Climate change is one of the most pressing societal and scientific challenges. Improving our understanding of climate change, in particular concerning its impacts on the country- or even city-scale, is therefore a global research priority. In this PhD project, you will utilise state-of-the-art machine learning techniques to make step-change improvements in our ability to pin down climate change impacts.
Taking an environmental data science perspective, you will work on innovative ways to understand and quantify changes in extreme weather events such as heat waves and floods. A central question will be how moving towards a more sustainable future, in contrast to “business-as-usual” climate change scenarios, could help control increased extreme event occurrences worldwide. We are further interested in related effects on air pollution, a key public health issue in cities, which is also affected by, and impacts on, regional climate change.
For this, you will evaluate the latest generation of Earth system models against Earth observations using machine learning, with a focus on extreme weather and air pollution events; identify physical drivers of model skill to reduce uncertainties in climate change projections; and use the improved projections to devise better, novel climate risk assessments in areas as potentially diverse as human health and ecology.
You will receive comprehensive training in
(a) how to analyse large climate datasets, such as from global Earth system models and NASA satellite observations.
(b) how to use and modify recently developed machine learning algorithms in Python.
(c) how to work on impact analyses and to handle associated uncertainties.
You will be involved in collaborations with colleagues at Imperial College London and the Data Science Institute of the German Aerospace Center in Jena, Germany. Your research will also include a visit to the Climate Informatics Group (https://climateinformaticslab.com/
) in Jena.
This is a PhD project.
The start date of the project is 1 October 2020.
The mode of study is full-time. The studentship length is 3 years with a 1-year registration period.
Acceptable first degree in any quantitative or natural sciences discipline (e.g. Physics, Mathematics, Computer Science, Environmental Sciences, Meteorology, Chemistry). A 1st class degree is desirable next to demonstrable research skills. A keen interest in developing interdisciplinary knowledge is required. Good programming experience and familiarity with some machine learning packages (scikit-learn, TensorFlow etc) would be an advantage, but are not essential. The ideal candidate should be able to demonstrate a keen interest in the physics of the Earth system and in testing out a number of different supervised and unsupervised machine learning algorithms.
The standard minimum entry requirement is 2:1.
1.European Environment Agency. Air quality in Europe – 2019 report, (2019).
2.Runge J, Nowack P, Kretschmer M, Flaxman S, Sejdinovic D. Detecting and quantifying causal associations in large nonlinear time series datasets. Science Advances, (2019).
3.Nowack P, Runge J, Eyring V, Haigh J. Causal fingerprints in climate models and observations. Under review in Nature Communications.4.Joshi M, Hawkins E, Sutton R, Lowe J, Frame D. Projections of when temperature change will exceed 2°C above pre-industrial levels. Nature Climate Change 1, 407-412 (2011).