The Ocean Engine: The Finale

It is now time to wrap up and conclude this blog. As stated, the aim of this blog was to explore past changes in ocean circulation; past, present and future. At the beginning of this blog, we sought to define the ocean’s role in the climate system. In particular, I stressed through the concept of the conveyor belt that whilst analogies are good for framing our imagination of ocean circulation, it is important to reconcile that the ocean is a complex mechanical engine in reality. Furthermore, this blog also reviewed techniques in palaeooceanography, to understand how past changes in ocean circulations are reconstructed. Upon reflection, given more time, I feel I could have perhaps added a few more detailed post attached to this topic as it is of crucial importance to understand past changes in ocean circulation.

This blog then explored paleaoclimatological evidence for past changes in ocean circulation. In particular, several episodic were analysed in relation to abrupt climate change; The Younger Dryas, 8.2 ka event and the Late Cretaceous. In this respect, I did narrow my research stand of the “ocean circulation” with respect to the North Atlantic circulation. However, this strand of blog posts could be nicely conceptualised by the posts I did on millennial-scale climate change. I found this strand of research particularly fascinating, given the complexity of feedback processes revealed from ice core studies such as the bipolar seesaw mechanism.

In this respect though, I feel I potentially could have explored two more strands of research in The Ocean Engine; notably the Southern Ocean and the El-Nino southern oscillation, both key are important components of the ocean system in relation to the past and present components. However, I feel that my definition of The Ocean Engine evolved over the course of this blog to run along the following theme; despite popular views advocated in films or by climate skeptics, ocean circulation is a complex science in reality. Thus through focusing on the North Atlantic thermohaline circulation, this blog sought to disentangle the myths attached to this component of the climate system by adopting a palaeo-approach and trying to understanding potential future impacts associated with anthropogenic climate change.

One of the key take-home messages from this blog is that by adopting a palaeo perspective, this can provide a critical position to interrogate questions surrounding climate change. The study of paleoclimatology differs somewhat in ideology to that of modelling, as emphasised in various posts. This divergence is a crucial point to convey a very key message; that anthropogenic climate change puts our understanding of the earth’s climate system to the limit, and we are currently at a stage to make fairly accurate predictions at the impacts but a significant distance away from the goal of quantification. I exemplified this point nicely in the last post, through the concept of noise, which I feel is useful analogy to help explain why uncertainty exists in climate change science; a crucial ingredient to any debate on climate change.

It has been an absolute pleasure writing this blog, and it has enriched my understanding of ocean circulation in relation to debates on climate change. To conclude, whilst we only are beginning to understand the components of The Ocean Engine, we do not understand its interior mechanics and complexities as of yet. Arguably, there are surprises in store as our understanding of the system develops and this warrants the seed for further research!

Conceptualising the ocean’s role in the climate system

Whilst we have been attempting to understand changes in ocean circulation due to climate change, it is useful to present a brief sketch of the earth’s climate system. The earth’s climate demonstrates variation on a variety of timescales, scaling from interannual, interdecadal, interannual, millennial and geological scales (Mann, 2007). Thus, the variation of the Earth’s climate can be conceptualised as the product of exogenous (factors independent of anthropogenic change and/or changes in other variables) and endogenous variables (factors affected by anthropogenic change and/or changes in other variables).

In relation to this conceptualisation, ocean circulation can be understood as an endogenous factor in the earth’s climate system.  Given that the current timescale of the Holocene (for purposes of simplicity this term is invoked) is shorter than that of orbital variations (eccentricity, precession and obliquity), termed “Milankovitch pacemakers”, events which occur at “sub-milankovitch periodicity” must be invoked to account for Holocene climate variability. Solar radiation and volcanic eruptions can be understood as examples of exogenous forcing factors and greenhouse gas emissions, EL Nino and Southern Oscillation (ENSO) are examples of endogenous forcing factors.

An interesting point in the debate is that of attribution; Is climate-change natural and /or anthropogenic driven? Crowley (2000) running a model similar to that of the IPCC, found that over the last 1000 years, solar forcing and volcanic forcing explained 41 – 59 % variance. However, when including greenhouse gas emissions and tropospheric aerosols in that scenario, it explained 41-64% of variance.

A question you may ask then is what accounts for the remaining 36 %?

