415 ppm CO2 threshold crossed May 2019

Some things in climate science are worryingly unpredictable – like when will the next ice shelf collapse off Antarctica? Or whether tropical storms will increase in frequency and strength in a warm world? Etc etc. However some things, like the ever increasing levels of CO2 in our atmosphere are getting all too predictable – Yesterday CO2 reached its annual maximum in the Mauna Loa Observatory, Hawaii crossing 415 ppm (13th May 2019).  Not that 415 ppm is an important figure, rather just a symbolic one.  Nonetheless, to honour this event I have updated our plot of CO2 over the last 3.5 million years.  And just to confirm, as has been the case for every May/June since the 1950s, atmospheric CO2 is higher NOW than it has been for around 2.5 million years. 

Plio-Pleistocene CO2 from ice core (red) and boron isotopes (blue). Sources indicated on text and listed  here . Data downloadable  here  and Rscript  here  to make your own version of this figure.

Plio-Pleistocene CO2 from ice core (red) and boron isotopes (blue). Sources indicated on text and listed here. Data downloadable here and Rscript here to make your own version of this figure.

To put this in context, this is a long time ago, e.g. this is what our human ancestors looked like back then (here).

New this year to the Pliocene to Pleistocene CO2 compilation: (i) new data from Dyez et al. (2018) for the interval around 1.5 million years ago; (ii) recalculation of Pliocene CO2 by Sosdian et al. (2018) and this.  This recalculation makes the data more accurate but also a little more uncertain. It also lowers the mean by 22 ppm – this doesn’t sound like a huge deal but makes the current CO2 higher than nearly all of the Pliocene.  Despite lots of talk it is still growing at 2-3 ppm per year - so I reckon by the time my kids leave school (ten years time) we will be outside of the Pliocene envelope and will have to go back 15 million years to the Miocene to see similar levels of CO2. It’s obviously not too late to act but we need to act fast!

Gavin takes part in award winning Radio programme

A BBC radio programme featuring Southampton Professor Gavin Foster has won an international award for science journalism.

The programme, broadcast on the BBC’s World Service in October 2017, was based around a listener’s question about ancient carbon dioxide levels compared to what they are today. Producer Cathy Edwards complimented Professor Foster on his contribution which she said was a “real highlight and played a part in the programme’s success”.

During his interview, Professor Foster, a specialist in Isotope Geochemistry, described how he is able to determine historic CO2 levels going back millions of years by studying ancient sea fossils drawn from sediment samples recovered from the sea floor. You can read more about this research here, and further information is also available here.

The Science Journalism Awards have been administered by the American Association for the Advancement of Science (AAAS) since their inception in 1945 to honour distinguished reporting for a general audience around the world. The awards are endowed by The Kavli Foundation which is dedicated to advancing science for public benefit and understanding.

You can listen to the full programme via the BBC website here.

PhD opportunities @theFosterlab

Through our new DTP at Southampton (INSPIRE) we have 3-fully funded PhD projects available this year. Please click on the links below for more details and how to apply through GSNOCS (our graduate school). The deadline is 04 January 2019. Please contact Gavin through this website (here) or email: gavin.foster@noc.soton.ac.uk

Boron isotopic composition of coralline algae at super-high resolution determined using laser ablation by    Fietzke et al. (2015, PNAS)   . Examining the boron & carbon isotope variability at this scale in    Project 1    is key for unpicking the health of coral reefs.

Boron isotopic composition of coralline algae at super-high resolution determined using laser ablation by Fietzke et al. (2015, PNAS). Examining the boron & carbon isotope variability at this scale in Project 1 is key for unpicking the health of coral reefs.

Project 1. Reconstructing the health and productivity of coral reef systems at high resolution: Laser ablation analysis of boron and carbon isotopes in corals. Here the student will develop and apply novel laser ablation methods to analyse boron and carbon isotopes in situ in corals (and potentially other carbonates) to determine the evolution of the ocean carbonate system at super high temporal resolution.

The Li, Sr and Os isotopic composition of the oceans over the last 70 million years from    Misra et al. (2012)   .    Project 2    investigates the role of the changing carbonate system in driving d7Li.    Project 3    revisits the variability in 87Sr/86Sr over the Cenozoic to better determine the relationship between silicate weathering and climate.

The Li, Sr and Os isotopic composition of the oceans over the last 70 million years from Misra et al. (2012). Project 2 investigates the role of the changing carbonate system in driving d7Li. Project 3 revisits the variability in 87Sr/86Sr over the Cenozoic to better determine the relationship between silicate weathering and climate.

Project 2. Lithium isotopes in foraminifera: a new proxy for the ocean carbonate system? Li isotopes in carbonates may be a tracer of weathering, but do they also reflect changes in the carbonate system? This project will examine this question through a combination of field, laboratory and culturing studies.

Project 3. Old proxy, New techniques: Reshaping the seawater strontium curve and resolving its implications for climate feedback processes. Sr isotopic composition of marine carbonates has long been used as a tracer of silicate weathering, new analytical techniques developed at Southampton offer the potential to use Sr to open a new window into the relationship between weathering and climate.

Hyperthermals - insights into our warm future from past rapid changes in climate

flood volcanism.jpg

There are few, if any, direct analogues for anthropogenic climate change in the geological record.  This is because it is occurring at a pace that is rarely seen naturally, short of those rare times when the Earth is hit by an asteroid (e.g. 66 million years ago). There are events that occur natural that are however relatively similar - these are known as the “hyperthermals”.  These are geological rapid, relatively short events (<1 million years) characterised by rapid warming and caused by the injection of carbon to the climate system – typically a doubling or more of CO2 (see table below for a list of the most recent ones).

