Dry deposition to snow and ice

[r-pound]

Name and Institution (Required)

Name: Ryan Pound
Institution: University of York

Confirm you have reviewed the following documentation

• [Y] Contributing guidelines

Describe the update

With the changes made in 13.4.0 (commit e593ccb, pull request #945) the method of calculating array IsSnow and other majority land cover types for grid boxes changed to use the native GEOS meteorological fields.

The result of switching from IsSnow being determined by surface albedo to the fraction of land covered with snow (FRSNO) has changed the behaviour of this array. Previously (due to surface albedo) this would be true for any grid box containing a majority of snow, land ice or sea ice.

This is no longer the case and only represents grid boxes with a majority of snow cover. Several sections of drydep_mod.f90 have historically used the IsSnow array to update dry deposition calculations to account for special cases of dry deposition to snow/ice. However since the land cover arrays were updated to use the native fields, these special case determinations have not been updated to reflect the current state of these arrays, apart from Update drydep code for assigning pH over land, ocean, and ice (pull request #1435) which updated the code to reflect the need to use both IsSnow and IsIce to determine if the tile is a majority snow or ice cover.

Expected changes

For the dry deposition of ozone, this restores dry deposition velocity calculation behaviour to that expected from #997. This increases surface ozone over predominantly ice and snow-covered regions, with everywhere else experiencing negligible change.

This also results in similar changes for other species, however, this should be reverting model behaviour back to that of pre-version 13.4.0 where IsSnow was calculated based on surface albedo.

I also recommend the inclusion of change in dry deposition plots within the benchmarks. I can supply the Python script used to generate this analysis if that would help.

[msulprizio]

Thanks @r-pound. I have added this bug fix to our model development priority page for inclusion in 14.3.0.

Also tagging Surface-Atmosphere Exchange Working Group Co-Chairs @jgeddes85 @cdholmes @amospktai

[lizziel]

Given this section treats snow and ice the same I think we should update this comment to be "Special treatment for snow and ice". Having "vs" makes it look like they are treated differently.

[r-pound]

Agree. The comment felt a bit ambiguous to me so I left it as is in case anyone more knowledgeable about this specific case wanted to suggest a different approach. Based on how IsSnow used to be defined it will have always treated snow and ice the same until 13.4 so my gut feeling is that it should read

Special treatment for snow and ice

as you suggested

[msulprizio]

This comment is now fixed in c23b1ef.

[lizziel]

Good catch that the pre-13.4 version of IsSnow included ice and thus there were parts of drydep_mod.F90 that needed updating but weren't. Thanks for finding this!