search

enram / data-repository

Data quality assessment
https://enram.github.io/data-repository/
MIT License
3 stars 1 forks source link

nominal time versus start time #96

Open adokter opened 4 years ago

adokter commented 4 years ago

see https://github.com/adokter/vol2bird/issues/134 for details

BerendWijers commented 4 years ago

Hi,

I'm also noticing this behavior when processing data from Baltrad.

Prior to running the polar volumes through Vol2bird we do a bunch of checks; after which we rename the pvols based upon our naming convention: CORAD_type_YYYYmmddTHHMM_wmo.h5

Renaming the pvols is done based upon the attribute data of the .h5 file. The nominal time issue then causes our translator to create identical files, as multiple files have identical time blocks.

I've noticed that German polar volumes seem to copy the starttime from the very first scan and use that as their nominal time.

I'm thinking of writing a check (on my end) that detects that a given day has duplicate time blocks and then applies a 'fix' by copying over the start time of the scans to the nominal time of the pvol.

Of course, this feels like a hack and I do not think we should be fixing this from our end. I really do think this is an issue in the generation of Baltrad data. However, until we can have information from Gunther about a possible fix (or exaplanation) I will start using this.

@adokter based on your issue @vol2bird; I've noticed you wanted to use startdate as a nominal time (From a Vertical Profile). I assume vol2bird creates the startdate block. How does vol2bird determine startdate?

bart1 commented 4 years ago

I have noticed similar issues for bejab. it prevents binding into vpts