Yvonne Firing (Scripps, USA)
Louise Newman (SOOS)
The first Southern Ocean section in the Bulletin of the American Metrological Society (BAMS) State of the Climate (SOC) was published in 2014 (Meredith et al. 2015). The task of the assessment was to put observations collected during 2014 in the context of Southern Ocean climatology. Compiling this first assesment presented numerous challenges arising both from the brevity required and from the sparse nature of most Southern Ocean observations. In response a SOOS-supported workshop was convened in June 2015 to gather community consensus on content and methods for future State of the Climate reports. The workshop was attended by ~50 people and consisted of a combination of invited speakers and panel discussion. All of the presentations and the panel discussion were recorded and can be watched on the SOOS YouTube Channel. Here we summarise the key ideas from the workshop across a number of themes that emerged from presentations and discussions, including:
- Importance of the Southern Ocean to global climate
- What we can currently assess and where are the most significant gaps
- What observations are most relevant to assessing the climate state
- Recommendations for climate assessment
Connecting the global oceans and being a primary site for communication between the atmosphere and the deep ocean, the Southern Ocean is a key component of the climate system. It is because of its place in the climate system that there is a need to understand its basic state, the processes that control that state, and how these are changing with time. As recognized by the Intergovermental Panel on Climate Change (IPCC), a poor understanding of processes and how they change in the Southern Ocean can propagate through global assessments of Earth’s climate, particularly when making long term climate projections.
The ability to assess the state of any component of the global climate system is highly dependent upon available observations. In order to provide a basis for determining change, these observations need to be both within the desired assessment period and comparable to a high quality time series. In parts of the Southern Ocean both these criteria are met, but in regions where observations are more sparse, particularly in the ice covered coastal seas, determining the difference between change and intrinsic variability is more challenging.
Historically, the Southern Ocean has been underobserved due to its size and remoteness. The Southern Ocean Observing System (SOOS) is seeking to rectify this situation, but addressing all of the goals for rigorously quantifying the Southern Ocean Climate remains a challenge. These goals include: (i) accounting for and gaining maximum information from all the relevant measurements; (ii) quantifying uncertainties from both measurements and inherent natural variability; and (iii) identifying all of the gaps in our knowledge. These points were noted frequently in the workshop. It was also emphasized that without understanding the governing dynamics of the Southern Ocean it is difficult to prioritize observations required for assessing climate change.
Several of the challenges relating to Southern Ocean observation systems were discussed, including the fact that many of the observational records in the Southern Ocean are short and may be sampled too infrequently. The resulting aliasing can further complicate analysis. Figure 1 summarises the time scales that different methods and instruments are able to measure, and the time scales they may alias.
Figure 1: Summary of time scales measured by different ongoing or recent measurement systems/observational programes in Drake Passage.
When this is combined with the sparse coverage of observations under sea ice, and the almost complete lack of observational coverage under ice shelves, there are concerns about the representativeness of the available data on the Southern Ocean. This problem can be compounded when interannual variability is neglected in forming climatologies. However, it may be possible to draw on long records from relatively quiet regions, such as the deep ocean, in order to detect change.
Good assessments of the state of the climate depend on adequate observations. Individually the observations need to be accurate and reliable. However, in order to produce low uncertainty in climate change assesment they also must be relateable to long time series. The time series length must be longer when the property has long decorrelation time scales or high natural variability (Leroy et al 2008). These criteria were mirrored within the workshop through the concepts of signal to noise ratios and power spectra of various climate-relevant properties.
Several common data sets were discussed in the workshop and their advantages noted. For example, satellite data was noted for its long record lengths, Argo for its lack of seasonal biases compared with ship observations, and seal data for providing a key data source from under sea ice covered regions. Hydrographic sections, while infrequent, provide data in the deep ocean, a region that would otherwise go unmeasured. Their high accuracy also provides a standard against which other platforms can be compared.
