ACCESS-AM2 Sea ice concentration ancillary dataset help

Atmosphere
, , , ,
1

Hi,

I’ve been looking at the sea ice concentration ancillary dataset for ACCESS-AM2 (suite u-cz934) found here:
g/data/access/TIDS/CMIP6_ANCIL/data/ancils/amip_fix/timeseries_1870-2018/SstSeaIce/seaice_amip_n96e.anc

Although this file is supposedly sea ice fraction (between 0-1), I’ve found that only the last year (2018) has values within 0-1; there seems to be values less than -1 and greater than 1 so it has me a bit confused.

I was wondering if anyone knows where/how this ancil file is generated and why there are values outside the range 0-1?

For example:

image
image796×667 43.9 KB

2

AMIP sst and sea ice are adjusted so that the monthly means match a prescribed dataset - AMIP Details index. Do the monthly mean values look more realistic?

3

Thanks Scott!
Just checking, do you mean the monthly model outputs for sea ice concentration? They do !

4

Hi, just letting you know this question is outside the scope of the ACCESS-NRI support as the ancillary file you described is not one that is created or supported by ACCESS-NRI.

With that said, since this is related to ancillary files of the UM perhaps @inh599 or @MartinDix may be of help?

5

@alanah.chapman @MartinDix is the better person to respond here.

However - my initial take would be that the ‘odd’ values are (likely) a consequence of the regrid/interpolation step needed to get from the source sea-ice concentration data onto the land-sea mask that AM2 uses. This can be sensitive to detail of the interpolation and the fill value. I know we had similar issues with coastal SSTs because of mismatches in the masks in our early CMIP6-AM2 runs.

1 Like