Observations indicate that the ocean is losing oxygen and models suggest that this will accelerate under persistent climate warming. Current indications are, however, that CMIP5 models under-estimate the magnitude of ocean oxygen loss relative to our best observational estimates.
Oxygen loss as a function of temperature change between 1970 and 2014.
The models suggest that warming-driven reductions in solubility are reinforced by increases in AOU in the extra-tropics, leading to large oxygen loss and relatively robust multi-model prediction. Oxygen loss in the tropics is more complicated, however, as changes in tropical ventilation lead to reductions in AOU that partially or wholly compensate for the direct effects of warming on solubility. The competition between these drivers yields divergent predictions across the models and contributes to ambiguous results at the global scale. Tropical ventilation, however, is poorly simulated in many (most?) models, so it is difficult to assess confidence in these results.
Proposed Hacking
This project will aim for as near-comprehensive characterization of CMIP6 oxygen projections as possible in the 3-day timeframe. We will seek to produce compelling visualizations of trends and phenomenology and aim to bundle a collection of Jupyter Notebooks into an accessible, reproducible computational narrative.
Here are some example research questions that will guide the analysis.
How well do the models simulate the mean state and observed trends in upper ocean dissolved oxygen?
What are the trajectories of future ocean oxygen loss and how do these manifest regionally from solubility, circulation, and oxygen utilization rate (OUR) drivers?
What are the projected change in hypoxic and anoxic volumes globally? How much will the vertical extent of viable pelagic habitat be compressed by thermocline O2 loss?
Can we explain AOU-driven changes in tropical oxygen loss in terms of ventilation dynamics? Do we expect that these dynamics are well simulated by the models?
What is the value of climate mitigation to avoiding impacts of oxygen loss?
Is it possible to develop emergent constraints or other clever means of reducing uncertainty in future oxygen projections via exploiting interannual to decadal variability, for instance? Are there relationships between simulated decadal variability and the forced response?
Additional fields to look at ventilation question…
Anticipated Software Tools
We will use the standard suite of Python analysis tools: xarray, dask, etc.
We will rely on packages like xgcm, esmlab, etc. to enable concise expressions for operations like integrals, etc. It is possible shortcomings in these packages will need to be addressed to enable writing elegant, performant, exemplary notebooks.
Desired Collaborators
People with interest and expertise in ocean biogeochemistry and tropical ocean circulation.
Looks like a very interesting and well-posed problem. I’m stunned to see that the observed relationship is completely outside the simulated one. It’ll be interesting to see whether CMIP6 models are any closer. Tempted to join despite this not being my area of expertise.
Following on the ventilation dynamics angle, it would be interesting to evaluate the tropical O2 response in the context of past and future changes in the coupled ocean-atmosphere response in this region, i.e. changes in surface winds, SST gradient, upwelling, and the equatorial current system characteristics, and links to the extra-tropics. Ocean circulation fields (u,v,w, stream function) and atmospheric surface winds, wind stress, SST, and air-sea O2 flux would be other helpful variables that other groups/projects may be using as well.
This is a nicely-scoped proposal, and one I’d be interested in contributing to. Probably good to be careful about expanding the scope too far, but it would be great to include model validation against oxygen observations (e.g. WOA).
A logistical question I have (not related only to this project) is how you’re envisioning collaboration working across the different sites. Would it make most sense for folks to work in project groups located at a single site, or do you think it would work to have people working on this project from multiple locations?
This project looks very interesting. I would be interested in contributing to it. Do you have a sense for whether the difference between models and observations relates to changes in biological community composition and reaction rates and opposed to representation of the flux of a passive (but non-reactive tracer)?
This is a great project idea and I am extremely interested in joining this project. In particular, I am curious about the representation of the Equatorial Current system and the Equatorial Undercurrent between the different models, something I have looked at recently for several CMIP5 models. @yassir.eddebbar summarized the variables needed for that nicely. Are full oxygen budgets available for any of the models?
From a technical standpoint I agree with @matt-long:
We will rely on packages like xgcm, esmlab, etc. to enable concise expressions for operations like integrals, etc. It is possible shortcomings in these packages will need to be addressed to enable writing elegant, performant, exemplary notebooks.
One of the major challenges of a project like CMIP6 is that the various models provided have different grid architectures, both in the horizontal and vertical. In order to analyze the output between different models we require tools that can calculate properties on the native grid of the model and avoid interpolation and associated errors. The syntax of these tools should be as „agnostic“ as possible so that common operations like a gradient or divergence operator can be applied with a simple syntax akin to an analytical equation. The proposed tools would detect if e.g. the velocities are located on a B or C-grid and then execute the numerical calculations as close to the internal model code as possible.
I would propose the implementation of common vector calculus operators like grad, div, curl into the xgcm module. Recent addition of automatic grid metric handling builds the foundation for these operators and this workshop presents a fantastic opportunity to develop these new function on top while simultaneously testing them with a variety of model outputs.
I am unsure if I should propose this as a seperate project, since it will surely benefit other proposals as well. @matt-long, @rabernat, do you think it would make sense to split out the more techincal aspects?
Hey everyone, I just posted in @rabernat’s project, that I would be really interested in looking at the Equatorial Undercurrent in the different models.
Our recent paper suggests that the EUC can influence the variability of the upper Oxygen Minimum Zone in the eastern Pacific, thus I thought it might be beneficial to interface with this project on that.
