2017 US AMOC Science Team Meeting Highlights

2017 US AMOC Science Team Meeting Summary

In May 2017, the US AMOC Science Team met in Santa Fe, New Mexico, to discuss progress and evolving science questions, benefiting from almost a decade of coordinated activities. Speakers and posters presented a range of new research findings and highlighted the value from longer observational time series analyses, such as western boundary currents using data from Line W near 40°N and higher resolution simulations down to 1/50° that show the role of submesoscale eddies in circulation.

A primary discussion topic was how the US AMOC Science Team will create a legacy when it sunsets in 2020. Gokhan Danabasoglu, Science Team Chair, shared a proposed set of review or synthesis papers and asked for the Task Teams to consider and expand upon these topics, informed by the meeting discussions. Authorship of the papers will be open to coordination with international colleagues, especially with the UK RAPID program. Volunteers and leads are being identified, with drafting slated to begin this summer and fall. The Executive Committee has set a goal of early 2018 for papers to be submitted.

Another outcome from the meeting was the Executive Committee approving a 2016 Paleo AMOC Workshop recommendation to form a Task Team 5 to better integrate cross-disciplinary collaborations between the paleo and modern oceanographic communities. With acknowledgment of the federal agency paleoceanographic programs to include the participation of their funded PIs, this Task Team will be formed in 2017.

The first special session on the Atlantic Multidecadal Variability (AMV) provided an opportunity to discuss the relative contributions of ocean dynamics and stochastic atmospheric forcing to the creation and driving of the AMV. During the session, two contrasting views were presented and discussed. In one view, it is argued that the AMV is a direct response to the North Atlantic sea surface temperatures (SST) to atmospheric stochastic forcing and/or external radiative forcings because the SST anomaly pattern associated with the AMV can be largely reproduced by a slab ocean model, implying that the ocean dynamics do not play a major role in creating the AMV. In contrast, the second view argues for a prominent role for ocean dynamics. Specifically, this argument suggests that the observed AMV includes coherent multidecadal variations in Atlantic SST and salinity, upper ocean heat and salt contents, and ocean-driven surface turbulent heat fluxes. These key observed features cannot be explained as direct responses to stochastic atmospheric forcing and/or external radiative forcings but are, instead, consistent with an important role for ocean dynamics. The amplitude of the AMV and associated multidecadal atmospheric variations, as well as its connection with the AMOC variability, are often underestimated in coupled climate model simulations, likely due to biases in models’ mean state and in representations of modes of variability. Vivid and interesting discussions appreciated both views and generated ideas for future experiments that could help reconcile the competing conclusions. 

The second special session was on AMOC stability metrics, which provided an opportunity to review the state of knowledge and discuss new results regarding the role of AMOC-induced freshwater flux into the Atlantic (mostly across its southern boundary at 35°S) on the stability of the AMOC. Commonly referred to as Fov, this freshwater flux is thought to be an indicator of whether the AMOC is in a mono-stable state (hence stable to finite-amplitude perturbations) or in a bi-stable state (hence prone to collapse, if perturbed strongly enough). The accuracy of this metric in models is rather surprising given that its fundamentals are captured by simple box models. One presentation discussed to what extent the assumptions in these box models carry over to comprehensive circulation models. It was suggested that the stability of the AMOC in the current generation of coupled climate models may be overestimated, due to common biases in the salinity field of the South Atlantic. A flux-correction was suggested as a specific solution to reduce this bias and make the stability of the AMOC in climate models more realistic. But the discussion naturally turned to the actual causes of the salinity bias, as biases in tropical precipitation were cited, as well as misrepresentation of Agulhas Leakage. It is generally thought that the move towards higher resolution models will reduce some of the biases seen in low-resolution models, and several studies were presented that probe the AMOC response to freshwater forcing in models that explicitly represent eddies. Indeed, new results showed that the AMOC is prone to collapse in an eddy-permitting model, suggesting that spatial resolution may affect the processes that influence the stability of the AMOC, although it is too early to say why.