I would argue that most of the uncertainty in the climate change debate is that it is difficult to attribute any degree of climate change conclusively to internal or external factors due to “noise” in the climate system. This can be defined by the fact that multiple possibilities exists for conceptualising the response of the climate system to a specific exogenous or endogenous forcing factors (Maslin and Christensen, 2007) (Figure 1). The climate system may respond directly in response to a variable (Figure 1a), in a delayed fashion (Figure 1b), a muted fashion (Figure 1c) or after passing a threshold (Figure 1d). To complicate this further, bifurcations exist, whereby forcing required to go through a threshold is different to the reverse, implying that once a threshold has crossed, it is difficult to reverse. This can be inferred from previous posts in relation to the impact of meltwater on the deep-water circulation.

Figure 1: Four Possible responses of the global climate system to forcing factors a) Linear b) Muted C) Non-Linear and D) Threshold.

Source: Maslin and Christensen (2007).

Whilst each element of the climate system may respond in variety of ways described above, this is compounded by the fact that each factor in the climate system responds on a different timescale. In this sense, the climate system is anti-essentialist in a way, the nature of climate system respond is multiple and complex. 

This conceptualisation illustrates the complexity of climate change science in particular in response to the question of attribution; detangling the greenhouse signal from that of natural climate variability.

Was Wally Right?

As the end of this blog approaches, I would like to draw a rather interesting paper for my last discussion on ocean circulation. Alley (2007) provides a substantial literature review on the evidence to support Wally Broecker’s ‘conveyor belt’ hypothesis which has advanced the study of abrupt climate change, particularly in relation to millennial-scale events.  Earlier posts in this blog have explored the scientific basis (see post 3) in addition to evidence used to support Broecker’s hypothesis (e.g. The 8.2 ka event and The Younger Dryas), these are not of the main concern here.

Rather, I would like to focus on what Alley (2007: 242)  describes as the ‘predictive power’ of Broecker’s hypothesis.  Alley (2007) argues that Broecker’s hypothesis can allow us to make certain predictions about the impact of anthropogenic climate change in relation to abrupt climate change. These predictions are particularly of relevance to the last few blog posts on the notion of “surprises”.

 This is that shutdown of the meridional overturning circulation (MOC) although unlikely, is still a crucial issue. This argument can be related to the last post on the a potential collapse of the thermohaline circulation, and that typical global warming “runs” over the next century appear to produce a slowing down of the MOC without it shutting down.  Secondly, Alley (2007) notes that meltwater inputs from the melting of the Greenland Ice Sheet could be sufficient to trigger a slowdown or shutdown. The findings from the Antarctic blog post, to highlight demonstrated a bipolar seesaw effect whereby melting from the Antarctic Ice Sheet seemed to have a stabilization effect on the MOC.  The key observation from that blog is the notion that a shutdown of the MOC will not be sufficient to trigger an ice age (As the Day after Tomorrow had us believe!). This reason will be explained in a forthcoming blog post.

Predicting the impacts of future anthropogenic climate change on ocean circulation is complex. As demonstrated, through the examples of the Antarctic, Greenland and thermohaline circulation; providing definitive and quantifiable answers are difficult. This uncertainty can be attributed to the fact that these questions seem to test our understanding of ocean circulation in using the analogue of the past. Thus, Broecker’s analogy of the conveyor belt can used as a sensible analogy to fill this void to conceptualise future impacts of global warming on ocean circulation. However, the non-zero possibility and potential large of abrupt climate change sows the seed for warrant research which is likely to in the long-term provide crucial knowledge to policy makers. 

Six Degrees: A worthwhile read...

Whilst this blog does not cover the science of climate change in its entirety, I would like to draw your attention to a very worthwhile read. Six degrees written by Mark Lynas in 2007, explores the scientific rationale underpinning the climate change debate that global temperatures will increase between one and six degrees.  Lynas (2007) undertakes a thorough review of the academic literature and rather masterfully as the temperature increases by 1 °C, on a chapter by chapter basis, takes the reader through six stages of global warming according to projections.  I believe it is an excellent read for anyone interested in climate change.  

This book was particularly inspiring for me and was one of the key factors for pursuing my interest in the science of climate change.

As Lynas (2007: xxii) writes:

“Climate change is the canvas on which the history of twenty-first century will be painted. Forewarned and forearmed”.

Surpises part 3: A Collapse of the thermohaline circulation?

Given the potential of the  north atlantic thermohaline circulation (NATHC)  to influence climate change as evident from study of abrupt climate changes from paleo-records, the following question is thus posed; will the NATHC collapse due to a rise in greenhouse gas emissions? The concern over the impact can be understood by the logic that an increase in sea surface temperatures due to thermal expansion of the oceans, may in turn reduce the solubility of CO₂ in warmer waters (Stocker et al. (2000 in Seidov et al., 2000). This in turn, may create a positive feedback effect whereby warmer waters hold less inorganic carbon, thereby causing a CO₂ release into the atmosphere (Stocker et al. (2000 in Seidov et al., 2000).

