Re: [las_users] how to set up an interface for 6-hourly data

[Martin Schmidt <martin.schmidt@xxxxxxxxxxxxxxxxx>]

Hi Roland,

we did so and I have got an (almost) sufficient result. THREDDS 
aggregation works and several data sets could be included.
More than 16000 time slices are glued correctly.
Many thanks for this nice tool. You may have a look at the final result 
visiting

http://phy-51.io-warnemuende.de/

The order and the names of some variables needs to be improved, but this 
will be
done when other things are ready.

There remain three major questions:
1. "expressions" does not work.
2. analysis does not work for data set "GENUS Atlantic model surface 
fluxes", but only for
some of the fields. For those fields see also 3.
3. Some fields in "GENUS Atlantic model surface fluxes" are displayed 
with wrong levels.
for example the long wave heat flux..
The data span is indicated correctly at the colour key, but the level 
span is set to zero.
Setting the levels by hand as plot options helps. There may be a "work 
around" using
"properties" to set initial levels, however, there seems to be some 
minor bug in the code.

Using the opendap access directly with ferret everything is fine for 
this data set.

The last data set "HIX" is experimental because it contains time series 
at a fixed point. I did not find out
how to remove the lon-time or lat-time plot option. I made a special 
ui.xml, and tried to
remove all 2D stuff from the timeline default. To get it running I had 
also to modify the lat/lon axes to cover
some rectangular area with the point in the middle where the data are valid.

Do you have any suggestions how to improve this ?

Many thanks,
Martin



Roland Schweitzer wrote:
> Martin,
>
> The short answer to your question is to up-grade to the latest 
> release.  The release of the UI client you are using did not work 
> correctly with intervals smaller than a day.
>
> Roland