HR: 0800h
AN: OS41A-0594    [Abstracts]
TI: The Probability of the Collapse of the Thermohaline Circulation in an Intermediate Complexity Model
AU: * Challenor, P
EM: P.Challenor@noc.soton.ac.uk
AF: National Oceanography Centre, Southampton, European Way, Southampton, SO14 3ZH United Kingdom
AU: Hankin, R
EM: R.Hankin@noc.soton.ac.uk
AF: National Oceanography Centre, Southampton, European Way, Southampton, SO14 3ZH United Kingdom
AU: Marsh, R
EM: R.Marsh@noc.soton.ac.uk
AF: National Oceanography Centre, Southampton, European Way, Southampton, SO14 3ZH United Kingdom
AB: If the thermohaline circulation were to collapse we could see very rapid climate changes, with North West Europe becoming much cooler and widespread impacts across the globe. The risk of such an event has two aspects: the first is the impact of a collapse in the circulation and the second is the probability that it will happen. In this paper we look at latter problem. In particular we investigate the probability that the thermohaline circulation will collapse by the end of the century. To calculate the probability of thermohaline collapse we use a Monte Carl method. We simulate from a climate model with uncertain parameters and estimate the probability from the number of times the model collapses compared to the number of runs. We use an intermediate complexity climate model, C-GOLDSTEIN, which includes a 3-d ocean, an energy balance atmosphere and, in the version we use, a parameterised carbon cycle. Although C-GOLDSTEIN runs quickly for a climate model it is still too slow to allow the thousands of runs needed for the Monte Carlo calculations. We therefore build an emulator of the model. An emulator is a statistical approximation to the full climate model that gives an estimate of the model output and an uncertainty measure. We use a Gaussian process as our emulator. A limited number of model runs are used to build the emulator which is then used for the simulations. We produce estimates of the probability of the collapse of the thermohaline circulation corresponding to the indicative SRES emission scenarios: A1, A1FI, A1T, A2, B1 and B2.
DE: 1605 Abrupt/rapid climate change (4901, 8408)
DE: 1626 Global climate models (3337, 4928)
DE: 1630 Impacts of global change (1225)
DE: 1635 Oceans (1616, 3305, 4215, 4513)
DE: 4962 Thermohaline
SC: Ocean Sciences [OS]
MN: Fall Meeting 2005