One issue in trying to understand this prediction through modelling is that it is difficult to achieve such a timescale. However, one of the earlier studies; undertaken by Manabe and Stoffer (1994), use a comprehensive atmospheric-ocean general circulation model (AOGCM) to simulate the impact over many centuries. Their results appear to indicate that a critical threshold lies between a doubling and four-fold increase in CO₂ concentrations. However only models of reduced complexity can be used to aid quantification of threshold, as this may allow for long-term simulations.

It has found that in addition to the stabilization level of greenhouse gas concentration in atmosphere, rate of greenhouse gas emission increase may also help determine a threshold. This is illustrated in Figure 1 with a simulation run by Stocker et al. (2000 in Seidov et al., 2000). They set climate sensitivity at 3.7 ° C for doubling of CO₂, and simulations of run concerning the rate of CO₂  increase a in 1 %/yr, 2 % year and 0.5 %/yr scenario (Figure 1a)

Figure 1a) Atmospheric CO₂ simulations for five experiments b) Simulated global mean temperature c) Simulations of maximum meridional overturning of the North Atlantic in Sverdrup (1 Sv= 10⁶ m³/s).

Source: by Stocker et al. (2000 in Seidov et al., 2000)

The global mean surface air temperature in simulations is not dependent on emission history for maximum CO₂ concentration (Figure 1b). However, a bifurcation appears apparent concerning the maximum meridional overturning of the North Atlantic. Reduction is apparent in all scenarios with the level dependent on maximum CO₂ concentration and CO₂ increase according to scenario. The circulation collapses at 750 ppmv with increase at rate of 1%./ yr. It then subsequently, appears to recover and settle to reduce value if increase is slower (0.5 %/yr ) or if the CO₂ level is reduced to 650 ppmv. At a rate of 2 %/yr, the circulation appears to collapse. The simulations appear to indicate that once the THC collapses it may settle a new equilibrium, with irreversible changes occurring, independent of CO₂ concentrations.

However, it must be noted that there is huge degree of complexity in understanding the impact of future changes in the thermohaline circulation due to anthropogenic climate change. These include; a lack of understanding concerning the variety of feedback mechanisms associated with changes in the THC and palaeoclimate modelling. There is a need to quantify components of climate-related components to a signal and test hypotheses regarding science of abrupt climate change. Changes in THC are likely in future and it is known the slowing down of THC moves system closer to thresholds. The uncertainty and importance attached to this as detailed in past blog posts highlight the need for further research.

Surprises part 2: The collapse of the Antarctic Ice Sheet

In contrast to Greenland ice sheet (GIS), potential melting of the Antarctic Ice Sheet (AIS) may have implications for ocean circulation. If you may recall from the posts on millennial-scale climatic changes,  potential freshwater outburst  (in particular from ice shelves around the Ross and Weddell seas) may affect the production of Antarctic Bottom Water (AABW) which in turn may level deep-water formation and in turn potentially affect global climates (Swingedouw et al., 2008). Swingedouw et al. (2008) attempted to quantify future AIS melting on climate by running 5 different experiments using the model LOVECLIM, a three-dimensional earth system model of intermediate complexity (Figure 1). These included a control simulation (CTRL) whereby forcing was set constant to pre-industrial conditions, notably with the CO₂ concentration in the atmosphere set to 277.6 ppm, where CO₂ concentration increased by 1 % per year until it reaches 4 times initial values, remaining unchanged till the year 3000. Scenario iAiG has fully interactive ice sheets over Antarctica and Greenland, whilst scenario fAfG, components are forced with a fixed ice-sheet component with simulated warming. Scenarios fAiG and iAfG show fixed and interactive GIS respectively.

Figure 1. Time series of annual mean value of the minimum of oceanic global meridional overturning streamfunction at 30°S ( 1 Sv= 10⁶ m³/s) emphasising the export of Antarctic and Circumpolar Deep Water (AABW and CDW) at 30°S and the maximum of Atlantic meridional overturning streamfunction at 30 °S, representing the export of North Atlantic Deep Water (NADW) at 30°S.

Source: Swingedouw et al. (2008).

