I am interested in statistical and numerical modelling of geologic processes, their interactions, and spatial/temporal variability. I have also worked extensively on probabilistic hazard and risk assessment, expert judgement elicitation and decision making under uncertainty.
A large part of my work involves the use of Bayesian Networks to model and understand conditional dependencies, primarily using UNINET - software developed by TuDelft/LightTwist for performing uncertainty analysis and modelling high dimensional distributions. Core skills include programming and data analysis (mainly R and C++) such as: data mining, Monte Carlo simulation, time series and geospatial analysis (QGIS/GRASS/SAGA GIS). Previously, I have worked on modelling and forecasting tools for volcanic hazard assessment and carried out expert elicitations and statistical analyses for projects funded by the UK NDA, World Bank and DFID.
I am currently working on the FELS Knowledge Exchange and Enterprise funded project "Machine learning to unpick evolving sensitivity to induced seismicity". This extends the work of Hincks et al. (2018, "Oklahoma’s induced seismicity strongly linked to wastewater injection depth") to characterise time lags and temporal variation in susceptibility to induced earthquakes.
From mid-2022, I will commence work on the NERC funded project "GLObal Suspended Sediment (GLOSS): Drivers, trends and future trajectories" (NE/W001233/1) with the School of Geography & Environmental Science at Southampton. Here, I will perform probabilistic modelling and analysis of global remote sensing data, to quantify the strength of influence of multiple environmental and anthropogenic drivers of global sediment flux. The aim is to better understand and predict sediment flux to the ocean (key drivers, interactions and uncertainties), and develop scenarios to quantitatively characterise the impact of future changes in climate and land use.