[Paraview] curvilinear CF conventions

Tue Dec 12 11:51:11 EST 2017

David,

The ParaView netCDF/CF reader does support the coordinates specified in Section 5.2. As the name of that section implies, this specification of coordinates applies explicitly to latitude and longitude coordinates. The rational, as far as I can tell, is that the direction of each coordinate is determined by the “units” parameter being either latitude or longitude. (At least, this is how the ParaView reader identifies each coordinate.) Even if we were to generalize to non-lat/lon coordinates, this specification provides no mechanism for applying the height field that you want to apply.

This limitation might seem weird, but you have to remember that the CF convention is designed for climate and forecast data and is therefore really geared toward geospatial coordinates. It is a testament to how well thought out the CF convention is that we are able to leverage it for so many different use cases, but sometimes you run into limitations of this nature.

-Ken

From: David Deepwell [mailto:ddeepwell at uwaterloo.ca]
Sent: Monday, December 11, 2017 7:16 PM
To: Moreland, Kenneth <kmorel at sandia.gov>
Cc: David Deepwell <ddeepwell at uwaterloo.ca>; paraview at paraview.org
Subject: [EXTERNAL] Re: [Paraview] curvilinear CF conventions

Hi Ken,

So section 5.2 (http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#_two_dimensional_latitude_longitude_coordinate_variables) doesn’t apply for arbitrary coordinate variables? I would have thought that it would be general and not just for latitude and longitude.

I have used the warp by scalar before, but I would prefer it if Paraview simply understood the geometry immediately. I have also found another method to adjust the grid which is quite good. The calculator has a Coordinate Results option which creates the vector field of the grid.

Thanks,
David

On Dec 11, 2017, at 7:09 PM, Moreland, Kenneth <kmorel at sandia.gov<mailto:kmorel at sandia.gov>> wrote:

David,

The way you are attempting to specify coordinates does not follow the CF convention. Your variables have a “coordinates” property that simply says “zc.” That is not sufficient in the CF convention to use that as a Z coordinate to add to the other independent coordinates. In fact, I cannot find any part of the CF convention (http://cfconventions.org/latest.html) that allows you to arbitrarily assign a Z coordinate to a 2D grid.

That said, accomplishing what you want is easy with ParaView. Just load your data like you are now (so it comes in as a flat rectilinear grid). Then run the “Warp By Scalar” filter and set the “Scalars” property to “zc”. That will apply the elevation to your data like you want.

-Ken

From: David Deepwell [mailto:ddeepwell at uwaterloo.ca]
Sent: Saturday, December 9, 2017 12:54 PM
To: Moreland, Kenneth <kmorel at sandia.gov<mailto:kmorel at sandia.gov>>
Cc: David Deepwell <ddeepwell at uwaterloo.ca<mailto:ddeepwell at uwaterloo.ca>>; paraview at paraview.org<mailto:paraview at paraview.org>
Subject: [EXTERNAL] Re: [Paraview] curvilinear CF conventions

Hi Ken,

I’ve attached a file which I think satisfies the curvilinear CF conventions. The grid is 2 dimensional where the z coordinate (called zc) depends on the horizontal and vertical dimensions (x and z).

Cheers,
David

On Dec 9, 2017, at 11:34 AM, Moreland, Kenneth <kmorel at sandia.gov<mailto:kmorel at sandia.gov>> wrote:

David,

The netCDF/CF reader should be able to read curvilinear coordinates. There is not enough information in your email to determine whether the issue is with the ParaView reader or an issue with the data file. It would be helpful if you could send us an example file so we can replicate your problem.

-Ken

Sent from my iPad

On Dec 9, 2017, at 9:12 AM, David Deepwell <ddeepwell at uwaterloo.ca<mailto:ddeepwell at uwaterloo.ca>> wrote:
Hi all,

I have a netcdf file that follows the CF conventions in curvilinear (structured) coordinates. Paraview however doesn’t recognize the auxiliary variable as the grid variable on which the rest of the fields should be plotted on. I’m not sure if this a bug with paraview misinterpreting the CF convention or if I just happen to have an attribute missing that Paraview is looking for.

Cheers,
David
_______________________________________________
Powered by www.kitware.com<http://www.kitware.com/>

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView

Search the list archives at: http://markmail.org/search/?q=ParaView

Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20171212/24116a54/attachment.html>