Ed Hawkins

Ed Hawkins is a post-doctoral researcher working with Rowan Sutton at the University of Reading, and his research has two aspects related to decadal prediction. "Firstly, we have separated and quantified the uncertainty in the CMIP3 climate projections into its three sources, i.e. model uncertainty scenario uncertainty, and internal variability. The importance of each source can then by examined for different lead times, regions, as well as different spatial and temporal averaging. This provides a possible basis for assessing the potential value of new investments in climate science. We demonstrate that, on regional scales, investment in climate science could potentially deliver significant improvements in predictions, especially for the next few decades. Secondly, we are estimating 'climate relevant singular vectors', which are defined as the small ocean perturbations which grow the most rapidly with time. These singular vectors could be used to design efficient ensembles in a decadal climate forecast system by optimally perturbing the initial conditions. They also indicate where additional ocean observations would be of most benefit to constrain climate predictions of the next few decades. The results will therefore aid the design of any future operational decadal climate forecast system, and inform decisions on observational programmes."