Professor Myles Allen CBE FRS

Abstract:

Estimation methods for the interpolation of sea surface temperature (SST) were discussed. SST data was collected using the Along Track Scanning Radiometer (ATSR). Dynamic estimation of SST was carried out using a recursive estimation algorithm which emulates the Kalman filter. Key steps in the estimation process included removal of underlying mean SST field, characterization of spatial and temporal statistics appropriate to SST in Tropical Pacific, establishment of suitable treatment of daytime and nightime data, and the assimilation of the data over time. The capability of assimilating data over time and dealing with spatially-varying prior model was demonstrated by the study.

Abstract:

Skin sea-surface temperatures from the first Along Track Scanning Radiometer (ATSR) are compared with coincident bulk temperatures from the Tropical Atmosphere Ocean (TAO) moored buoy array in the equatorial Pacific Ocean. The response of the skin-bulk sea-surface temperature difference (ΔT) to variations in wind speed and surface heat flux is examined. The use of remotely-sensed skin temperatures for this purpose is enabled by ATSR's unique design which permits the independent retrieval of ocean skin temperature to an accuracy of 0.3 K. For the four-year period considered (August 1991-August 1995), almost 6000 coincident skin and bulk sea surface temperature (SST) measurements were available; at night, the mean value of ΔT is -0.20 ± 0.46K, with a daytime mean value of +0.05 ± 0.51K. ΔT is found to depend on both net heat flux and local wind speed as predicted by the Saunders [1967] model and other formulations, and an estimate of the Saunders λ parameter is obtained.

Abstract:

The first principal component of Northern Hemisphere sea level pressure, known as the Arctic Oscillation (AO) index, has increased significantly in recent winters, and this change is associated with ~30% of Northern Hemisphere January-March warming. We examine the AO in a model used to detect anthropogenic influence on climate, and find that it exhibits no systematic trend in response to greenhouse gas, sulphate aerosol, or ozone forcing. To test the significance of this discrepancy for anthropogenic climate change detection, we include the spatio-temporal pattern of temperature change associated with the observed AO in the set of forcing-response 'fingerprints' used to account for observed changes, thus separating temperature change associated with the AO from a residual. We find that the detection of a global response to both anthropogenic greenhouse gases and sulphate aerosols is robust to this exclusion of AO-related warming.