skip to content

Cambridge NERC Doctoral Training Partnerships

Graduate Research Opportunities

Lead Supervisor: Peter Haynes, Applied Mathematics and Theoretical Physics

Co-Supervisor: Adam Scaife, UK Met Office

CASE Partner: UK Met Office

Brief summary: 
The factors controlling time variability of jets in simple models will be studied, motivated by the connection between such variability and the ‘signal-to-noise’ paradox which suggests that skilful predictions can be extracted from seasonal to decadal weather and climate forecasts notwithstanding the large spread of predictions across individual ensemble members.
Importance of the area of research concerned: 
The properties of the atmospheric and oceanic circulation mean that useful forecasts of weather on a particular day more than a few days ahead are not possible, but useful long-term forecasts of the weather statistics, e.g. whether a particular month or season being unusually dry or wet, or hot or cold, are possible and indeed centres across the world are demonstrating concrete progress in making such forecasts. Alongside this progress, some fundamental challenges remain. It seems that model simulations show behaviour that is highly unpredictable, yet when ‘historical’ forecasts are investigated, the average of several such simulations appears to contain useful skill in the prediction of the year-to-year variation. This has been described as the ‘signal-to-noise’ paradox [1]. The problem is particularly acute for making forecasts in the North Atlantic/European region and makes it difficult to answer basic questions such as ‘has the decline in Arctic sea ice affected European weather?’ There are hints that this problem will reduce if model resolution increases. But there is little understanding of the precise physics that is important for this aspect of model behaviour.
Project summary : 
The project will be based on the hypothesis that at the heart of the signal-to-noise paradox is the ability of models to predict the intrinsic variability of the mid-latitude atmospheric jet. Such variability is not only a major component of forecast ‘noise’, but its properties potentially affect the response to external forcing, or equivalently to provision of external information, i.e. the forecast ‘signal’ [2]. A simple minimal model of this jet is provided by ‘beta-plane turbulence’ in a single-layer or two-layer flow [3]. The formation of jets in this model has been much studied, but the factors controlling the variability of such jets have not. Natural questions include the effect of spatial resolution on jet variability, and whether any sensitivity to resolution results from the representation of energy transfer from large scales to small scales or vice versa.
What will the student do?: 
The student will investigate variability of beta-plane jets in one and two-layer models, with geometrical constraints and configurations of imposed forcing to capture the important properties of the midlatitude circulation. The primary approach will be by carrying out sequences of numerical simulations, using different dynamical theories to aid formulation of relevant hypotheses. Simulations will be carried out using existing numerical codes, or simple extensions to such codes. New statistical models of turbulence will also be exploited. The relation between properties of simulated jet variability and predictability will be explored as will the relevance of different ideas concerning the signal-to-noise paradox, including the existence, or non-existence, of persistent states. Regular discussion will be held with the CASE supervisor and colleagues in the Monthly to Decadal Forecasting Group at the UK Met Office who will host two or three extended research visits as the project proceeds. The investigation of simple models will feed into ongoing work at the Met Office on resolving the signal-to-noise paradox in state-of-the-art models and vice versa.
References - references should provide further reading about the project: 
Scaife, A. A. and Smith, D. M., 2018: A signal-to-noise paradox in climate science. npj Climate and Atmospheric Science 1, 28.
Cooper, F. C. and Haynes, P. H., 2011: Climate sensitivity via a non-parametric fluctuation-dissipation theorem. J. Atmos. Sci. 68, 937-953.
Scott, R. K. and Tissier, A.-S., 2012: The generation of zonal jets by large-scale mixing. Physics of Fluids, 24(12):126601.
You can find out about applying for this project on the Department of Applied Maths and Theoretical Physics (DAMTP) page.
Department of Applied Mathematics and Theoretical Physics PhD Admissions
Professor Peter Haynes