Adam Scaife

Met Office Hadley Centre

Geoffrey Vallis, Ruth McDonald, David Fereday, Marika Holland, Fred Kucharski, Franco Molteni

The North Atlantic Oscillation (NAO) is the dominant pattern of variability on timescales from days to decades in the Atlantic region of the Northern Hemisphere and climate change patterns project positively onto the NAO. Despite this, estimates of the NAO response to anthropogenic forcing are poorly constrained. Similarly, the magnitude and nature of internal variability in the NAO and the role of coupling with the ocean and the stratosphere are only just beginning to be understood. We therefore propose the following analysis to reduce some of these uncertainties for the IPCC AR4 and to reach the latest consensus using current models.
a) Climate change and climate variability in the NAO. Timeseries of gridpoint and pattern based measures of the NAO will be used to determine a multi-model mean response of the NAO to anthropogenic climate forcing and its effect on surface temperature. Uncertainty in the expected NAO changes due to model formulation and internal variability will be assessed, and any changes in NAO structure will be examined. Comparison will also be made with recent observed NAO changes. b) The link between changes in the NAO and eddy statistics. The IPCC Beijing workshop concluded that intercomparison of mid-latitude cyclones in models was difficult because different tracking methods have been used. Here we will apply a single mid-latitude storm tracking routine to data from each model to analyse changes in frequency, intensity and life-times of mid-latitude storms. A model inter-comparison of changes in storms will also be carried out and the changes in storms will be linked to changes in the NAO. c) The role of boundary conditions. There is current debate about the role of lower and upper boundary conditions on the NAO from the sea-surface and the stratosphere. We will carry out a preliminary search for links between these boundary conditions and changes in the modelled NAO. To carry out this analysis we require monthly-mean sea level pressure fields, monthly mean land and sea-surface temperature fields, high frequency sea-level pressure fields (12 hour intervals or higher frequency), monthly mean lower stratospheric zonal winds. Data are required from control and at least one scenario experiment.