please keep in mind
that this will not make it CF compliant, but rather "good enough" for the IDV.

The main issue:

The time variable, t, has a unit attribute set to hours and a attribute 
ReferenceTime
set to the reference time - in order to be CF compliant, the reference time must
be part of the unit:

attribute unit : "hours since 1991-03-19 18:00:00"
(see 
http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.6/cf-conventions.html#time-coordinate)

All this NcML file does is change the attribute unit on the time variable t. 
Note that
if you want to use this on a different file, you will need to edit the second 
line of the
NcML file to correctly reference the new netCDF file. To make life easier, you 
may
want to contact the modeling group and see if they would fix the unit for t in 
their
code.

Also, I would strongly recommend that when working with this data, you use the
"Predefined -> South America" projection and turn off the option "Auto-set 
Projection" - both are found under the Projections menu just above the main
view window. If you don't do this, you will get some pretty funky artifacts due
to the prime meridian.

Cheers!

Sean

> Dear support-idv:
> 
> I am unable to import a time-varying 2-d gridded netcdf variable into idv
> and was wondering if you might have some advice as to what I'm doing wrong.
> 
> For example, I am using the following netcdf file:
> ftp://ftpext.usgs.gov/pub/wr/wa/vancouver/Mastin/Ash3d/3d_tephra_fall.nc
> 
> I am attempting to read the 2-D, time-dependent variable cloud_load, whose
> properties are given as follows when I type "ncdump -h"
> 
> double cloud_load(t, lat, lon) ;
> cloud_load:long_name = "Cloud Load" ;
> cloud_load:units = "T/km2" ;
> cloud_load:missing_value = "-9999.f" ;
> cloud_load:_FillValue = -9999. ;
> 
> When I open this file using the IDV data chooser, only three variables are
> visible, as you can see in the attached screenshot:
> 
> Deposit arrival time
> Airborne ash arrival time
> Cell area
> 
> These variables are 2-D gridded, non-time dependent variables, having the
> dimensions:
> 
> double depotime(lat, lon) ;
> double ash_arrival_time(lat, lon) ;
> double area(lat, lon) ;
> 
> The netcdf file contains several other variables with dimensions (t, lat,
> lon), like cloud_load, but none of these show up.
> 
> Our modeling group wrote the code that creates the netcdf file and we think
> the file is nf-compliant.
> 
> Do you think this is an issue involving compliance with netcdf conventions,
> or is there a trick to viewing time-dependent gridded netcdf data that I
> haven't yet found?
> 
> Your help would be appreciated.
> 
> Thanks,
> 
> Larry Mastin
> 
> 
> --
> U.S. Geological Survey
> David A. Johnston Cascades Volcano Observatory
> 1300 SE Cardinal Court, Bldg. 10, Suite 100
> Vancouver, WA 98683 USA
> tel. (360) 993-8925
> fax (360) 993-8980
> 
> 

Ticket Details
===================
Ticket ID: ALX-750694
Department: Support IDV
Priority: Normal
Status: Open