DWD composite files with new reference system produce different results

[Joshua K]

Hi,

https://www.dwd.de/DE/leistungen/opendata/neuigkeiten/opendata_mar2022_1.html

Quote:

---

A static set of test data and information for the planned switch of the geodetic reference system for some radar products to WGS84 is now available at https://opendata.dwd.de/test/weather/radar/.

The change is postponed to earliest end of April 2022. A more precise date for the switch will be given in a later newsletter.

The planned switch to WGS84 affects products WN, HG, and RV in folder https://opendata.dwd.de/weather/radar/composit/ as well as MCD (mesocyclone detection) in folder https://opendata.dwd.de/weather/radar/mesocyclones/.

---

Using wradlib.io.radolan.read_radolan_composite to read WN products still works but the data is slightly distorted in my test.

Am I using wradlib incorrectly, is this a bug or an effect of the new reference system?



(Attachments contain test files and a jupyter notebook.)

[Kai Mühlbauer]

Hi Joshua,

that's a bit tricky.

I've contacted DWD and they told me that the new/updated grid document(s) will be released soon.

What I can tell from now the changes only affect the earth model. The change is made from earth model "sphere" to earth model "wgs84".

That has some really annoying side effects as the underlying grid is shifted to some extent. You will not be able to directly compare the two grids any more.

There is a marker (VS) in the datafiles (header) which accounts for this change (sphere for VY <= 4 and wgs84 model for VS>= 5). wradlib will have to adapt it's calculation code to generate the correct grid coordinates in stereographic as well as wgs84.

I'll add more details to this thread when this transition is done.

Best,

Kai

most likely a problem with the new projection information.

[Kai Muehlbauer]

Hi Joshua, All,

I've attached a plot where I tried to visualize the difference of the
DWD grids. A notebook is attached too, which creates this plot.

Shown is the lower left corner of the domain. Red dots are the center
points of the new wgs84-based grid, blue dots are the center points of
the new sphere-based grid. Cyan dots are the lower left corner points of
the old sphere-base grid. New sphere-base grid and old sphere-based grid
are essentially the same, the old projection being in km without false
easting/northing and the new projection in m with false easting /northing

Doing this I found that the current xarray radolan reader uses the lower
left coordinates of the grid points. That leads to a shifted (by
resolution / 2) representation while plotting using xarray. I'm
considering that a bug which will be fixed (using center points in the
future)

The wradlib code base will get an update regarding the projection
information. This will be available in the next version.

Best,
Kai


Am 26.04.22 um 13:15 schrieb Kai Mühlbauer:

> colored_diff.gif
>
> (Attachments contain test files and a jupyter notebook.)
>

> --
> You received this message because you are subscribed to the Google
> Groups "wradlib-users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to wradlib-user...@googlegroups.com
> <mailto:wradlib-user...@googlegroups.com>.
> To view this discussion on the web, visit
> https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de
> <https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de?utm_medium=email&utm_source=footer>.

--
Kai Muehlbauer
Institute of Geosciences, Meteorology Section, University of Bonn
Auf dem Huegel 20 | +49 228 739083
D-53121 Bonn | kai.mue...@uni-bonn.de

[Joshua K]

Hi Kai,

thank you.

Can I assume that the next wradlib version will be released before DWD switches to WGS84?

[Kai Muehlbauer]

Hi Joshua,

there will be a wradlib release later this month. Hope it will be in time.

Best,
Kai

Am 09.05.22 um 15:00 schrieb Joshua K:

> <https://opendata.dwd.de/test/weather/radar/>.

> >
> > The change is postponed to earliest end of April 2022. A more
> > precise date for the switch will be given in a later newsletter.
> >
> > The planned switch to WGS84 affects products WN, HG, and RV in
> > folder https://opendata.dwd.de/weather/radar/composit/
> <https://opendata.dwd.de/weather/radar/composit/> as well as
> > MCD (mesocyclone detection) in folder
> > https://opendata.dwd.de/weather/radar/mesocyclones/

> <https://opendata.dwd.de/weather/radar/mesocyclones/>.

> >
> > ---
> >
> > Using wradlib.io.radolan.read_radolan_composite to read WN products
> > still works but the data is slightly distorted in my test.
> >
> > Am I using wradlib incorrectly, is this a bug or an effect of the
> > new reference system?
> >
> > colored_diff.gif
> >
> > (Attachments contain test files and a jupyter notebook.)
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "wradlib-users" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send
> > an email to wradlib-user...@googlegroups.com
> > <mailto:wradlib-user...@googlegroups.com>.
> > To view this discussion on the web, visit
> >
> https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de
> <https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de>
>
> >

> <https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/wradlib-users/98b6a310-3650-b0ae-4314-39e9699c7ee4%40uni-bonn.de?utm_medium=email&utm_source=footer>>.
>
>
> --
> Kai Muehlbauer
> Institute of Geosciences, Meteorology Section, University of Bonn

> Auf dem Huegel 20 | +49 228 739083 <tel:+49%20228%20739083>
> D-53121 Bonn | kai.mue...@uni-bonn.de

>
> --
> You received this message because you are subscribed to the Google
> Groups "wradlib-users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to wradlib-user...@googlegroups.com
> <mailto:wradlib-user...@googlegroups.com>.
> To view this discussion on the web, visit

> https://groups.google.com/d/msgid/wradlib-users/168fdb65-27fc-4554-aae3-526b3efe153dn%40googlegroups.com
> <https://groups.google.com/d/msgid/wradlib-users/168fdb65-27fc-4554-aae3-526b3efe153dn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Joshua K]

Hi Kai,

I reran my jupyter notebook with wradlib 1.16.2 and the output of the "wgs84" and "sphere" versions is still different.

Shouldn't that have been fixed or is that what you meant by:

> That has some really annoying side effects as the underlying grid is shifted to some extent. You will not be able to directly compare the two grids any more.