Jobs

Projects for Students

We welcome expressions of interest from potential DPhil (=PhD) candidates who would like to join our group. UK students can be funded through the University of Oxford's Doctoral Training Partnership in Environmental Research. There may be alternative funding opportunities for non-UK candidates.

We also welcome expressions of interest from undergraduate students who are interested in working in our group for their MPhys project or during the summer vacation.

MPhys project "Retrospective forecasts of winter and summer large-scale circulation changes during the 20th Century"

Forecasts of seasonal-mean anomalies of the climate using dynamical atmosphere-ocean circulation models based on the laws of physics are now routinely made at many operational meteorological forecast centres around the world. Such seasonal predictions provide estimates of seasonal-mean statistics of weather, typically up to four months ahead. In order to estimate how skilful seasonal forecasts are, the models are run in so-called retrospective forecast mode. This means that a period in the past, that can be verified with observations, is predicted using only information that would have been available at the time of the start of the forecasts. This project works with a long seasonal retrospective forecast data set that covers the entire 20th Century. The unusually long model forecast record allows the analysis of dominant modes of large-scale atmospheric variability, predictability and their changes on multi-decadal time scales. The student is going to analyse the existing ensemble forecast data set and compare it with a proxy data set of global observations.
Supervisors: Dr. Antje Weisheimer, Prof. Tim Palmer, Dr. David MacLeod, Dr. Chris O’Reilly (all AOPP)
Contact: Dr. Antje Weisheimer

2016 Summer Programme

Atmospheric, Oceanic and Planetary Physics will run a summer research programme for undergraduate students. We anticipate taking about eight students from the second year and above.

Students will work with a supervisor in the Department, usually a postdoctoral researcher or lecturer, on a self-contained research project. Students are encouraged to take part in Departmental life, joining researchers for coffee, discussions and seminars.

The projects run for up to 10 weeks, nominally from late June to August. The duration may be shorter to accommodate summer travel. Students will be paid via a stipend (Usually £200 - £300 per week depending on hours worked). The project is full-time but hours can be discussed with your supervisor. Note that foreign applicants must already have the right to work in the UK. See here for more information.

Available projects in the Predictability of Weather and Climate Group:

Large-scale atmospheric circulation changes during the 20th Century in re-analyses and seasonal climate forecasts

This project looks at large-scale atmospheric circulation changes that occurred during the last century. A global new data set of the climate evolution of the 20th Century, the ERA-20C reanalysis from the European Centre for Medium-Range Weather Forecasts (ECMWF) will be studied together with retrospective climate forecasts of winter and summer seasons during the same period made with the atmospheric component of ECMWF’s seasonal forecast model. An initial analysis of predictive skill of the North Atlantic Oscillation (NAO), which is the dominant mode of climate variability in the Euro-Atlantic area, indicated a high level of multi-decadal variability both in the NAO and in the skill to simulate these changes with the seasonal forecast model.

Forecasts of seasonal-mean anomalies of the climate using dynamical atmosphere-ocean circulation models based on the laws of physics are now routinely made at many operational meteorological forecast centres around the world. Such seasonal predictions provide estimates of seasonal-mean statistics of weather, typically up to four months ahead, and are performed as ensembles of forecasts because of the chaotic nature of the atmosphere. In order to estimate how skilful seasonal forecasts are, the models are run in so-called retrospective forecast mode. This means that a period in the past, which can be verified with observations, is predicted using only information that would have been available at the time of the start of the forecasts. This project works with a long seasonal retrospective forecast data set that covers the entire 20th Century.

The summer project will build on existing experience within the group to analyse the reanalysis and retrospective forecast data sets. It is planned to study both extratropical and tropical modes of variability to enhance our understanding of the predictability of the atmosphere in connection with a changing climate. The methodology to do so relies on statistical techniques to analyse and visualise large gridded data sets. Experience in programming and visualisation languages would be of advantage. Knowledge of basic statistical concepts is required to understand the methods and to help interpret the findings. The project will provide opportunities for developing independent research ideas. Results of the project could potentially lead to outputs like publications in the scientific literature.

Supervisors: Dr. Antje Weisheimer, Prof. Tim Palmer, Dr. David MacLeod, Dr. Chris O’Reilly (all AOPP)
Contact: Dr. Antje Weisheimer Antje [dot] Weisheimer [at] physics [dot] ox [dot] ac [dot] uk
Application: Please email Antje Weisheimer with your expression of interest and a short CV. We accept applications by the end of May 2016. Suitable candidates will be interviewed during the first half of June 2016.

Extreme Weather and Circulation Regimes in the Weather@Home Ensemble

Extreme weather events, such as droughts or floods, are often associated with weather regimes, or persistent large-scale atmospheric flow conditions. In general, North Atlantic/European weather regimes are poorly represented in climate models, though globally increasing the resolution of these models can significantly improve the regime structure and statistics. In the Weather@Home (W@H) project, a high-resolution regional model is nested within a lower resolution global climate model. This allows us to ask important questions about the changes in regional weather events that could occur in a changing climate. However, it is not known whether nesting a high resolution model in this way will improve the regime behaviour of the parent model.

The W@H high-resolution model is nested within the lower resolution model used in the climateprediction.net project. In both projects, an ensemble of simulations is produced to allow us to predict the possible range of climate trends over the coming decades. It is important to represent the uncertainty in the simulation due to shortcomings of the model – the climateprediction.net project uses a perturbed parameter ensemble to account for this. There is evidence that the choice of model uncertainty representation can matter. For example, an alternative representation of model uncertainty (stochastic parametrisation schemes) improves the regime behaviour of both simple atmospheric models and comprehensive climate models, whereas the impact of perturbed parameter schemes had a detrimental impact on the regime behaviour of a simple atmospheric model.

This project has two main goals. The student would firstly evaluate the realism of the North Atlantic weather regimes in the W@H data set. Does using a nested high-resolution model improve the regime behaviour of a climate model by representing small-scale processes important for regime transitions and structure? Secondly, the student would investigate how perturbing the parameters in the climate model impacts the observed regime structure.

The student would gain experience of analysing data from complex climate models and using advanced statistical methods to do regime analysis, such as k-means clustering. They would also learn about the importance of representing model uncertainty for making calibrated climate change predictions and different approaches to doing this.

Skills Required

Experience with a data-analysis language such as Matlab or Python would be advantageous, but is not essential.

How to apply

Supervisors: Dr. Hannah Christensen, Dr. Peter Watson, Prof. Tim Palmer (all Predictability of Weather and Climate group)

Contact: Dr. Hannah Christensen: h.m.christensen 'at' atm.ox.ac.uk

Application: Please email Hannah Christensen with a short cover letter explaining your interest in this placement, together with a copy of your CV. Applications will be reviewed on an on-going basis until the position is filled.

Jobs

We are not currently advertising any jobs.