The meeting was noteworthy for its interactions among the 80 participants and wealth of new information, as captured in the oral presentations, posters, and tweets. Using overarching questions established before the meeting, the Task Team leads integrated these into plenary and breakout discussions. They also built their discussions on near- and long-term priorities and recent activities. Below are summaries from each Task Team.  

 

Task Team 1: AMOC observing system implementation and evaluation

The Task Team 1 (TT1) session highlighted the influence of eddies and mesoscale variability and how these features extend across vertical layers. Presenters also showed how meanders in the Gulf Stream influence the exchange between the Deep Western Boundary Current and the interior, and how variability in the North Atlantic Current over Charlie Gibbs Fracture Zone influences the westward outflow of Iceland-Scotland Overflow Water through the fracture zone. At the 16°S array, deep eddies exhibit a strong influence on the total transport across the line. Furthermore, high-resolution model studies are used to assess the impact of resolving eddies on ocean circulation, but more work is clearly needed.

One of the TT1 priorities is to improve understanding of the meridional coherence (and/or lack thereof) of the AMOC and the mechanisms that control AMOC changes. This remains an important near-term priority. Latitudinal coherence in the observational time series is low, partly due to the limitations of record lengths and likely also due to low signal-to-noise ratio (noise being the local mesoscale variability) of the anomalies that might be communicated between the arrays. There is some coherence between the RAPID-MOCHA and MOVE arrays, both of which are located in the North Atlantic subtropical gyre. This coherence is also evidenced by the GRACE bottom pressure data. Wind forcing appears to be the mechanism of variability in the North Atlantic subtropical gyre, while buoyancy forcing is suggested as the predominant cause of variability in the North Atlantic subpolar gyre. In model studies, the variability in density of the Labrador Sea Water is correlated with AMOC strength on longer timescales, but it is not clear yet how this will translate into signals observed at different latitudes. At the southern boundary of the Atlantic Ocean, wind forcing and Agulhas leakage both seem to play a role in transport variability.

The continuation of the existing observing arrays remains a topic of concern within TT1, as many of these projects are currently seeking renewal of funding. Even though the US AMOC Science Team is planned to sunset in 2020, the community hopes that these time series will be continued. This will depend on the willingness of PIs to propose a continuation of their projects, demonstrating their value, as well as the favorable review of the proposals and their selection by funding agencies for continued support. Better outreach to show the necessity of these observations, both within the (non-observing) community and to society in general, will help to sustain these observations into the future.

TT1 discussed the possible review/synthesis papers, and the following topics emerged:

• A review/synthesis paper highlighting the US AMOC observations, why they were deployed, what has been learned about the AMOC, and best practices in AMOC observations.

• A review/synthesis paper showing what has been learned from the observations about meridional coherence, using the multiple long time series of the Deep Western Boundary Current as one example.

• A review/synthesis paper looking into the future, such as which new technologies and/or novel combinations of technologies and data can be used for sustained AMOC observations.

Several names for lead authors were put forward during the discussion and these will be finalized in the coming weeks. The topic of the second paper may also be addressed in collaboration with Task Team 2. It is a goal that these papers can be used to provide input for the OceanObs’19 Conference.  

New action items:

1. Investigate how to make data more accessible. The US AMOC Science Team now hosts a webpage with a collection of AMOC time series. OceanSITES is also considering options for making datasets easier to access. While these resources are good options for the AMOC community, they are not always the best resource for other communities (e.g., non-academic) who may find the data useful.

2. Advocate for the available observations to be used more widely by the modeling and reanalysis communities. The use of observations in these studies will be beneficial for two reasons. First, it will increase understanding of AMOC variability and coherence across multiple observing systems. Second, the use and citation of these datasets by the wider community will help PIs to make the case to sustain these observations.

3. Investigate possible new data products (e.g., density fields) to help achieve action items 1 and 2. These data products would need to be relatively easy to produce across arrays and would be an addition to the currently provided transport time series (and their uncertainties).

4. Finalize synthesis paper topics and authors.

  

Task Team 2: AMOC state, variability, and change

The focus of Task Team 2 (TT2) is to identify common patterns and mechanisms to explain the present AMOC circulation state, its high-frequency variability, and long-term change. As each individual research study contains these timescales, progress comes from finding similarities between multiple observational and modeling studies. Therefore, the work of TT2 attempts to integrate across multiple studies. The discussion and breakout sessions identified the following commonalities and discrepancies.

