Profile
Dr Stefan Siegert
Senior Lecturer
Telephone: 01392 724058
Extension: (Streatham) 4058
My research is on statistical forecasting in complex systems, particularly weather and climate. I am especially interested in how to best use information from computer simulations to say something about the real world.
- I am currently a Lecturer in Mathematics. I teach Basic Statistics, Statistical Inference, Advanced Statistical Modeling, and Applied Data Science to students at all stages.
- I have previously worked in the Real Projections project (NERC) where leading UK climate researchers and statisticians work together to provide robust spatial projections of real-world climate change. I have developed statistical methodology for making reliable spatial projections of future climate change based on historical observations and ensembles of climate simulation models.
- I have worked for the SPECS project (FP7) for the development of seasonal-to-decadal climate forecasts for European climate services. My role in this project was to develop new statistical methodology and software for forecast quality assessment.
- I am developing and maintaining the R package SpecsVerification for evaluating the quality of weather and climate forecasts.
- My Dr. rer. nat. (PhD equivalent) in Physics was awarded with distinction by the Technical University of Dresden, Germany (Title: "Rank Statistics of Forecast Ensembles").
Student supervision and PhD projects: I am interested in supervising students at all stages, from 1st year up to and including PhD students, on projects related to mathematical-statistical methodology (including machine learning, data science, artificial intelligence), in application areas related to environmental science (such as weather forecasting, climate modelling, environmental extremes). If you have an idea for a project in these areas that you would like to work on, please contact me. Here are some specific projects I am currently offering:
- Developing spatio-temporal machine learning models for the improvement of medium-range weather forecasts
- Bayesian modelling for combining climate simulation and observation data to produce reliable projections of future extremes
- Physical-statistical modelling for precipitation radar nowcasting
- Bayesian methods for quantifying and reducing uncertainty in life cycle assessment of renewable energy