A summary of the most significant hyperthermals of the last 300 million years.&nbsp; Italics indicate a high degree of uncertainty.&nbsp; From Foster et al. (2018) -  click here , also see original paper for references.

A summary of the most significant hyperthermals of the last 300 million years.  Italics indicate a high degree of uncertainty.  From Foster et al. (2018) - click here, also see original paper for references.

We @thefosterlab, with Celli Hull, Dan Lunt, and Jim Zachos organised a discussion meeting at the Royal Society last September to bring together scientists from many branches of the Earth Sciences to advance our understanding of these events and crucially try and fathom what they can tell us about our warm future.  This research has now been written up and turned into a special volume of the Philosophical Transactions of the Royal Society (here). I recommend you go and read them as many are open access and free to download. 

(Briefly) What can we learn from hyperthermals?

Exactly how the climate will respond to anthropogenic forcing is currently uncertain because our understanding of the climate system is incomplete. Hyperthermals, however paint a very consistent picture of how the Earth has responded in the past intervals of rapid and massive carbon addition (see table above).  These features include:   

(i) rapid global warming of >3 C. 

(ii) a reduction in oceanic oxygen content leading to ocean anoxia and/or euxinia. 

(iii) ocean acidification of around 0.3 to 0.4 pH units. 

(iv) the hydrological cycle intensified with wet regions generally getting wetter and dry regions drier.

(v) continental erosion/weathering rates were enhanced.

(vi) relatively large biotic responses occurred in the first half of the Phanerozoic (Paleozoic and early Mesozoic mass extinctions often associated with hyperthermals), and muted or mixed responses in the latter half of the Phanerozoic. But in each case the hyperthermals are associated with biotic disruptions. 

(vii) the Earth system takes 100's of thousands of years to recover once C-emissions have stopped.

These changes to the system were likely caused by the introduction of 10,000 to 40,000 Pg of C over a couple of millennia.  Where this carbon came from is debatable but CO2 from flood basalt eruption and emplacement is likely involved (see here and here).  Although this is rapid for geology, the rates of carbon addition are still likely 10x less than current rates (i.e. less than 1-2 Pg C per year vs. 10 Pg C per year; though this is controversial – see this paper in our special volume). 


As the figure above from Gattuso et al. (2015, Science) shows, we have seen a very similar response to the Earth System to current anthropogenic emissions, for instance in response to increasing atmospheric CO2 by 40% over the last 150 years we have seen sea surface temperature increase by 0.5 C, ocean acidification by 0.1 pH unit and a decrease in ocean oxygen content.  Projections by the IPCC and others suggest we are heading on a similar path as the ancient hyperthermal events (see figure above). 

A key difference between what we are currently doing to the Earth System and what happened during these hyperthermal events however is the rate of current change is much much faster.  The magnitude of CO2 change from any C emission, and hence the full magnitude of warming, anoxia, ocean acidification and extinction that occurs, is a function of rate, due to the timescales of a number of key negative feedbacks (see this paper).  Why the Palaeozoic hyperthermals are associated with significantly greater extinction rate than more recent hyperthermals is currently not known (see table above).  However, a consensus is emerging that it is the extreme heat and anoxia that are the likely “kill mechanisms” (see this paper in our special volume).  Given that the rate of carbon addition during our “anthropogenic hyperthermal” eclipses that of the Palaeocene Eocene Thermal Maximum (PETM), at the very least we are likely looking at a potential future with a more severe impact on life on Earth than any climate change event of the last 56 million years.  Exactly how severe however remains perhaps the biggest unknown.       

An unprecedented degree of undersaturation?

Many people are familiar with Michael Mann’s famous Hockey Stick of global surface temperature over the last 2000 years, clearly showing the influence of anthropogenic climate change.  In a recent paper published in Earth and Planetary Science letters this week led by Sindia Sosdian from Cardiff University we show a series of “ocean carbonate system” hockey sticks showing how anthropogenic and future change in CO2, pH and aragonite saturation compare to what the Earth has experienced over the last 20 million years or so.

Surface water pH (A), atmospheric CO2 (B) and surface water aragonite saturation state (C) over the last 20 million years.&nbsp; Bands encompass the mean and 1sigma uncertainty.&nbsp; The different colours represent different scenarios (see the paper for more detail).&nbsp; The plots on the right-hand side show historical (grey) and future projections from  Winklemann  et al. (2015).&nbsp;

Surface water pH (A), atmospheric CO2 (B) and surface water aragonite saturation state (C) over the last 20 million years.  Bands encompass the mean and 1sigma uncertainty.  The different colours represent different scenarios (see the paper for more detail).  The plots on the right-hand side show historical (grey) and future projections from Winklemann et al. (2015). 

From the long-term trends in the above figure, pH (the –log10 of the H+ concentration) has increased and CO2 decreased, on the whole, over the last 20 million years (with some interesting structure that is for another post for another day).  CO2 and pH are clearly tightly coupled over this time, the reason is actually a bit complex, but can be simply thought of arising because CO2 is an acidic gas, so more CO2 = lower pH and vice versa.  Furthermore, the CO2 content of the ocean dictates the CO2 content of the atmosphere, so this tight coupling between pH and CO2 is not a surprise.  We are not sure why CO2 declined and pH increased through the last 20 million years but it most likely relates to a long term decline in CO2 outgassing from the mantle or a gradual increase in the weathering of silicate rocks in the Himalaya.

