We are currently transitioning to a new website, please check our new webpage
Climate-carbon cycle
The oceans are a large sink for anthropogenic carbon
dioxide (CO2) and the processes responsible for the
absorption of CO2 by the ocean are influenced by, and
sensitive to, the atmospheric CO2 concentration, changes
in atmospheric temperature and winds, ocean temperature,
ocean circulation, and ocean nutrient supply leading
therefore to several potential oceanic feedbacks.
I am currently quantifying CO2 uptake by the ocean,
investigating the sensitivity of the marine carbon cycle
and acidification levels to processes such as surface
boundary conditions (e.g, pCO2, SST and freshwater
fluxes), ocean mixing, sea-ice extent, and ocean
circulation and deriving new dynamical constraints to
diagnose the vertical structure of the dissolved inorganic
carbon (DIC), or any other biogeochemical tracer in the
oceans (with David Marshall and Gideon Henderson). These
different projects involve theoretical and modeling tools
including adjoint-based methods, the latter providing new
insight on the amplitude and spatial pattern of CO2 uptake
by the ocean and future acidification levels.
Air-Sea carbon flux evaluated from an ocean general circulation model (MITgcm) with a biogeochemical component.
Relevant publications:
Atlantic Predictability
In order to estimate the predictability limits of
Atlantic climate, we explore the fastest growing
perturbations leading to the amplification of large-scale
meridional overturning circulation (MOC) and sea surface
temperature (SST) anomalies in an ocean general
circulation model. The analysis of the optimal
perturbations provides with information on the sensitivity
of Atlantic ocean to perturbations and on the error growth
(uncertainties) in the model and therefore the implied
limits on predictability. The results showed an
amplification of MOC anomalies on a timescale of 19 years
when the perturbations are constrained to the upper ocean
compared to 7 years when the perturbations are permitted
over the entire ocean depth. These results indicate that
predictability experiments in which only the atmospheric
state is perturbed (equivalent to perturbing the upper
ocean only) may lead to an overestimate of the ocean
predictability time. In addition, optimal perturbations
of upper ocean temperature can be amplified a factor of
1.6 after 15 years in the northern part of the basin. The
ocean actively participates in the amplification of the
anomalies rather than just integrating over the
atmospheric forcing however it is not necessarily
correlated with the growth of MOC anomalies.
Moreover, I am constructing an empirical model using observations over the past several decades in order to predict interannual fluctuations of the North Atlantic sea surface temperature anomalies. Preliminary results show that the model has predictive skill up to 3 to 4 years, forecast skill on longer timescales is greatly reduced. Additionally, the model is used to diagnose how different ocean regions impact the predictions.
Optimal perturbations of upper ocean temperature and salinity in the ocean general circulation model.
Relevant publications:
North Atlantic Variability
In the North Atlantic, the large-scale ocean circulation
transports heat from low to high latitudes. The strength
of the meridional overturning circulation (MOC), similarly
to Atlantic climate variability, is believed to vary over
a wide range of time scales. We explored the sensitivity
and variability of the Atlantic MOC by investigating the
non-normal ocean dynamics in an ocean general circulation
model. Although the linearized dynamics are found to be
stable, initial temperature and salinity anomalies can
generate a large amplification of MOC anomalies after
several years. Deep density anomalies in the northern part
of the basin are found to excite the largest amplification
of MOC anomalies after about 7 years. The growth of MOC
anomalies can be understood by examining the time
evolution of deep zonal density gradients related to the
MOC via the thermal wind relation. The propagation of the
density anomalies, which depends on the mean flow velocity
and the mean density gradient, determines the growth time
scale of the MOC anomalies.
Using simple box models of the circulation, we found that stochastic salinity anomalies were far more efficient than temperature anomalies at sustaining MOC variability; and that density gradients at high latitudes were more efficient than low-latitude ones at inducing MOC variability over a wide range of frequencies.
Meridional overturning circulation streamfunction anomalies at time of maximum amplification.
Relevant publications:
Intra-seasonal and Interannual Tropical Sea Surface Temperature Variability
The interannual variability of ocean tropical sea surface
temperature (SST) is commonly believed to be dominated by
coupled ocean-atmosphere feedbacks. Even though
atmospheric excitation of the observed tropical SST
variability may be significant, the precise role of
ocean-only dynamics in this variability remains unknown
due to lack of observations. A few studies invoked ocean
dynamics including the meridional overturning circulation
and wave instabilities to explain this variability.
Recently, we proposed a mechanism relying on the
excitation of deep salinity anomalies in the vicinity of
the western boundary (for example by mesoscale
eddies). After a rapid geostrophic adjustment, the
propagation of the coastal and equatorial Kelvin waves
creates SST anomalies in the tropical Atlantic on a
timescale of 3 to 4 years. This mechanism is solely driven
by the ocean dynamics (without the participation of the
MOC nor tropical wave instabilities) and does not rely on
any atmospheric feedback.
On shorter time scales, sea surface temperatures in the tropical can reach unusually large values in localized regions (tropical "hot spots"). We analyzed possible feedbacks and perturbations (deterministic or stochastic) responsible for their formation using a 1D idealized coupled ocean-atmosphere model.
Hovmoller diagrams of equatorial upper ocean temperature and salinity anomalies.