| Wind Energy | A recent project centred around the validation of computational fluid dynamical (CFD) models for wind turbine wakes using SCADA observational data from a wind farm. This project entailed the writing of a novel geometric algorithm to determine, for a given wind direction, which turbines in the wind farm have wakes that impinge on a given machine of interest. The data at that machine is filtered to be at times when the identified group of machines are working correctly. This facilitates comparison with the CFD model which assumes the machines to be operating correctly. Previous work in the wind energy field involved the development of advanced statistical (Vector Auto-Regressive) models of the wind in order to estimate capacity factors for wind farms of interest, in wind integration studies. They can also be used for wind resource assessment. This project also involved the processing and handling of much data from the pan-European COSMO meteorological model, the British Atmospheric Data Centre's MIDAS meteorological data and the use of the MERRA reanalysis set. Selected Publications |
| Clinical Trials and Epidemiology | Analysis of maternity care survey data employed logistic regression techniques in mixed effect models and used weights to address non-response issues. An overview of urinary incontinence involved meta-analyses making forest plots and network maps from published trials. A separate project investigated nursing interventions on the wards, using linked data from the Safe Haven at Dundee, Scotland. Previous work included an epidemiological project analysing health risks from naturally occurring radioactive (residential) radon gas in a major pan-European project, pooling case control studies.Selected Publications |
| Ocean Sciences and Environmental Science | A methodology was created, incorporating an optimisation technique in an inverse method, for estimating sediment parameters and provided a way of decoupling the advection of a background concentration gradient from the resuspension component. This was in addition to estimating bed parameters, erosion rates and settling velocities, or eddy diffusivities. Calibration software for the backscatter signal from an Acoustic Doppler Current Profiler to measure suspended sediment concentrations was also written and applied. Oceanic work included tidal modelling creating Digital Elevation Maps (DEMs) of the inter-tidal zone by utilising shorelines taken from Synthetic Aperture Radar (SAR) data at different times of the tidal cycle in conjunction with modelled sea elevations to create the DEMs. Experience also includes detection and attribution of climate change studies, applying statistical techniques to ocean-atmosphere models. This involved a technique called optimal fingerprinting (a regression method with an expansion of the control in terms of EOFs) to distinguish anthropogenic signals from natural ones in vertical atmospheric temperature patterns. Selected Publications |