The key new record in this latest study however is the evolution of the saturation state of calcium carbonate (CaCO3) over the last 20 million years (in figure above expressed as omega aragonite – the saturation state of the aragonite polymorph).   When saturation state is greater than 1 CaCO3 can precipitate easily, when its below 1 CaCO3 dissolves.   Organisms that make their shells and skeletons out of calcium carbonate, like corals and shell-fish, require a high degree of CaCO3 oversaturation.  Similarly, carbonate structures like coral reefs exist in a delicate balance between dissolution and accretion, so any decline in saturation state can start to weaken and dissolve the reef. 

This really great figure below from Honisch et al. (2012; Science), firstly shows the close relationship between atmospheric CO2 (panel A) and surface ocean pH (panel B) in a model where you double CO2 on different timescales (warm colours – fast, cold colours – slow). In panel C the mean surface ocean saturation state of aragonite is shown.  Although it looks similar to the other two on short timescales, on long timescales of CO2 addition it becomes decoupled from pH and CO2 and doesnt change very much.  This is perhaps more clearly shown in panel D.  It’s not easy to explain why this happens and those of you interested should look at the Honish et al. (2012) paper in more detail (or see wiki or this Royal Society report)   


Our new record of aragonite saturation state shows that as predicted, there isn’t much of a trend over the last 20 million years, consistent with this idea of a decoupling of pH and CO2 from saturation state when CO2 change is slow – i.e. on thousands to million year timescales.

However, over the last 150 years or so, CO2, ocean pH and saturation state have all changed in tandem because the changes are so fast.  What the figure from our new paper shown above shows is that attempts to mitigate the effects of climate change by restricting CO2 rise at 2100 to <500 ppm (RCP2.5) or so keeps ocean pH and saturation state to well within the range of the last 5 million years or so.  The RCP8.5 scenario – the often called “business as usual” – risks tipping saturation state to lower values than have been seen in the last 14 million years, and maybe longer.

What does this mean for calcifying organisms? Well I guess we just don’t know for sure without further study. But the message is clear – if we continue to emit CO2 at the current rates we risk taking the Earth to a state not seen for many millions of years…

410 ppm CO2 for April 2018 - first time in millions of years....

So with spring having sprung we reach the highest monthly CO2 at the Mauna Loa for 2018 - April 2018 was 410.26 ppm, thats 1.26 ppm above last years peak (Mauna Loa data here). 

@thefosterlab we determine the levels of CO2 over the last 50-60 million years using the boron isotope proxy.  It is often said that at 410 ppm CO2 is now higher than any level the Earth has seen for at least 3 million years.  As this revised plot shows (see here for Rscript and here for data), this is likely true, but we are now beginning to creep over even the highest values in the Pliocene 3 million years ago.

Atmospheric CO2 from AD 1000 to AD 2018 (right) from a mix of ice core records and measuresments of the astmosphere from Mauna Lao.&nbsp; On the left is a compilation of ice core CO2 (red) and boron isotope based estimates (blue).&nbsp; Note the age scales are different but y-axis is the same.   See  this document  for references.

Atmospheric CO2 from AD 1000 to AD 2018 (right) from a mix of ice core records and measuresments of the astmosphere from Mauna Lao.  On the left is a compilation of ice core CO2 (red) and boron isotope based estimates (blue).  Note the age scales are different but y-axis is the same. See this document for references.

Beyond the Pliocene we have to go to the Middle Miocene Climatic Optimum around 16 million years ago to see CO2 >450 ppm. There have been several papers recently on this topic (e.g. this one by James Super of Yale) and I desparately need to find time to update the compilation on p-CO2.org...at current rates of CO2 rise though I probably have a couple of years before Earth crosses 450 ppm....

Research Experience Placement for Summer 2018

The micro-environment around the planktic foram  O.Universa  being analysed for O2 and pH (top).&nbsp; A close up of a live  O. universa  collected Nov 2017 from the Sargasso Sea, Bermuda (bottom)

The micro-environment around the planktic foram O.Universa being analysed for O2 and pH (top).  A close up of a live O. universa collected Nov 2017 from the Sargasso Sea, Bermuda (bottom)

Larger view of the  O. universa  being probed in the laboratory using micro-electrodes

Larger view of the O. universa being probed in the laboratory using micro-electrodes

TheFosterLab has a fully funded Research Experience Placement for the summer of 2018 (17 June to 21 September 2018).  We are looking for a quantitative student from a discipline not normally funded by NERC (e.g. from chemistry, physics, engineering, biology NOT geology, earth science, environmental science) who is keen to apply their skill set to help understand foraminifera vital effects.

Details of the scheme which is run out of our NERC DTP can be found here: (http://www.spitfire.ac.uk/spitfire-dtp-research-experience-placement-scheme-2018), the project is described below. Please contact Gavin if you are interested (gavin.foster@noc.soton.ac.uk) or have any questions

Opening the black box: the influence of environment on foraminiferal physiology

The foraminifera are a group of amoeboid sub-mm sized protists with an extensive fossil record that play an important role in global carbon cycling due to their ability to form calcium carbonate shells (known as tests).  The chemical composition of their tests is predominantly determined by: (i) the composition of the seawater they grew in and (ii) a number of environmental factors (e.g. temperature, salinity, and pH).  The latter forms the basis of many quantitative reconstructions of climate over the last 140 million years.  Such reconstructions inform our understanding of how the climate system works and help to improve our predictions of future climate in the face of anthropogenic climate change.  Assessing the reliability and uncertainty of our quantitative reconstructions of past climate using the chemical and isotopic composition of foraminifera is therefore key.

