CMIP Data Request for Snow on Sea Ice Variables

[anton-seaice]

In v1.2.2.2 of the CMIP7 data request, snow variables are set with either of these cell methods

For sisnmass : area: time: mean where sea_ice over all_area_types

For sisndmass* : area: time: mean where sea_ice (mask=siconc)

however the processing notes do not say anything about applying a weighting (of sea ice area fraction) when averaging over time.

Most places online say these cell_methods mean that an "area weighed mean" should be calculated.

An area weighted mean of sisnmass isn't consistent with the Notz et al. 2016 definition.

So we have saved snow variables in two forms:

1. sisnmass is a normal mean, assumes 0 snow when aice = 0. (This is referred to as "extensive" in Notz et al)
2. sisnmass_intensive, is an area weighted mean ("intensive" in Notz et al)

Ill try and reach out to the SIMIP folk again to get some clarity on this !

[anton-seaice]

This is the link with the change on cell_methods:

CMIP-Data-Request/Harmonised-Public-Consultation#81 (comment)