The results are of particular interest. Without AIS melting, AABW export at 30 °S weakens during first 300 years and recovers, enhanced by changes compared to CTRL scenario after 1000 years. This can be related to sea ice changes due to freshwater forcing relating to the retreat of sea ice cover. Net annual mean sea ice melting in Weddell and Ross seas are lower in fAfG compared to CTRL. This in turn increases sea surface salinity and sea surface density, counter to density loss from projected temperature loss, increasing AABW formation in seas in fAfG compared to CTRL after 3000 years. However I would like to focus on particular observation in the above, that with AIS melting, this affects NADW export in all the scenarios but then this recovers after 1000 years in iAfg in contrast to the fAfg, indicating a potential stabilizing effect of AIS melting on the weakening of the NADW cell. AIS melting cause NADW cell weakening by 1.2 SV in iAiG compared to fAiG. This stabilization effect of the AIS, the authors note, is similar to that of the bipolar seesaw; a mechanism for explaining millennial-scale climate change (explained in blog post 17). This can be explained by the fact that a reduction in AABW density appears to enable NADW to penetrate further and deeper south in the Atlantic with the associated cell.

The findings of this study are important for two reasons. Firstly, the findings of the study highlight the importance of using palaeoscience to understand environmental problems; through the application of mechanisms used to explain millennial-scale climate change to help understand and try to predict future impacts of changes projected with  anthropogenic climate change (such as AIS melting) on changes in ocean circulation. Finally, the study highlights the importance of using models in addition to palaeoscience, to not forecast but to offer an insight into the feedbacks between ice sheets, ocean processes under a warming scenario. This point is important to communicate (to climate skeptics!) the manner in which science of climate change is being developed and understood; in this case through understanding the impact of the Antarctic Ice Sheet on changes in ocean circulation.  

Surprises part 1: The collapse of the Greenland ice sheet

Adopting a paleoclimatic perspective in our study of ocean circulation has highlighted the concern over the stability of the North Atlantic meridional overturning circulation (AMOC) given the discovery of millennial-scale climatic changes. To add to this, further concern must now be raised over the impact of greenhouse gas emissions on the collapse of the Greenland Ice sheet (GIS). It has been well documented that the GIS has been losing mass since the early 1990s, approximately 200 gt net loss (Wu et al., 2011). However, the retreat of this ice mass poses great concern for freshwater balance in the North Atlantic and in turn the AMOC. As can be inferred from the study of abrupt climate change through millennial-scale climate processes, a weakening of the AMOC has implications for global climate.

A vast body on literature exists on projections of the collapse of the Greenland ice sheet on AMOC, However, a study by Wu et al. (2011) is particularly of interest. They use a national centre for atmospheric research community climate system model to assess the influence of a shrinking GIS on AMOC, surface climate and sea level rise. However, the impacts on AMOC are of focus here. The authors use the IPPC A1B scenario, a mid greenhouse gas emissions scenario whereby CO₂ concentrations change from 368.5 ppm at year 1999 to 688.5 ppm at the year 2099, then a constant is assumed from the year 2099 to 2199 in conjunction with three other scenarios to predict the impact on global mean temperature.   This  scenario was used as a baseline which do not included GIS melting and subsequently scenarios were developed where rates of melting were 1 % and 3 % until the year 2099 and then kept at 2099 level until 2199 ( termed 1 % and 3 % exp) and then 7 % till 2050 and kept constant till 2199 ( termed 7 % exp) (Figure 1.)

Figure 1. The global mean surface temperature

Source: Wu et al. (2011).

The projections of the scenario indicate that a weakening of AMOC has a moderate effect on global climate. Mean global climate increases by 2. 43 °C in the last 20 years in the 21^st century, relative to last 20 years of 20^th century and an additional increase of 0.43 °C by end of the 22^nd century is observed in the A1B exp. scenario. Upon incorporating GIS melt-water, global mean temperature is not significantly affected in the first half of the 21^st century in all cases. However, at the end of the 21^st century, small fraction of a degree less is shown in the 7 % exp. scenario and the end of 22^nd century shows warming is 0.25 °C  and 0.50°C in the 3 % and 7% exp. scenarios. The 1 % exp. scenario seems to be in a similar time step to the A1B exp scenario. This has implications for our understanding of the impact of a collapse of the GIS on AMOC. That is that upon melting intensification of the GIS, AMOC weakens, global mean temperature is projected to rise due to increase in greenhouse gas emissions and that a weakened MOC will not reverse the current trend and trigger a global cooling trend.

However, the impact of melting GIS on global sea level means that the issue is still of pressing concern. We now turn to a more complex issue; predicting the impact of a potential collapse in the Antarctic ice sheets..