2015-11-04 Speed Dating / A-L: Atmosphere-Land

Notes by Forrest M. Hoffman (Unlicensed), November 4, 2015 at 8:40 a.m.

David C. Bader mentioned an issue of something being deposited on land that it was not epxecting. This may be the aerosol deposition on snow issue that arose previously. That issue was due to a known bug in the land model. Hailong Wang has been changing the snow code in the land and committed code, but the Land Group is not aware of this work. David C. Bader wants to know how the disconnect occurred. The aerosol development was in the proposal, but was not communicated across the Groups. Peter Thornton suggested using the Code Review Process so that it appears in the table and lists the components or features that are affected by the feature being developed. Gautam Bisht suggested also using Request Hub when appropriate to reach out the Land Group and to create an Issue if there are bugs that need to be fixed on land.

When should BGC coupled simulations begin?

Peter Thornton and William Riley (Unlicensed) asked about if the delay in BGC simulations was agreed upon, or is it a rumor that is propagating through Confluence? Phil Rasch (pnl.gov) believes it is easier to configure and tune the low resolution model. There is concern, brought up by Chris Golaz, that the atmosphere model may have to be tuned separately for the physical model than for the BGC configuration. Peter Thornton said we want a model tuned to work well in the BGC configuration, and the Land Group may soon ask that the Atmosphere Group to use the carbon cycle model, then it could be used for all the tuning to keep the number of atmosphere configurations small. Phil Rasch (pnl.gov) needs a recommendation about what configuration to use. William Riley (Unlicensed) suggested the model code will be there for January.

Philip Cameron-Smith (Unlicensed) asked if the active carbon cycle would be used for the water cycle experiments. Peter Thornton said this could be a possibility and could keep the number of configurations small.

Phil Rasch (pnl.gov) would be most satisfied if the Land Group could provide a prototype, which might change in January, to work with for calibration and tuning.

Ruby Leung asked if there would be a spun-up land carbon cycle at high resolution. Peter Thornton suggested the Land Group could provide something that would work better than SP mode.

Peter Thornton suggested sticking with SP mode for now, but it may change by January.

Ruby Leung asked if transient land use change was in the model. Peter Thornton said yes, and the Land Group could spin up to whatever date is desired.

Philip Cameron-Smith (Unlicensed) asked if the coupling could be changed to speed up the simulations. At high resolution, the timestep may get down to 10 min. Peter Thornton indicated this might be acceptable.  Phil Rasch (pnl.gov) suggested deferring this topic for future discussions.

Phil Rasch (pnl.gov) brought up data sets that are trusted by the Atmosphere Group. And these data sets are now in netCDF with appropriate metadata, and Phil Rasch (pnl.gov) would like to discuss these with the Land Group.  Xubin Zeng said it would be nice to have the same data sets for forcing the land model and evaluating the atmosphere model. Shaocheng Xie suggested we need a Confluence page and a future conference call. Phil Rasch (pnl.gov) suggested we form a group to discuss this that includes a few people from the Atmosphere Group, William Riley (Unlicensed), Peter Thornton, Forrest M. Hoffman (Unlicensed), Ruby Leung, and any other interested people.

Xubin Zeng discussed previous difficulties in coupling land and atmosphere component models and suggested considering a combined solver. Chris Golaz said they had one at GFDL and it is quite complicated. Is it worth the difficulties? William Riley (Unlicensed) suggested possibly doing an interactive canopy that interfaces with the atmosphere and land models.

Completed at 9:20 a.m.