Carbon–climate feedbacks have the potential to significantly
impact the future climate by altering atmospheric CO
By modifying the future atmospheric CO
We show that simulated carbon–climate feedbacks can significantly impact the
onset of undersaturated aragonite conditions in the Southern and Arctic
oceans, the suitable habitat for tropical coral and the deepwater saturation
states. Under the high-emissions scenarios (RCP8.5 and RCP6), the
carbon–climate feedbacks advance the onset of surface water under saturation
and the decline in suitable coral reef habitat by a decade or more. The
impacts of the carbon–climate feedbacks are most significant for the medium-
(RCP4.5) and low-emissions (RCP2.6) scenarios. For the RCP4.5 scenario, by
2100 the carbon–climate feedbacks nearly double the area of surface water
undersaturated with respect to aragonite and reduce by 50 % the surface
water suitable for coral reefs. For the RCP2.6 scenario, by 2100 the
carbon–climate feedbacks reduce the area suitable for coral reefs by 40 %
and increase the area of undersaturated surface water by 20 %. The
sensitivity of ocean acidification to the carbon–climate feedbacks in the low
to medium emission scenarios is important because recent CO
Ocean acidification, the measurable consequence of increasing atmospheric
CO
Future carbon–climate projections generally show global warming alters the
efficiency of carbon dioxide (CO
For this study, we consider future projections of atmospheric CO
The structure of the paper is as follows. In the next section, we briefly
describe the ESM used and the simulations performed. In the subsequent
section, we present the results from the historical and the future
simulations. We show that the carbon–climate feedbacks accelerate ocean
acidification in all future emissions scenarios. Importantly, it is in the
low and medium emissions scenarios where ocean acidification is most impacted
by the carbon–climate feedbacks. For the low and medium emissions scenarios,
ocean acidification is sensitive to the additional CO
In this study, we used the CSIRO Mk3L Carbon Ocean, Atmosphere, Land (COAL)
ESM
The land module (CABLE) with CASA-CNP
The ocean component of the ESM has a resolution of 2.8
The ESM was spun up under pre-industrial atmospheric CO
From the spun-up initial climate and carbon state, the historical simulation
(1850–2005) was performed using the historical atmospheric CO
The future simulations were repeated using the CO
The COAL simulations of the future carbon–climate feedbacks made with RCP8.5
and RCP2.6 were discussed by
In all our simulations, the vegetation scenario used by
An assessment of the simulated carbon and climate was made in
Year 2002
surface aragonite saturation state
For the various RCP scenarios, the atmospheric CO
For 2002, we compare the simulated annual mean surface ocean aragonite
saturation state to the values estimated from GLODAPv2 observational dataset
For the various RCP scenarios, the cumulative difference in
For the future, the ESM simulated higher atmospheric CO
Since the differences in land and ocean carbon uptake between the
corresponding EP and CP simulations reflect their different atmospheric
CO
For the various RCP scenarios, the global decadal averaged surface temperature change from the present day for the EP (dotted) and CP (solid lines) simulations.
For the
various RCP scenarios, the surface ocean aragonite saturation state for the
decade of the 2090s. CP simulations (left column):
For the ocean, our ESM feedback parameters (
For the land, our ESM warming feedback (
A recent analysis of 11 ESMs of the RCP8.5 scenario
The higher atmospheric CO
For the various RCP scenarios, the CP simulations (solid lines) and
their corresponding EP simulations (dotted lines) for
For the
year 2100, the change in the depth of the aragonite saturation horizon
between the emission simulations (EPs) and the concentration simulations
(CPs) for
Figure
For undersaturated aragonite surface water, the EP simulations all display a
similar evolution to the corresponding CP simulations but with a more rapid
onset of undersaturated conditions. For RCP8.5, the EP simulation leads the
CP simulation by about 5 years. For RCP6, the EP simulation leads the CP
simulation by about 10 years. For RCP4.5, the lead is nearly 20 years. While
for RCP2.6, there is a similar 20-year lead in the emissions simulation but
the area of undersaturated water is small due to the low atmospheric
CO
For all scenarios, the carbon–climate feedbacks accelerate the onset of undersaturated aragonite conditions. However, it is in the medium- to low-emissions scenarios (RCP4.5 and RCP2.6) where the differences between EP and CP simulations are greatest and, hence, where the carbon–climate feedbacks are most significant.
For the surface ocean area suitable for coral reefs, the evolution of the EP
simulations is similar to the corresponding CP simulations, but, again, they
lead the CP simulations. The more rapid onset of ocean acidification produces
the largest difference in the RCP4.5 scenario where, by the end of the
century, the suitable area for coral reefs in the EP simulation (18 %) is
less than half the CP simulation (37 %). Under the high-emissions
scenarios (RCP6.0 and RCP8.5), there is no suitable habitat for coral reefs
by 2100, with the time of disappearance occurring 15 and 6 years earlier in
the EP simulations than in the CP simulations for RCP6 and RCP8.5,
respectively. With the highest emission scenario (RCP8.5), there is such a
large and rapid release of CO
The differences between the EP and CP simulations extend into the ocean
interior. By 2100, the EP simulations show a shoaling of the aragonite
saturation horizon (depth of where the aragonite goes undersaturated) than
the corresponding CP simulations (Fig.
Here we employ an ESM to investigate the potential consequences of carbon–climate feedbacks on the future evolution of ocean acidification. With the emissions-driven (EP) simulations, we show that carbon–climate feedbacks can significantly accelerate the future rate of ocean acidification. Therefore, accounting for carbon–climate feedbacks is important in projecting future ocean acidification impacts and trajectories.
The other salient point is that carbon–climate feedbacks have the greatest impact
under the medium- to low-emissions scenarios (RCP4.5 and RCP2.6). For the
RCP4.5 scenario, the carbon–climate feedbacks nearly double the area of
undersaturated surface water, and halve the area of surface water suitable
for coral reefs by the end of the century. While less dramatic, in the RCP2.6
scenario, the carbon–climate feedbacks reduce the area suitable for coral
reefs by 40 % and increase the area of undersaturated surface water by
20 %. If we aim to track a low-emissions scenario
Here, we have only considered ocean acidification impacts, but carbon–climate
feedbacks also lead to faster global warming. This would accelerate impacts
like ocean warming and deoxygenation
While it is natural to compare the impact of climate-carbon feedbacks on
ocean acidification to previous estimates of intermodel variability from CP
simulations, we emphasise that all the CP simulations prescribe the future
atmospheric CO
Regionally,
The World Climate Research Program (WCRP) identified
Recent studies show the carbon–climate feedbacks are dominated by the land
carbon cycle response to warming
An ESM with a weaker land
The large differences in the carbon–climate feedbacks are not only a key
uncertainty in climate projections
Correspondence and requests for materials should be
addressed to Richard J. Matear (email:
richard.matear@csiro.au). Data are available on request and a persistent URL will be created on the CSIRO data portal site
The authors declare that they have no conflict of interest.
This article is part of the special issue “The Ocean in a
High-CO2 World IV”. It is a result of the 4th International Symposium on the
Ocean in a High-CO
Richard J. Matear and Andrew Lenton would like to acknowledge the financial support of CSIRO Ocean and Atmosphere and the CSIRO Decadal Climate Forecasting Project. Edited by: Jean-Pierre Gattuso Reviewed by: two anonymous referees