However, rather than being passive recorders of the environment they live in, the composition of the foram tests are heavily influenced by the physiology (i.e. life processes) of the foram and its photosymbionts.  Indeed, it is the influence of environment on physiology that often imparts an environmental sensitivity to test composition. This influence comes about predominantly because foraminiferal calcification and respiration and symbiont photosynthesis modify the pH in the immediate 1 mm or so around the growing foram, such that it is no longer simply growing in seawater but seawater with a composition that is modified by the growth of the foram itself.

This short project will aim to better understand the role of the external environment on the physiology of foraminifera, and hence test composition, by using microelectrodes to make measurements of the pH, [Ca2+], and O2 in the micro-environment around growing foraminifera under controlled conditions. The student will therefore gain experience in the maintenance and study of foraminifera in laboratory culture and in the development and use of ion-selective microelectrodes to measure pH, Ca and O2 at micron-scale resolution in biological samples.

Experiments will be performed at the National Oceanography Centre under the guidance of Prof. Gavin Foster, Dr. Tali Babila (OES), Dr Glen Wheeler (MBA), Dr Gerald Langer (MBA). The results of this study will feed into a larger  NERC funded project SWEET aimed at reconstructing climate 50  million years ago.  

The task the student will be to perform the measurement of pH, O2 and Ca2+ gradients in the micro-environment around multiple specimens of two benthic foraminifera species Ammonia sp. (non-symbiont bearing) and Amphistegina (symbiont bearing). The magnitude of the chemical gradients measured are indicative of the fluxes of ions and molecules in the micro-environment allowing a quantification of the magnitude of calcification, respiration and photosynthesis for each individual.  By manipulating the environment the foraminifera are inhabiting, e.g. by changing the temperature, pH and chemical composition of the culture media, we will gain unique insights into how environment influences foram physiology. 

An approximate plan is:

Weeks 1-2: ambient conditions

Weeks 2-5: modified temperature

Weeks 5-7: modified pH

Weeks 7-10: modified chemical composition

No prior knowledge of foraminiferal biology is required but experience with the use of micro-electrodes and an understanding of how they work is desirable. The candidate should also be numerate and comfortable with data processing.

Narrowing in on Equilibrium Climate Sensitivity

A post by Gavin, Phil Goodwin and Eelco Rohling

A fundamental variable in describing how warm our future will be is the equilibrium climate sensitivity (ECS) – the global mean surface temperature change in response to a doubling of atmospheric CO2, once the system has reached equilibrium.  Assessing the impact of anthropogenic climate change has been framed in these terms for nearly 120 years since Svante Arrhenius in 1896 (LINK). However, a precise value for Earth’s ECS has stubbornly resisted determination over this time – see this figure I have modified from Steve Schwartz

The evolution of the estimates of equilibrium climate sensitivity over the last 130 years.&nbsp; From Steve Schwartz with the addition of the 2013 AR5 IPCC report.&nbsp;   http://www.ecd.bnl.gov/steve/schwartz.html

The evolution of the estimates of equilibrium climate sensitivity over the last 130 years.  From Steve Schwartz with the addition of the 2013 AR5 IPCC report.  http://www.ecd.bnl.gov/steve/schwartz.html

This apparently lack of scientific progress was exemplified in the last IPCC report in 2013 when, instead of quoting a most likely value as previous reports had done, the best estimate of ECS based on multiple methods was that it was likely (i.e. with a 66% probability) in the range of 1.5 to 4.5 C per CO2 doubling.  This is arguably something we have known since the late 70s (although by luck rather than judgement; see figure above).  With the international agreement in Paris in 2015 to limit global warming to 1.5 or 2 C this uncertainty in the true value of ECS has come into sharp focus - quite simply in order to ensure we don’t go above 1.5 or 2 C we first need to know how sensitive the climate system is.

In the last year, with an eye on these targets, there have been a number of attempts to narrow in on the real value of the Earth’s ECS.  For instance, Armour (2017; doi:10.1038/NCLIMATE3278) showed that because climate feedbacks operate on different spatial and temporal scales, simple energy balance approaches, though elegant, tend to biased low (by ~26%).  Brown and Caldeira (2017; doi:10.1038/nature24672), confronted climate model output with a series of observations, and those models that best matched the observations were found to describe a narrower range in ECS.  Finally, last week Cox et al. (2018, doi:10.1038/nature25450) found that only a subset of the available climate models exhibited a similar level of year-on-year variation in climate to what is observed in the historical record.  Interestingly, this subset of models also had similar and narrow range of ECS.  Taken together, these recent advances suggest that an ECS of <2 or >4.5 is unlikely (summarised nicely by a Nature news and views article by Piers Forster and this figure below).


Recent estimates of equilibrium climate sensitivity.  The estimate from the IPCC (2013) is based on several lines of evidence.  Bars depict the ranges for which there is a 66% likelihood of the value being correct.  Best estimate is a blue line.  From Forster (2018).

Our latest paper, published today (22/01/2018; Goodwin et al. 2018, doi:10.1038/s41561-017-0054-8), continues this recent trend of narrowing in on the real value of ECS for the Earth in its current state.  In contrast to these other studies we used a much simpler model (the WASP Earth System model, http://www.waspclimatemodel.info) but one that could be run many times with different input parameters.  The accuracy of the outputs could then be tightly constrained by observations of anthropogenic climate change (e.g. surface warming, ocean heat up take etc.), thereby identifying the most realistic parameter sets and so most accurate model runs.  We started from the premise that the geologically constrained estimates of ECS from the Palaeosens study (Rohling et al., 2012, doi:10.1038/nature11574), represent a good, if uncertain, approximation of the Earth’s true sensitivity (in Bayesian terms this is our prior).  We found that, given the uncertainties in climate forcing over the last 100 years, only a subset of the initial geologically-determined ECS was able to generate climate change in WASP that adequately agreed with the available observational datasets (e.g. ocean heat uptake and surface warming).  Like in the other studies discussed above, this subset (our posterior) had a much narrower distribution of ECS: a most likely value of around 2.7 C per CO2 doubling and a 66% likelihood range of 2.3 to 3.2 C.  This agrees remarkably well with estimates of Cox et al. (2018) as shown in this figure.


