My Research

Please get in touch for more information about some of the exciting things I have been working on or for any full texts

Research Themes

My research is focussed on applied Data Science / AI, focussing on environmental and health impacts of air pollution and carbon neutral energy production.

Machine Learning in Geophysics
Computational Geoscience
Advancing Sustainable Infrastructure Through Data-Driven Seismicity Research
Machine Learning and Data Science in Health

5-Year Research Outlook

Ongoing or Proposed Projects

Although my previous work involved computational geology and numerical modelling, my future research plans involve a more data-centric approach, utilising skills in data science.

Theme 1

using a RaspberryShake to quantify traffic volumes from anthropgenic noise – IoT application that will allow a dense network of traffic measures to be deployed cheaply and that can be implemented into smart road networks, such as traffic lights and speed signs. Taylor, R,L., Gerber, L. and Healy, D. Predicting urban traffic patterns using AI and seismic signature: Insights for sustainable urban planning.
using a RaspberryShake and air quality sensors to predict local greenhouse gas emissions from anthropgenic noise – IoT application that can inform policy and decision-making in Smart Cities. Taylor, R,L., Gerber, L. and Healy, D. Using AI to predict local air pollution and its effect on human health.

Theme 2

petrophysical and mechanical modelling of fault rocks, comparing 2D versus 3D approaches to fracture characterisation – that will compare methods that allow the fracture state from observations at the surface to be modelled using expensive and labour intensive techniques to relatively cheap practices that require little expert knowledge and training. Taylor, R.L., Seers, T., Hodgetts, D., Rutter, E.H. Petrophysical and mechanical modelling of fault rocks: A comparison of 2D versus 3D approaches to fracture characterisation.
using XRCT tomography images to detect hydraulic-induced fractures – that will have implications for improved fracture detection in rocks of transitional green energies. Taylor, R,L. and Chandler, M. Rock fracture detection using deep learning from micro-CT tomography images of shale.
do preprints hinder or facilitate the number of citations a paper receives once published? - analysing metadata from the EarthArxiv preprint server to establish patterns in the number of citations. Narock, T., Taylor, R.L., Goldstein, E., Boston, A. and Irawan, D.E. EarthArxiv: Today and Tomorrow.

Theme 3

predicting earthquakes from solid earth tides - that can help us mitigate risks in geoenergy applications and improve sustainability of our energy infrastructure. Taylor, R.L. Unlocking earthquake predictions: Investigating the role of solid earth tides through statistical analysis.
high-level nuclear waste repository and carbon-capture and storage solutions - understanding the subsurface fluid flow pathways and probable scenarios for potential fluid leakage and contamination. Taylor, R.L. et al. Quantitative analysis of the influence of fractures on fluid flow in rocks, based on the use of digital outcrop analogues.
high-level nuclear waste repository and carbon-capture and storage solutions - leveraging machine learning and deterministic forward modelling to predict the effect of climate controls on glacial seismic and seismic risk to protect the sustainability of our energy infrastructure. Taylor, R.L. et al. A data-driven approach to climate controls on neotectonics.

Theme 4

using machine learning to predict patient mortality from cardiopulmonary exercise test measurements – that will have implications for the metrics that are used to assess patient eligibility for major abdominal surgeries. Bagley, L., Taylor, R.L., Maudsley-Barton, S., Bryan, A. and Han, L. A machine learning approach to predicting patient mortality from CPET to assess patient eligibility for major abdominal surgeries.
biophysical modelling of skeletal muscle – a review of recent modelling approaches to the nine main muscle subsystems. Hodson-Tole, E., Taylor, R.L., Adeniran, I. and Degens, H. Reviewing recent advances in biophysical modelling of skeletal muscle.

Data Science

The application of data science techniques into data analysis of science research is ever increasing. Using these powerful tools we can, as scientists, use vast quantities of past data to predict future events.

My particular interest in Data Science lies within the fields of Health/Sports Science and Environmental Science.

I primarily use tabular and time series data to solve environmental problems.

Virtual Outcrop Modelling

Applications of virtual outcrop modelling are far-reaching. Not least in terms of accessibility and inclusion. Viewing rocks in the virtual space can be used as an outreach tool, open access by reaching people on a global scale, and more inclusive. It can aid learning in ways we can only imagine.

I have experience in photogrammetric modelling of outcrops and I am further exploring this research avenue to incorporate a more integrated and immersive experience. I would welcome academic collaboration.

Fracture Modelling

High resolution data from core samples cannot provide a representative elementary volume (REV) of a reservoir as the volume necessary to define a REV within a fracture system is large. Field seismic data therefore cannot be forward modelled or validated from core data alone. The effects of fractures on P-wave velocity in fractured basement reservoirs must be investigated by integrating fracture outcrop studies and laboratory testing of core samples to produce deterministic forward models.

Carbonate Petrology

and Petrophysics

In an effort to gain a better understanding of petrographic and petrophysical data from carbonate lithofacies, fracture distributions have been linked with mechanical and textural properties, and petrophysical data linked to textural properties. Fractures have a significant contribution to fluid flow within rocks, particularly where matrix porosity is low, such that this type of research is applicable to hydrocarbon exploration and production from carbonate rocks.

Fluid Flow Properties

The commercial exploitation of shale-gas using hydraulic fractures depends on the rate at which gas can flow through the pores of the rock matrix. The matrix permeability of gas shales is of particular importance because production requires gas to be released from pore spaces that are present on a size-scale commensurate with the grain size of the rock, and to flow towards a fracture network. It is therefore essential to understand the physical and chemical processes that influence such permeability, so that meaningful reservoir models can be made, leading to estimates of gas-in-place and rate of production.

To date, my research has been focused on 4 main areas (in geology):

detailed field studies on the structure and geophysical properties of strike-slip fault zones to understand how seismological properties are influenced by fault zones, to enable better interpretation at depth,
numerical modelling of the effect of fractures on seismic rock properties,
computational modelling and rock typing of sedimentary rocks based on the measurement of petrophysical properties, and
measurement of the fluid flow properties of tight rocks to understand better gas draw-down effects during well production and likely yield of a shale gas reservoir.

Page updated

Google Sites

Report abuse