Dr Jill Johnson
- Position: Research Fellow
- Areas of expertise: Statistics; Applied statistics; Uncertainty quantification; Computational statistics; Bayesian inference; Extreme value theory.
- Email: J.S.Johnson@leeds.ac.uk
- Location: 11.121 Earth and Environment Building
I am an applied statistician working as a research associate in the aerosol research group at the Institute for Climate and Atmospheric Science, University of Leeds.
I studied at Newcastle University, graduating in 2010 with a PhD in Extreme Value Theory: “Modelling Dependence in Extreme Environmental Events”. My PhD work involved a spatial approach to modelling extreme values and extremal dependence for a complex data set of daily rainfall and mean wind speed observations over 20 years covering 25 sites across the UK. Since then, I worked as a research statistician for 3 years at the government’s Food and Environment Research Agency, looking at uncertainty quantification and risk analysis for applications including food safety and land-use change.
I joined the aerosol research group at Leeds in December 2012, and my work focusses on the quantification and constraint of key uncertainties in complex aerosol and cloud models.
My research interests lie in the application of statistical methods to explore the behaviour of complex environmental processes and systems, with a focus on modelling extreme events and uncertainty quantification in complex models.
My research involves working with atmospheric models on many different scales, including a model that simulates an individual cloud, an LES (Large Eddy Simulation) model that simulates a cloud field and a large aerosol-climate model that simulates the global distribution of aerosols in the atmosphere. I apply and develop statistical methodologies to quantify the key parametric uncertainties in these complex atmospheric models and constrain these uncertainties using real-world observations.
This work involves using statistical experiment design techniques, surrogate modelling (emulation) approaches and sensitivity analysis to quantify the important uncertainties, and history matching techniques to compare the models against observations. The practical application of these techniques to real-world atmospheric models provides many challenges, and in my research, I aim to develop further statistical methodologies and tools to explore and overcome the following key challenges:
- How to constrain an uncertain and at times structurally incorrect complex global model using disparate observational data (e.g., measurements from aircraft, ground stations or satellites) with sparse spatial and temporal coverage.
- How to emulate a multi-dimensional model output surface that is non-stationary. Natural systems are inherently noisy, they may be chaotic, and they often contain discontinuities.
- How to evaluate the robustness of a complex model. How can we determine the usefulness of predictions given uncertainty in the underlying processes?
- Can environmental simulators predict extreme system behaviour in the presence of uncertainty?
- PhD in Statistics (Newcastle Univ.; 2010): 'Modelling Dependence in Extreme Environmental Events'
- MMATHSTAT: Master of Mathematics and Statistics (Newcastle University; 2004)
- Royal Statistical Society
- American Geophysical Union
Research groups and institutes
- Institute for Climate and Atmospheric Science
- Atmospheric Chemistry and Aerosols