Research Projects

Early Pliocene Climate

As new data emerges, our interpretation of the early Pliocene climate continually evolves. More and more data from the last few years indicate a persistence of a significantly reduced zonal temperature gradient in the equatorial Pacific, which is associated with climate conditions usually referred to as a "permanent El Niño". We have collected together data from a variety of regions in both Pacific and Atlantic oceans to reconstruct the meridional temperature gradient in the ocean during the early Pliocene. Our results indicate a substantial reduction in the meridional temperature gradient from the equator to the subtropics (in comparison with today's climate) and a large poleward expansion of the warm-water pool in the tropical Pacific.

I have used an atmospheric general circulation model to look at the implications of a reduced meridional temperature gradient ( by forcing it with hypothetical surface temperature boundary conditions, representative of our new understanding of the early Pliocene). I've explored the changes in the global circulation and precipitation patterns, along with inferences about the ocean heat transports. The Walker circulation is virtually removed by the imposed SST boundary conditions. The latitudinal extent of the Hadley cell increases only slightly, but the strength of the Hadley circulation is substantially reduced, which results in a weakening of the poleward heat transport by the atmosphere in the low- to mid- latitudes.

In his previous research, my supervisor (Alexey Fedorov) has found that oceans models predict a reduction in the ocean heat transport in the early Pliocene, rather than the warming my results would indicate. The issue of poleward heat transport in the early Pliocene has been called the "Pliocene Paradox" (Fedorov et al, 2006), and I am currently looking for methods to resolve this paradox.

Ocean Parameter Uncertainty

Atmosphere-ocean general circulation models are the best tools available to provide policy-relevant predictions of the climate's response to a change in the amount of atmospheric CO2. They include parameterisations of physical processes that cannot be resolved explicitly. There is uncertainty in the numerical values (parameters) involved in the parameterisation of ocean physics. This is called ocean parameter uncertainty. My PhD work was the first investigation of the effects of ocean parameter uncertainty in a complex climate model.

I created a database of ocean parameters and their uncertainty ranges for a complex model (HadCM3) through expert consultation. A perturbed physics ensemble was then created from the highest priority parameters. That ensemble represents an upper bound of ocean parameter uncertainty. Each ensemble member has been run to a preindustrial quasi-steady state. I looked at the effects of ocean parameter uncertainty on the long-term climate state: there is surface temperature uncertainty of greater than 2.5oC at high latitudes.

I then performed a climate change experiment with CO2 increasing by 1\% per year with the ensemble. The effects of ocean parameter uncertainty on the climate change signal can be detected above natural variability. The global mean effects are compared to those seen from other sources of uncertainty. I've investigated the regional patterns of the effects on the surface climate change signal: they are over 100\% of the ensemble mean signal in regions of strong vertical ocean heat transfers. I also looked at the effects on the vertical structure of the ocean.

Analysis of the evolution of ensemble spread with increasing CO2 reveals an anti-correlation between the preindustrial global mean temperature and the transient climate response. This is an interesting feature of the ensemble, because the majority of climate change studies are designed to completely remove any such dependence. The relationship is mediated by variations in ocean heat uptake processes and climate feedbacks in approximately equal measure.