Estimates of ECS from Goodwin et al. (2018) and Cox et al. (2018).  NB – for plotting we have assumed a normal distribution from Cox et al. (2018).  Our 66% range is 2.3 to 3.2 C per CO2 doubling with a most likely value of 2.7.  Cox's 66% range is 2.2 to 3.2 C with a most likely value of 2.8 C per CO2 doubling..

At face value, this appears to be great news.  For instance, the extreme global warming implied by an ECS >5 C now appears to be very unlikely, as does the possibility of a “lukewarm” sensitivity.  However, there is a downside too, because by reducing uncertainties in ECS we are also reducing our “wiggle room” in climate negotiations and emission-reduction actions.  To illustrate this in Goodwin et al. (2018), we take our subset of observationally consistent models and see how quickly they approached 1.5 and 2 C above pre-industrial, given current emission rates (see figure below).  What we find is startling – within 17-18 years we are likely to have reached the 1.5 C target and by 35-41 years we will have reached 2 C above pre-industrial levels.  So, rather than being relieved that our study rules out extreme climate change on century timescales, we place the need to rapidly decarbonise our society into ever more sharp focus. 

Cumulative Emissions.PNG

Cumulative carbon emissions and warming projects from our observationally consistent ensemble.  From these plots, given a known rate of emissions, the time at which a target T is crossed can be calculated. From Goodwin et al. (2018).


2017 - The highest CO2 for millions of years

As 2017 draws to a close it is set to be 2nd hottest year on record.  With an annual average of around 407 ppm, the concentration of CO2 in the atmosphere was also a record breaker - in fact its the highest its been since the Pliocene around 2.5 to 3 million years ago (see figure below)

Atmospheric CO2 from AD 1000 to AD 2017 (right) from a mix of ice core records and measuresments of the astmosphere from Mauna Lao.&nbsp; On the left is a compilation of ice core CO2 (red) and boron isotope based estimates (blue).&nbsp; Note the age scales are different but y-axis is the same.   See  this document  for references.

Atmospheric CO2 from AD 1000 to AD 2017 (right) from a mix of ice core records and measuresments of the astmosphere from Mauna Lao.  On the left is a compilation of ice core CO2 (red) and boron isotope based estimates (blue).  Note the age scales are different but y-axis is the same. See this document for references.

Here @theFosterlab we use the boron pH proxy to reconstruct CO2 in the geological past beyond the reach of the ice core record (the last 800 thousand years, kyr). What the figure above shows is a compilation of the latest boron data including some data we published this year compared to actual measuresments of the atmosphere and bubbles of ancient atmosphere trapped in the ice cores.  The first time we made a plot like this was a couple of years ago when CO2 was still just below 400 ppm.  Given the relentless rise in atmospheric CO2 it looks like it will be around 20 years or so until even the highest values in the last 3 million years are exceeded.  Pliocene - here we come!

The R-script to make the above figure can be downloaded here, the data here and the relevent references here. If you are interested in coming and working with us to reconstruct CO2 in the past we have several fully funded PhD studentships on offer (see here).  

PhD projects @TheFosterlab

This year we are advertising two PhD projects via our @NERC DTP SPITFIRE (http://www.spitfire.ac.uk/).  These studentships are fully funded, but funding is allocated on a competitive basis (based on interview and application pack).  See links below for details how to apply.

Project 1 – The Role of CO2 in the first ice ages: insights from very high-resolution boron isotope records.

The ice core records of CO2 only stretch back 800 kyr.  To find out how CO2 and climate are related before this we need to rely on indirect proxy methods – here we will apply the boron isotope proxy.  By examining the first big glacial intervals following the intensification of northern hemisphere glaciation 2.7 million years ago, this project aims to address the following questions:

·         Does the sensitivity of the climate system vary as background climate state changes?

·         What is the magnitude of CO2 variability in the early Quaternary on orbital timescales? And how does this constrain the mechanisms responsible for orbital changes in CO2?

·         What are the phase relationships between the major components of the climate system in the early Quaternary? What does this reveal about how the Earth System functions when warmer and colder than today? 


High-resolution SST records from the last glacial cycle and the target interval for this project (Marine-isotope stage 100).  Taken from Shakun (2017). 

Project 2 - Can silicate weathering regulate atmospheric CO2 during periods of rapid climate change? Testing the negative feedback hypothesis in the geological record

Over geological timescales the silicate weathering feedback process is thought to have kept Earths climate within habitable bounds by regulating atmospheric CO2 levels.  Theory suggests that it the process of weathering of continental rocks that does this. A debate remains however regarding how much weathering is a negative feedback vs. a driver of climate change.  This project aims to characterise the link between silicate weathering and climate by applying weathering tracers (e.g. d7Li, radiogenic and stable Sr-isotopes) to examining the response of weathering to rapid warming events in the past. 