Given the expense of working in the Southern Ocean and the power and weight budgets of instrumentation - particularly animal-borne sensors - a cost-benefit analysis including a variety of aspects is always likely to be required when prioritizing observing programmes. For example, adding dissolved oxygen sensors to seal Conductivity-Temperature-Depth tags would provide additional data but reduce sampling life, and where dissolved oxygen is tightly linked to temperature would provide little additional knowledge.
The on-going collection of observations is important for determining the state of the climate in future years. The design and prioritization of observations, however, is dependent upon the aims of the SOC assessment and the metrics through which this is best achieved. This is a question of understanding both the observations and the metrics: is the aim to be able to make a statement on how the Southern Ocean is changing, or is it to provide the mechanics for detection and attribution of change? In the workshop no final answer was given for this question.
The workshop did yield some clear pointers as to which observations we need to continue, and why. We need to continue observing deep ocean sections, as they provide the clear baselines that are needed for SOC assessments. Argo and satellite observations will soon cover much of the rest of the Southern Ocean. As successive years make these records longer, the issue of having short climatologies for each of these these records will resolve itself, although in the short term it will continue to be a problem. When it comes to the Southern Ocean, using a combination of observational methods and working on a long term timeframe is the best approach.
Workshop participants raised several areas where oberservational improvements could be achieved with relative ease. These include working to increase the number of observations - particularly of air-sea flux - from vessels of opportunity; trying to better resolve bathymetry, particularly in Antarctica’s coastal seas; and, most crucially, collecting observations in the ice-covered seas. It was acknowledged that funding and logistic support will both pose a challenge to implementation. An ongoing effort is required in order to continue improving the observational network.
The requirement of brevity in the BAMS state of the Southern Ocean document resulted in important discussions being omitted. Key omissions included assesment of the carbon system and the status of ice shelves. Documentation of methods was also limited. Thus the workshop became an opportunity to prioritize what should be assessed in future reports and best practices for doing so. Three recommendations were put forth. The first is that a paper be published with a comprehensive assessment of the Southern Ocean climate which includes extensive discussions of methods employed. This paper will provide a vital citation for future BAMS SOC assesments. The second recommendation is that given the brief nature of the BAMS report it is practical not to repeat the same calculations each year. Every year the community must determine interesting Southern Ocean climate “news”. This could cover an observed event (e.g. ice shelf calving) or be devoted to a topic that hasn’t been covered in recent assesments (e.g. the carbon content of the Weddell Gyre). A common view was that the BAMS SOC can provide a progress report on the Southern Ocean and how the community is contributing to the broader climate observation picture. The final recommendation is that specific themes or regions should be highlighted in their own section. A number of aspects of the Southern Ocean climate system deserving of annual reporting were discussed (e.g. Antarctica’s ice shelves and ice sheets). It was recommended that the community rally support around themes and lobby the BAMS SOC editors for inclusion of additional sections. The various topics covered in the Arctic section may provide a useful model for future reports on the Antarctic and Southern Ocean.
The SOOS State of the Climate workshop was a huge success in that many members of the community were able to give input regarding the content and direction of future Southern Ocean climate assesments. Efforts to carry out the recommendations are underway.
We would like to thank the workshop sponsors: Southern Ocean Observing System; Institute of Marine and Antarctic Studies, University of Tasmania, Australia; and the Deep South National Science Challenge, New Zealand. We would also like to thank the other workshop speakers and panelists: Nathan Bindoff, Karen Heywood, J-B Sallée, and Anna Wåhlin.
Leroy, S. S., Anderson, J. G., & Ohring, G. (2008). Climate signal detection times and constraints on climate benchmark accuracy requirements. Journal of Climate, 21(4), 841-846.
Meredith, MP, Mazloff M, Sallée J-B, Newman L, Wåhlin A, Williams MJM, Garabato NAC, Swart S, Monteiro P, and M. M. Mata SS, Schmidtko S. 2015. Southern Ocean [in “State of the Climate in 2014”]. Bull. Amer. Meteor. Soc.. 96(7):S157–S160.