My research is about working with imperfect data and models to better monitor and model the environment. 

To do this I focus on developing hybrid approaches of statistical methodology and machine learning, working across mathematical and biological disciplines on environmental problems. I enjoy working collaboratively with experts, including ecologists, climate and remote sensing scientists, to ensure the methodological tools I create are genuinely useful and impactful both in and beyond the statistical domain. 

My methodology research interests include spatial, Bayesian and adaptive sampling design methods, with a focus on making these scalable and meaningful for big data and remote sensing analysis. I also work on enhancing machine learning methods by combining these with traditional statistical approaches to account for spatial information and quantify uncertainty of predictions. This is important in many applications, and particularly environmental monitoring where systems can fluctuate and data are often missing, disparate or limited.

My environmental interests include, but are not limited to, forest cover monitoring, modelling biodiversity, weather data and environmental change of our terrestrial and marine environments over time.