Rachael goes to BATS

In October 2017 while on placement at Bermuda’s Institute for Ocean Science (BIOS) PhD student Rachael Shuttleworth had the amazing opportunity to take part in one of the monthly Bermuda Atlantic Time Series (BATS) cruises on board R/V Atlantic Explorer. The BATS cruises have taken place pretty much every month since they began in 1988, where they take monthly measurements of hydrographic and biogeochemical parameters through the water column at sites within the Sargasso Sea. This time series is an invaluable climate dataset which helps us tackle big-picture questions and better understand global climate change and the oceans’ response to this.

Rachael tells us about it:

When at sea the scientists and crew work shifts around the clock and it is impossible to keep up with all the different projects that are going on! The first job was to locate and recover lost glider “Jack” who had lost a wing, and suffered a shark attack!

Searching from the bridge for glider Jack

Searching from the bridge for glider Jack

Leaving BIOS

Leaving BIOS

Once Jack was safely on board we set off for the BATS sample site where the sediment traps were deployed. We waved goodbye to these for the week and began on the first (of many) CTD casts. The BATS team collect samples from the CTD’s for a wide range of analyses including nutrient levels, alkalinity, oxygen content, salinity, (the list goes on…!). I however was collecting samples to characterise the carbon and oxygen isotopic composition (d13C and d18O) of the water column. The water sampling came to be my favourite job on the cruise as all the scientists get together and it’s a great time to catch up with everyone!

CTD as it enters the water and begins its’ descent to 5500m!

CTD as it enters the water and begins its’ descent to 5500m!

Deploying a zooplankton tow

Deploying a zooplankton tow

While the weather was still calm we were able to do a couple of zooplankton tows, however the winds soon picked up and conditions got so rough that all science was called off for the next day at least. I was secretly grateful for this, since I’d had little sleep over the previous 2 days, however my stomach was less grateful…

Sediment traps

Sediment traps

Things had calmed down by Wednesday evening and we were able to get back to work (which consisted of many, many more CTD casts, plankton tows, radiance measurements, and collection of the sediment traps). We returned back to BIOS late Friday evening with boxes upon boxes of samples to be run – writing this a week later I don’t feel I’ve even began to scratch the surface of mine…

Science Week 2017 @ Oakwood Primary School

On Tuesday 4th July Hannah and Rachael visited Oakwood Primary School as part of their Science Week to give an assembly to our youngest climate scientists yet – Key Stage 1 (4-7 year olds)! We have made our presentation (here) available to the public, so please feel free to use it as you wish.

Hannah Donald and Rachael Shuttleworth visit Oakwood Primary School

Hannah Donald and Rachael Shuttleworth visit Oakwood Primary School

It is so important that the key causes of the global issue that is climate change is taught in Primary Schools, as these children are part of the generation who will be living with the consequences, and finding solutions to it! The assembly took the children through a journey from learning about the fundamentals of the ‘greenhouse effect’ and how greenhouse gases have affected Earth’s climate through geological time. We then played a great game of “natural vs. man-made”, in which we introduced where greenhouse gas emissions come from, and discussed concepts such as deforestation, how our diet affects global emissions, and made the key link between electricity use and emissions. It is also important to highlight that it is not all doom and gloom, and that by working together, we can help to combat climate change! There are many tiny things that everyone (even 4 year olds!) can do every day such as switching off lights, recycling as much as possible, and opting to walk, cycle, or use public transport as much as possible.

The National Oceanography Centre in Southampton runs a scheme called ‘Eco-Schools’ which is aimed at Junior (KS2) Schools in Hampshire. This consists of a day-long visit to NOC and includes an introduction to climate change, a visit to the NOCS aquarium and several carbon footprint activities. The Foster Lab has maintained a strong tradition for STEM outreach involvement, with Hannah and Rachael helping to organise these award-winning Eco-Schools events since 2013 (more info here)

Hannah and Rachael are also STEM (science, technology, engineering and maths) ambassadors. The STEM Network (https://www.stem.org.uk/) is another fantastic way for schools to access local scientists and engineers who are enthusiastic and knowledgeable role models to inspire young people into the world of STEM. We encourage scientists and schools alike to get involved!


A really useful way of visualising the relationship between different aspects of the ocean carbonate system is through plots of alkalinity (ALK) vs. dissolved inorganic carbon (DIC).  For instance, the one below made by James Rae clearly shows that pH and CO2 are tightly related - as DIC or ALK (or both) change pH and CO2 change similarly.  It is this relationship that largely underpins the utility of the boron isotope pH proxy (see here for e.g.). 

ALK vs. DIC from Foster and Rae (2016; 10.1146/annurev-earth-060115-012226

ALK vs. DIC from Foster and Rae (2016; 10.1146/annurev-earth-060115-012226

I tend to use the R for making plots and analysing data these days and normally if there is something you don't know how to do you can just google it and someone would have uploaded some script somewhere.  But after many hours of searching i couldn't find something to make plots like these, so had to make my own using the seacarb package.  I have uploaded it here with some annotations so you can make one if you like.  I dont think mine is as pretty as James' but its not bad for a first attempt -  Let me know if you make it better! 

My trip to New Zealand

Post By: Rachael Shuttleworth

What drives glacial – interglacial CO2 level changes is what is known as the “holy grail” of palaoceanography. Though the question as to whether these changes are caused by biologically or circulation dominated process is still hotly debated. To better quantify and enhance our understanding of the processes and mechanisms which control CO2 on these timescales we must look to the past for answers. This is the focus of my PhD @thefosterlab and we are initially targeting cores taken from variety of latitudes in the southern Pacific Ocean, ranging from subtropical to polar waters. These cores are stored at NIWA in Wellington, New Zealand – someone had to do the tough job of collecting the samples...

