BGC Parallel Session Meeting notes

Date and Time

Nov 19, 2019: 3:30 to 4:15 ET

Nov 20, 2019 3:15 to 4:15 ET

Call Info

None. In person meeting.

Attendees

@Katherine Calvin (Unlicensed)

Goals

Discussion items

Time	Item	Who	Notes

Time	Item	Who	Notes
3:30-3:35 (Tuesday)	Introduction	@Katherine Calvin (Unlicensed)	Goals for the session
3:35-3:55	v2 Land Model Updates (LBNL stoichiometry, photosynthesis, allocation)	@Qing Zhu
3:55-4:15	Ocean model: v1 biases and v2 planning	@Nicole Jeffery	Biases v2 improvements Spinup
3:15-3:35 (Wednesday)	v2 Scenarios	@Katherine Calvin (Unlicensed)
3:35-3:55	Discussion: what worked and didn’t work for v1	@Susannah Burrows	What worked especially well during the v1 campaign, in terms of organization, process, communication, or technical approach? Were there any problems that occurred during the v1 simulation, which we could potentially avoid next time by doing something differently? Was the system for organizing the simulations useful (Simulation overview page, POC for each simulation, verification spreadsheet, approach to running diagnostic packages)? Is there anything we might want to consider changing about this approach for the v2 campaign? How was your experience with zstash? Are there any new challenges we anticipate for the v2 campaign? Any other comments / concerns?
3:55-4:15	Discussion: looking forward to v3		v3 will include atmospheric chemistry in the BGC configuration, and also coupling of GCAM surface emissions (e.g, CO2, BC/OC, SO2), which will open up new opportunities to examine feedbacks involving atmospheric constituents (greenhouse gases, aerosols, atmospheric oxidants) Are there gaps in v3 process coupling (not yet being addressed)? Coupling crops with GCAM and with atmospheric chemistry (NH3 emissions)? Coupling fire model with BC/OC emissions? Any other potential gaps? If there are gaps, what can be done about this? Are there science questions that will be enabled by these v3 features, which might be “low-hanging fruit” to address? Are any early planning steps needed to make it possible to address those questions? Are there important opportunities to address science questions with these new capabilities, which may be outside of our project’s scope, but where we could work towards getting them addressed through collaborations with external projects or partners? Is there analysis that could be done on the v2 model that would help with planning for v3 (e.g., evaluation of methane fluxes?) What near-term or long-term strategies could we pursue to build credibility and visibility for BGC capabilities and activities in E3SM? Any other comments / concerns?

What Actually Happened

Tuesday notes from @Beth Drewniak

Susannah has questions for v3 simulations to answer by tomorrow breakout.

Qing:

Changes in environment = changes in allocation and stoichiometry

Allocation based on resource availability

Flexible C:N:P stoichiometry

Little field data to constrain, but good understanding of relationships

Tropics can uptake more carbon through more efficient phosphorus use.

Discussion:

Oak Ridge and Duke face experiments compared with other models, no consistent result existed between sites. No agreement for static vs. dynamic for allocation/stoichiometry. How does this model fit into that uncertainty?

Only two sites, need more data, not just elevated CO2, but other perturbations. Two sites aren’t convincing for constraining models. Other dynamic allocation schemes exist, we should do parallel testing. We should add to models to test this.

Ocean thinks of stoichiometry to optimize growth, does the database give estimates of this?

We use data to constrain data, want plants to adjust under changes. WE have limited knowledge what should be good target: reproduction, other issues? It’s not clear

Amazon swing seemed less than allowed?

Current vs. future conditions are stronger.

Any molecular biology evidence for carbon adjustment?

There is strong evidence for that, depending (for example) daily timescale

Allometry depends on season. Other models apply other approaches for optimization?

Game theory, etc. This is designed for a longer time scale.

Phenological stages: heat stress in crops have effect and nitrogen effect.

Nicole: Slides for discussion

V1 BGC simulations

Bias 1: Low surface nutrient biases, arctic is strong, but in other areas as well. Ocean and sea ice underproductive

Bias 2: Extensive Labridor sea ice

Bias 3: river nutrients are low

Bias 4: ocean chlorophyll is low, important for ocean sink of CO2

Dissolved oxygen is too low

Unrealistically shallow mixed layers in southern ocean

High resolution does better

Working to improve the these

River nutrients: Virtual fluxes in POP, that minimized this. Freshwater inputs cause problems. Nutrient fluxes with MOSART is needed, might not be reasonable in arctic region (MOSART can’t handle ice). Alternatives? Not right now.

Export production? We want it, but can’t get it. Matt can calculate particulate, it’s low

Will any approaches help with arctic intermediate water simulated? Hope so, could use help. Small changes in KPP. Lots of physics issues will help with biogeochemistry. (High resolution near coastal regions could help). Expecting new grid that hasn’t seen biogeochemistry yet.

Need to fix sea ice nutrients. Nicole added N mineralization in sea ice.

E3SM Conferences and Meetings