[Paraview] Time Variables in CF Conventions

[Moreland, Kenneth]

Thanks, Ken. I think I have the problem figured out. If you look closely at the time variable (which gives the time values), you will see that there are several invalid 0 values appended to the end. The netCDF reader is being stupid and add these time values, which gives an invalid time bounds. I think ParaView is seeing the invalid time bounds (begin time greater than end time) and determining that there should not be time. If you strip off the 0 values, the time loads fine.

I will make a bug report on this. Is it OK to attach the example file you gave me to it, or would you rather it not be open?

-Ken


From: Ken Sheldon <KSheldon at slb.com<mailto:KSheldon at slb.com>>
Date: Tuesday, April 26, 2016 at 3:25 PM
To: Kenneth Moreland <kmorel at sandia.gov<mailto:kmorel at sandia.gov>>
Subject: [EXTERNAL] RE: [Paraview] Time Variables in CF Conventions

Hi, Ken.
I will email you a ZIPX file , size approx. 800Kbytes.

----Ken


From: Moreland, Kenneth [mailto:kmorel at sandia.gov]
Sent: Tuesday, April 26, 2016 5:17 PM
To: Ken Sheldon <KSheldon at slb.com<mailto:KSheldon at slb.com>>; paraview at paraview.org<mailto:paraview at paraview.org>
Subject: Re: [Paraview] Time Variables in CF Conventions

That is very odd. I just had a look at the code for the reader, and both the units and the axis properties should have caused the reader to treat the dimension as time.

Is there a way I can get a copy of the netCDF file? I tried going to the web site you posted at the beginning of this thread, but I couldn't figure out how to download a netCDF file.


From: Ken Sheldon <KSheldon at slb.com<mailto:KSheldon at slb.com>>
Date: Tuesday, April 26, 2016 at 12:36 PM
To: Kenneth Moreland <kmorel at sandia.gov<mailto:kmorel at sandia.gov>>, "paraview at paraview.org<mailto:paraview at paraview.org>" <paraview at paraview.org<mailto:paraview at paraview.org>>
Subject: [EXTERNAL] RE: [Paraview] Time Variables in CF Conventions

It looks like the units in the CLASS data are, “hour since 1997-01-01 00:00:00”
In the test data, (which works fine, I confirm) , the units are, “days since 2001-1-1”

I would expect the attribute axis = ”T” to indicate this is a time dimension.  Time as a UOM might be used in any property, not just dimensions.
What do you think?


TIME1 (-1051845271)
    double,    182
    Number of attributes = 5
        units = hour since 1997-01-01 00:00:00
        time_origin = 1-JAN-1997 00:00:00
        axis = T
        standard_name = time
        bounds = TIME1_bnds



time (-781830246)
    double,    24
    Number of attributes = 7
        standard_name = time
        long_name = time
        units = days since 2001-1-1
        axis = T
        calendar = 360_day
        bounds = time_bnds
        original_units = seconds since 2001-1-1

From: Moreland, Kenneth [mailto:kmorel at sandia.gov]
Sent: Tuesday, April 26, 2016 1:14 PM
To: Ken Sheldon <KSheldon at slb.com<mailto:KSheldon at slb.com>>; paraview at paraview.org<mailto:paraview at paraview.org>
Subject: Re: [Paraview] Time Variables in CF Conventions

The netCDF CF reader in ParaView does support time dimensions. For example, the netCDF file tos_O1_2001-2002.nc used for VTK testing (https://midas3.kitware.com/midas/item/206312<https://urldefense.proofpoint.com/v2/url?u=https-3A__midas3.kitware.com_midas_item_206312&d=CwMGaQ&c=uGuXJ43KPkPWEl2imVFDmZQlhQUET7pVRA2PDIOxgqw&r=PPcXhptMPNzBCwHx-o1zqg&m=A7bpyohhfVnNQoUi0GE2BJS2jE2OkxogntGvLqTmVGM&s=HGoAZZDUus7tmDI-abU30H7Ij8Qg_EeQwanbaawq2-w&e=>) has a time dimension and it loads just fine.

As I recall, the CF convention defines the time dimension using a units property on the variable that specifies some unit of time. One problem that has occurred in the past is that sometimes the time unit is something that the reader does not recognize as time. What is the time units used in your data set?

-Ken


From: ParaView <paraview-bounces at paraview.org<mailto:paraview-bounces at paraview.org>> on behalf of Ken Sheldon <KSheldon at slb.com<mailto:KSheldon at slb.com>>
Date: Saturday, April 23, 2016 at 9:51 AM
To: "paraview at paraview.org<mailto:paraview at paraview.org>" <paraview at paraview.org<mailto:paraview at paraview.org>>
Subject: [EXTERNAL] [Paraview] Time Variables in CF Conventions

I am trying to read a CF Conventions NetCDF file  (Sea Surface Temperature from NOAA).
The CF reader in ParaView recognizes the axes and the data correctly.  However, it does not seem to recognize the TIME dimension.

Is anyone here reading CF (Climate Forecasting) files successfully?


Reference: http://www.class.ngdc.noaa.gov/saa/products/search?datatype_family=SST14NA<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.class.ngdc.noaa.gov_saa_products_search-3Fdatatype-5Ffamily-3DSST14NA&d=CwMGaQ&c=uGuXJ43KPkPWEl2imVFDmZQlhQUET7pVRA2PDIOxgqw&r=PPcXhptMPNzBCwHx-o1zqg&m=A7bpyohhfVnNQoUi0GE2BJS2jE2OkxogntGvLqTmVGM&s=SRGkdReFFtTB-YX8wNBb3I5f3OaCiZ27AIVx3iqa78k&e=>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20160426/a3898ad6/attachment-0001.html>