In order to get the data we want, a lot of preparatory work on the samples has to be undertaken. Firstly we must sample to core itself, taking slices of mud at specific depths which correspond to the time periods of interest.

These cores are then put into an oven overnight to dry out.

These dried samples are then washed over a 63micrometer serve to remove the very fine mud particles. They are then put back into the oven to dry again.

The information that we want is stored in foraminifera within these sediments, and so the next step is to pick these micro fossils out one by one.

These samples are now prepped and will be sent back to @thefosterlab based at the National Oceanography Centre, Southampton where analysis on their isotopes will be undertaken. Many of these cores already have carbon, oxygen and nitrogen isotope records which can tell us about temperature, age, ice volume, and productivity. For my PhD I will focus on obtaining a boron isotope record which will tell us about the change in flux of CO2 into the oceans from the present day back to the last glacial maximum. By comparing this with dust flux and productivity records I hope to better quantify the role that relaxation of micronutrient (such as iron) limitation due to enhanced dust deposition played during the last glaciation. This will provide further insight into how much biology played a part in driving glacial-interglacial CO2 change, as well as the potential carbon storage via this mechanism of iron fertilisation.

Its still early days for my PhD, and I now have my first sample set.  Next stop – the lab!

Future relevant climate sensitivity (part deux)

A paper out this week by Friedrich et al. (2016) is the latest in what seems like a series this year determining climate sensitivity using the palaeoclimate record (see here).  This is a very powerful approach but has its difficulties many of which are discussed in our previous post and here.  The Friedrich et al (2016) study used a new empirical estimate of Surface Air Temperature (SAT) based on a compilation of Sea Surface Temperatures (as did Snyder recently) and a complete assessment of the processes “forcing” climate change over the last ~800 thousand years (e.g. CO2, land-ice albedo, and dust) to identify that climate sensitivity changed as a function of climate state: they found it was ~1.8 K per CO2 doubling when the Earth was substantially colder than today and ~5K per doubling when the Earth was only a little bit colder than the pre-industrial.  We @theFosterlab were just involved in a review on this subject (here) and a summary of the literature from that paper is shown below.  

Friedrich et al. go further than other studies and apply the palaeo-sensitivity to predict our warm future (see figure below).  They correct for the fact that over the next 100 years the climate isn’t in equilibrium with its forcing and show that the high “warmer-paleo” sensitivity yields temperature in 2100 AD, given a business as usual RCP8.5 emissions scenario, that could be as high as 6 K (5-7 K) compared to 5 K (3 -6 K) from the CMIP5 models (see figure below). 

The accuracy of such a paleo-sensitivity approach to predict the future climate is: (i) very dependent the accuracy of the temperature record used to determine sensitivity in the past (and others limitations, see http://julesandjames.blogspot.co.uk/); (ii) but is reliant on the assumption that the sensitivity estimated from climates slightly warmer and substantially cooler than today is applicable to Earth temperatures up to 5 K warmer.

I think it’s a great finding of this paper that climate sensitivity is state dependent over the Pleistocene glacial cycles, but this also means it’s probably an over simplification to assume that a similar state dependency doesn’t characterise the system when temperatures are substantially higher than the pre-industrial (see top figure). 

We attempted to investigate this in this paper in 2015 using boron isotope based CO2 data from the Pliocene.  We didn’t examine the data for a state dependency within the Pliocene (see here for why) but determined that the average climate sensitivity in the Pliocene was ~3.7 K per CO2 doubling (with an uncertainty of 2-6 K).  This overlaps with what Friedrich et al. determine as “warm-paleo” sensitivity.  However, my feeling is that, given the current uncertainties in determining climate sensitivity in the past, rather than indicating the Earth system is more sensitive than the future, studies like that of Friedrich et al. (2016) and our own (http://www.nature.com/nature/journal/v518/n7537/full/nature14145.html) are very important validators of our understanding of the behaviour of the climate system encapsulated by the CMIP5 models.  These are very different ways to understand the climate system yet they give the same results – this is very powerful and is a great illustration of the utility of palaeoclimate research.  As a community we now need to work hard to reduce the inherent uncertainties and get the most out of the rock record of past climate changes.

Future Relevant Earth System Sensitivity

A recent paper by Carolyn Snyder published in Nature this week presented a record of global average surface temperature (GAST) reconstructed from a compilation of sea surface temperature records.  This is a great approach and the GAST record, and the techniques used, will be very valuable in efforts to use the geological past to understand how the climate system works.

However, towards the end of the abstract, and the focus of the press release, was a statement about Earth System Sensitivity (the temperature change in response to CO2 change – subtly different to equilibrium climate sensitivity – see here for detail).  Earth system sensitivity, unlike equilibrium climate sensitivity, includes the action of all climate feedbacks (except those relating to the carbon cycle) and is very dependent on the background climate state.  Synder calculated that ESS was around 9 oC per CO2 doubling over the last 800 thousand years.  As discussed here, this is not a relevant measure of how our future climate system will behave because the high ESS during the cold Pleistocene was driven in part by the changes in surface albedo caused by the waxing and waning of the continental ice sheets of North America and Europe – ice sheets that are no longer present on the Earth so are not able to act as a feedback in the future.  ESS for the future is therefore much lower than what Synder calculates.  In a paper last year we @thefosterlab used new Pliocene CO2 data from the boron isotope proxy to show that a future relevant ESS is more likely in the range 2.2 to 5.2 oC per CO2 doubling (which is in good agreement with modelling studies; Lunt et al. 2010).  This suggests that estimates of equilibrium climate sensitivity from the IPCC (1.5 to 4.5 oC per CO2 doubling) do adequately describe the long-term (1000 year) response of the future climate system to CO2 change because the northern hemisphere ice sheets are currently at close to their minimum extent.

