Help from ACCESS-NRI on libaccessom2 netcdf packing issue for COSIMA ERA-5 runs

COSIMA
1

The effort to develop an ERA-5 forced version of ACCESS-OM2 has stalled due to a number of issues, most importantly this one on libaccess-om2 not dealing properly with netcdf packing in the NCI ERA-5 forcing dataset.

Is this an appropriate issue to get help from ACCESS-NRI on? Is there an appropriate channel for requesting NRI help (I guess a forum post)? Is there someone at the NRI who might have the skills and knowledge to help here? It appears beyond my knowledge of the workings of netcdf packing and libaccessom2.

1 Like
2

No harm in asking I guess. @MartinDix would be the first port of call I imagine.

3

I looked at the libaccessom2 code and things looks to be handled correctly in the (currently reverted) modification to handle packed data. If scale_factor and add_offset are found then they updated at the same time as the data is read.

The only thing I can think of is that there has been some naive concatenation of the files to handle midnight at the end/beginning of the month without taking into account the change in scaling. This could take the form of either directly concatenating an int16 or overflowing /wrapping values which go out of range when repacked with the new values.

1 Like
4

Thanks @russfiedler and @Aidan. The naive concatenation was what I thought might be happening. I’ll add some logging and look into it once Gadi is up again.