ProfessorAndrew Parnell

Climate and Weather Data Science

I work on the past, present and future behaviour of climate and weather extremes on a national and continental scale. My research in this field is funded through multiple sources, including the new €41.3m Research Ireland Co-Centre in Climate, Biodiversity and Water, where I am Deputy Director and lead the Evidence Synthesis platform. Recent funding from the Marine Institute (the A4 project) researches sea level rise in Ireland and the US east coast. The SFI Predict project covers environmental monitoring in Dublin bay. The work under this theme has led to publications in Science, Nature Communications, and Proceedings of the National Academy of Sciences. The research involves a substantial computing element, and combine ideas in stochastic processes, state-space modelling, and extreme value theory applied to cutting edge climate data sets.

Bayesian Machine Learning

I recently held an SFI Career Development Award which funds 4 PhD studentships in my research group to work on Bayesian aspects of probabilistic machine learning. I am particularly interested in developing Bayesian analogues of common machine learning techniques such as random forests and neural networks and exploring their computational and theoretical limits. A key challenge is developing computationally tractable approaches which can be adapted to novel data types including multivariate, spatio-temporal, and unsupervised data. The project has 4 industrial partners. I have continued this work through funding from the SFI Insight Centre for Data Analytics.

Machine Learning in Agriculture

As a co-Principal Investigator in the Nature+Energy project funded by SFI (through the MAREI centre for Marine, Energy and Climate) I lead a work package that uses artificial intelligence techniques to monitor biodiversity at wind farm sites. Through the €8m funded EU InnoVar project I lead several work packages looking at the impact of climate change on the resilience of wheat varieties. In a similar project funded through the Department of Agriculture, Food and the Marine (Mycotox-I) I am applying AI techniques to determine the spread of mycotoxins into oat varieties at Ireland’s farms.

Quantitative ecology and biodiversity

My interest in statistical ecology concerns Bayesian compositional modelling of data from stable isotope experiments on ecosystems. Through this research we have developed the widely-used R software packages SIAR and simmr (> 3000 citations and 40k downloads since 2010).