[PROJ] Preferred grid format for transformations?

[Martin Desruisseaux]

I will not insist for NetCDF, but just want to bring some clarifications:

Le 25/11/2019 à 23:44, Even Rouault a écrit :

> - HDF5 is indeed much more powerful than GeoTIFF. We would have to 
> restrict even more severely a profile than I proposed to do with 
> GeoTIFF to avoid having to deal with crazy formulations. Actually with 
> my hat of GDAL developer on, this very flexibility of HDF5 is a 
> serious problem because data producers tend to follow their own 
> personal inspiration of how to structure the data, and in particular 
> interoperability of geoferencing encodings is close to null.
>
HDF5 is only a binary container. How data are organized inside that 
container is defined outside HDF5. In our case, they are defined by 
CF-Conventions. This is the same separation than XML versus schemas: we 
can see HDF5 as a kind of binary XML + arrays, and CF-Conventions as a 
XML schema for NetCDF/HDF. The CF-Convention addresses the 
georeferencing encoding in various ways. One way standardized by 
CF-Conventions is to use WKT [1].


> the cloud-friendliness of HDF5 is unknown to me. On the contrary, I 
> know that COG is a technology used heavily.
>
I did not benchmarked HDF5 myself, but I attended a talk in an Apache 
Conference 2 or 3 years ago were such benchmarking in the cloud had been 
done for various formats (I do not remember if TIFF was among them 
however). HDF5 performances on the cloud were reported good except for 
arrays of character strings (a problem of data chunk size not matching 
the size of data block transferred on network).

Note that HDF5 is in active development (I see 9 commits yesterday, 82 
commits this month), is governed by a Technical Advisory Board since 
2018 [2] and has a massive community-built software ecosystem according 
their web site [3]. HDF was initiated by the National Center for 
Supercomputing Applications (NCSA) and is used heavily too (e.g. at 
NASA). It is the de facto standard in the scientific and research community.

But as said before I do not insist if the preference is still for 
GeoTIFF in PROJ. I just wanted to reply to some of the arguments that 
were advanced in this thread.

     Regards

         Martin

[1] http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#use-of-the-crs-well-known-text-format
[2] https://confluence.hdfgroup.org/display/HDF5TAB
[3] https://www.hdfgroup.org/community/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20191126/fef8d319/attachment.html>