The figure above is from Martinez-Boti et al. (2015) showing the Sensitivity Parameter for the Pleistocene (red) and Pliocene (blue) for two different temperature records - mean surface air temperature (left) and sea surface temperature (right).  The Sensitivity Parameter describes the temperature change for a given change in climate forcing.  Panels (a) and (c) show the sensitivity parameter due to CO2 change only.  ESS = this sensitivty parameter multiplied by 3.7 (the radiative forcing from a doubling of CO2).  ECS is approximately the senstivity parameter in panels e and g multipled by 3.7 (because here we also take into account the radiative effect of the ice sheets).  As in the Synder paper we also found a high ESS for the Pleistocene because of the large ice sheets present (ca. 2 x 3.7 = 7.4 oC per CO2 doubling).  The Pliocene however lacked large ice sheets and ESS is about equal to ECS, which both fall in the IPCC range (ca. 1 x 3.7 = 3.7 oC per CO2 doubling).

2016: the first year of 400 ppm? The Foster Lab interviewed on BBC Radio Weather Show

A paper published this week by Richard Betts of the UK Met Office predicted that 2016 will be the first year CO2 will be >400 ppm at the Mauna Loa observatory all year round.  Given the 2-3 ppm yearly increase this signals that CO2 is likely to remain >400 ppm for our lifetime.  This was featured by the BBC website that also included reference to the work fo the Foster Lab and one of our earlier blog posts.  On Wednesday this week Gavin was also interviewed for the Paul Hudson Weather Show about the last time CO2 was 400 ppm.  The show was broadcast this morning at 6am (18/6/16) on BBC Radio Humberside and will be broadcast on BBC Radio Lincolnshire on monday (20/6/16).  You can also listen to the entire show here:


or fast forward to 32 mins to hear Gavin!

The question is now not if we are heading for a Pliocene-like future, but how long will it take to get there...


Ancient marine sediments provide clues to future climate change

 /* Style Definitions */
	{mso-style-name:"Table Normal";
	mso-padding-alt:0cm 5.4pt 0cm 5.4pt;
  Fossil planktonic foraminifera (33-50 million years old) from Tanzania by Paul Pearson, Cardiff University is licensed under a Creative Commons Attribution 4.0 International License

Fossil planktonic foraminifera (33-50 million years old) from Tanzania by Paul Pearson, Cardiff University is licensed under a Creative Commons Attribution 4.0 International License

A paper led by the fosterlab’s Dr. Eleni Anagnostou and featuring Prof. Gavin Foster is published today in Nature (here). It represents the latest in our efforts to use the geological record to provide information about how the Earth’s climate system operates when significantly warmer than today (e.g. see here).  It was work funded as part of the “descent into the icehouse” project that involved researchers from a number of UK universities, and is part of the NERC life and planet theme.

Here is the press release from the University of Southampton.  And here is a short video that discusses the wider project:

Our principle finding in this new paper is that atmospheric CO2 was around 1400 ppm during the warmest part of the Eocene and approximately halved through the Eocene.  We determined this by using boron isotopes.

Last year Dr. Gordon Inglis from Bristol who is also involved in this work published this paper that better demonstrated that: (a) the early Eocene (~53 million years ago) was warm (the best estimates, here, are at +14 oC compared to the pre-industrial); (b) there was around 8 oC of cooling for the high latitude surface ocean and around 2 oC of low latitude cooling through the Eocene (from ~53 to ~34 million years ago).

We know from earlier work that some of the warmth of the Early Eocene was caused by changes in the position of the continents, vegetation change, and the lack of any ice sheets on the South or North poles.  This leaves around 9 oC of the early Eocene warmth to be explained.  CO2 is a potent greenhouse gas (see here for example; ) and when its concentration is doubled in the atmosphere (www.ipcc.ch) the Earth should warm by 1.5 to 4.5 oC.  Levels of CO2 of 1400 ppm is ~5 times the pre-industrial concentration of atmospheric CO2. This means, if the Earth’s climate system really behaves like we think, the early Eocene should be between 4 and 11 oC warmer than the pre-industrial (when CO2 was 280 ppm).  The 9 oC warmth of the early Eocene is thus entirely explained by the enhanced greenhouse effect due to the higher CO2 at the time. This finding also supports our earlier work that suggests the sensitivity of the climate system to forcing from CO2 doesn’t depend hugely on climate state. 

We also know from previous work that surface temperature in high latitude regions tend to change more than the global mean, and low latitude regions tend to exhibit a more muted change than the mean.  Given the nature of this “polar amplification”, we further demonstrated in this new paper that the CO2 drop through the Eocene (from 53 to 34 million years ago) was sufficient to drive the observed (2 oC low latitude and 8 oC high latitude) cooling in sea surface temperature.

These two findings (that CO2 drove much of the warmth of the Eocene, and its decline drove much of the cooling through the Eocene) confirm that not only are the IPCC estimates of climate sensitivity consistent with the geological record, but that CO2 change was a major player in driving the switch between the Cretaceous/early Cenozoic greenhouse climate state to the late Cenozoic ice house climate we currently find ourselves.   The next question of course, and one we are working on, is why did CO2 decline through the Eocene? A recent paper in Science by MacKenzie et al. suggest it’s all about volcanoes and their emissions of CO2. Whatever the cause, its only with CO2 records like we present here we will be able to provide a deeper understanding of the role of CO2 in natural and anthropogenic climate change.