Consistent AMOC signals:

• Wind input of momentum is consistent with observations and models in the subtropical gyre in driving sub-annual variability.

• Some internal flow components are consistent for the deep limb of AMOC, but only a few observed events exist to rely upon for interannual variability.

• Meridional coherence is consistent within gyres from observations and models.

• A horse-shoe shaped SST pattern for Atlantic multidecadal variability is common in models and observations for decadal periods.
   

Inconsistent AMOC signals:

• The driver of overturning in the subpolar gyre — wind versus buoyancy forcing — is not consistent between different models and observations, in particular, surface buoyancy forcing. The pending OSNAP results will help, but this points to a less-than-complete understanding of the small-scale processes at high-latitudes and how they relate to their large-scale consequences.

• The depth extent of the AMOC in models and observations is inconsistent, in part because models (especially at lower resolution) poorly resolve overflows and regions of deep convection, and in part because calculations are inconsistent (in depth or density space, depending on observational/model approach).

• There is inconsistency in the amount of variability between numerical models and observations (generally smaller variability in models). Further inconsistencies exist in long-term trends both among numerical models and observations. Observational studies can be constrained by how much of the basin the measurements cover and what assumptions (e.g., geostrophy) feed into the computations of AMOC.

• Coherence across gyres is not consistently represented in observations and models. Some models have identified basin-wide planetary wave modes as an important means of interannual variability, but their exact patterns, timing, and forcing have yet to be described consistently.

• The timescales of variability across different models are not robust, even though model parameters might be robust and/or large-scale mean AMOC pattern might be consistent.

• In the South Atlantic, numerical models and observations have mismatches for geostrophic flow. It remains to be seen whether this is due to misrepresentations in the simulations or inadequate knowledge of the reference level for the geostrophy in the observations. 

These findings in the AMOC signals lend themselves to being focal points of future studies that directly address TT2’s goals. There was a consensus that metrics are needed that can be consistently assessed across multiple data sources, both from numerical simulations and observations. Quantifying uncertainties will also aid comparison. The scientific community is most interested in transport of quantities (heat, freshwater, momentum, carbon, nutrients) that influence the climate and interact with the atmosphere. This highlights the conceptual mismatch between a focus on transport, and the kinematics that drive dynamics, and the integrated nature of temperature/heat content and salinity/freshwater. This mismatch also relates to describing the ocean in a Eulerian framework versus a Lagrangian one. Flow pathways and mixing are Lagrangian, whereas overturning strength or latitudinal coherence are implicitly Eulerian. Throughput of AMOC and the connection between gyres where dynamical balances differ requires understanding from each framework. Related outstanding questions are the nature of coherence in the upper AMOC branch, how this relates to observational coverage (well resolved upper ocean circulation and temperature/heat versus poorly sampled deep ocean and salinity/freshwater), the impact of eddies and indirect (and unknown) water parcel pathways, and the resulting transfer between gyres.

The TT2’s three guiding questions for the meeting provide a useful framework for review or synthesis papers, recognizing much overlap with other Task Teams, especially with TT1 and TT3. However, because the question of coherence between observations is an active research topic without many results, it is premature to include such efforts as part of the current set of review/synthesis papers. The two remaining questions for possible review/synthesis papers are:

TT2 will focus on the following priorities for the remainder of the US AMOC Science Team program:

• Use new and existing observations in combination with modeling experiments to refine the understanding of the present and historical circulation (and related transports of heat and freshwater as well as flow pathways) in the North and South Atlantic.

• Observational studies should focus on mechanisms and pathways that identify and explain coherent and incoherent signals between different study sites, thereby reaching consensus on which signals represent the large-scale AMOC versus more localized circulation patterns.

• Synthesize modeling and observational evidence to build scientific consensus on the variability and change of the AMOC over the last 50 years, using observable proxies as appropriate. This could be integrated with work to be done by TT5. Efforts within the data assimilation community should focus on reaching an accurate consensus (consistent with other lines of observational evidence) on the evolution of the AMOC over the last 50 years.

