The role of oscillating modes on the climate system

Supervisors: Dr Adrian Matthews and Dr Manoj Joshi.

The problem: The Earth's climate system contains many oscillating phenomena, usually called 'modes of variability'. Examples of these are the North Atlantic Oscillation (NAO), which influences day-to-day weather in Europe, or the Madden-Julian Oscillation (MJO; Zhang, 2005), which modulates tropical weather systems over timescales of a few weeks with further impacts on the extratropics (Matthews et al., 2004), to the El Nino-Southern Oscillation (ENSO), which can affect climate globally over months to years.

The conventional way of defining oscillating modes and their impacts is in terms of anomalies or perturbations from a climatological average. In this framework a single cycle has a positive phase and a negative phase, rather like a sine wave, so by definition, an average over one (or many) cycle(s) is zero, implying that these modes do not alter the long-term average climate. However, in a more complete framework, non-linear interactions can cause even oscillating modes to have an effect on climate when averaged in time; these oscillatory modes can then become integral to defining the mean climate.

The research: This project will address potentially important non-linear contributions of oscillatory modes to the mean climate. The analysis will be carried out using global observational data sets of the atmosphere-ocean system¹ and by designing and running numerical experiments with a global climate model². A framework will be developed to objectively quantify the impact that individual modes such as El Nino have on the mean climate system. This framework will then be used to attribute errors in mean climate simulation to errors in simulating specific weather/climate modes or phenomena such as El Nino (Bell et al., 2009) or the MJO. For example, the framework will allow us to make statements such as 'If there was no MJO, the jet stream over North America would be XXX m s^-1 weaker, with YYY consequences for weather over North America and Europe' (where XXX and YYY are not zero!), or 'The error in simulating El Nino in the climate model led to an error in the mean climate such that the mean temperature over AAA was BBB degrees Celsius colder.' Such statements can then be used to guide improvements of climate model formulation.

¹ERA-Interim atmospheric reanalysis, ECCO ocean reanalysis, satellite precipitation and sea surface temperature data sets, etc.

²Intermediate General Circulation Model (IGCM), a fast, stripped-down climate model that is well designed for carrying out numerical experiments (Forster et al., 2000).

Requirements, training and opportunites: We seek an enthusiastic, pro-active student with strong scientific interests and self-motivation. They will have at least a 2.1 honours degree in physics, mathematics, meteorology or oceanography or another branch of environmental science with good numerical ability. Experience of a programming language such as python, FORTRAN or matlab will be advantageous. They will be trained in meteorological, oceanographical and climate theory, and in the theoretical and practical aspects of computer modelling. The student will have the opportunity to present their work at an international conference. This project will suit an applicant intending to start a scientific career in meteorology, oceanography or climate science.

Further reading:

Bell CJ, Gray LJ, Charlton-Perez AJ, Joshi MM, Scaife AA, 2009: Stratospheric communication of El Nino teleconnections to European winter. J. Climate, 22, 4083-4096.

Forster PMde~F, Blackburn M, Glover R, Shine KP, 2000: An examination of climate sensitivity for idealised climate change experiments in an intermediate general circulation model. Climate Dyn., 16, 833-849.

Matthews AJ, Hoskins BJ, Masutani M, 2004: The global response to tropical heating in the Madden-Julian Oscillation during northern winter. Quart. J. Roy. Meteorol. Soc., 130, 1991-2011.

Zhang C, 2005: Madden-Julian Oscillation. Rev. Geophys., 43, RG2003, doi: 10.1029/2004RG000158.

The animation shows the cycle of the Madden-Julian Oscillation (MJO), with global maps of cloudiness/rainfall anomalies every day through its 48-day period. Blue colours correspond to regions where there is more precipitation than usual, red colours to where it is drier than usual. The way this particular version of the MJO life cycle has been mathematically constructed means that if you sum the cloudiness/rainfall anomalies over the entire MJO cycle, the positive values exactly cancel the negative values, and the net result is zero. Hence this type of analysis gives no information on the effect of the MJO on to the mean climate. Though, of course, it gives very useful information on to which regions may expect wet or dry conditions over the next few weeks. The aim of this PhD project is to move beyond 'linear' representations as shown in this animation, and to look at the nonlinear interactions that do not cancel out, whereby the MJO (and other oscillating modes such as the NAO, ENSO etc.) do have an effect on the mean climate.