  

Task Team 3: AMOC variability mechanisms and predictability 

Task Team 3 (TT3) focuses on the issues of AMOC mechanisms and predictability. The meeting covered a wide range of topics, and most of the research made use of either forced ocean or fully coupled climate models, with more presentations slanted towards mechanisms than predictability. Timescales of interest spanned interannual-to-multidecadal and model resolutions ranged from several degrees to 1/50°.

The breakout session covered several topics. First, TT3 revisited the webinar series that was carried out during the 2016/2017 academic year. It was widely viewed as a success, and everyone agreed to continue starting in fall 2017. The format of 30 minutes total time (15-minute presentations, 15-minute Q&A) will remain and topics are open to anything that falls under the general responsibility of TT3. Once details of the time are settled, the schedule will be sent to all members of US AMOC Science Team for their participation. The goal is to have approximately one presentation per month.

TT3 also discussed potential topics for synthesis/review papers. Two were identified that had both people willing to take the lead and sufficient progress made in recent years to warrant such an effort.

• A review/synthesis paper on AMOC stability, particularly focusing on the role of freshwater transport across 30° S.
• A review/synthesis paper on high-resolution models of the AMOC, including regional studies of key components such as the Southern Ocean, Gulf Stream, and high-latitude convection regions.

The near-term priorities were reviewed for both continued relevance and to assess progress. All seven were deemed to remain important and appropriate for the life of the Science Team. Currently, funded US AMOC projects are addressing at least parts of each priority. Based on the special science session on the AMV, an additional near-term priority was added to address a better understanding of the relative influences of white noise atmospheric forcing and ocean advection in controlling the low-frequency variability of SST and air-sea fluxes of heat, freshwater, and momentum.

The three long-term priorities were reviewed with no changes recommended. Progress is being made on all three.

  

Task Team 4: Role of AMOC in global climate and ecosystems

The focus of Task Team 4 (TT4) is the relationship between the AMOC and other components of the Earth system. During the meeting, there were two sessions devoted to interactions between the AMOC and other components of the Earth system, including a special session to understand the relationship between the AMOC and the AMV. There was vigorous debate regarding the relative roles of external forcing (e.g., aerosols), atmospheric forcing, and ocean dynamics in the AMV.

While TT4 has traditionally been focused on AMOC impacts, an improved understanding of the two-way nature of the climate system, as well as debate regarding causality (e.g., the origins of the AMV), led to a recast of TT4 priorities in terms of interactions between various components of the climate system and the AMOC. For example, TT4 will seek to understand interactions between the AMOC and sea surface temperature (SST), global and regional sea level changes, and the cryosphere, rather than just the impact of the AMOC on these components of the climate system. 

During the TT4 breakout, members discussed the near-term priorities in detail and found some of them to be too general. TT4 will now focus on specific interactions between the AMOC and other components of the climate system that are well-documented but where understanding is incomplete, thereby making these topics ripe for advancement. Three specific topics were isolated:

• The impact of the AMOC on the Inter-Tropical Convergence Zone (ITCZ).

• The interaction between the AMOC and the hydrological cycle, including clouds.

• The relationship between the AMOC and climate extremes, including hurricanes, droughts, and flooding events.

TT4 also modified the long-term priorities to include the teleconnections of AMOC variability with variability in other ocean basins (e.g., Southern Ocean). Additionally, TT4 would like to engage the paleoclimate community in order to understand the impacts of the AMOC on centennial and longer timescales, taking advantage of the establishment of a new Task Team on this topic.  

During breakout discussions, TT4 identified both a list of four topics appropriate for review/synthesis papers (three could be review papers and one a science paper) and volunteers to lead each effort. These topics are:

• A review/synthesis paper on the linkage between the AMOC variability and the AMV and associated impacts, including both modern and paleo observational linkages.

• A review/synthesis paper on the relationships between the AMOC and sea level changes.

• A review/synthesis paper on the impact of the AMOC on ocean tracers.

• A new scientific paper on “Are the recent temperature changes off Greenland related to the AMOC?”