- BSc (Hons) Marine Biology & Zoology, UWA.
- Applied Statistics
- Variable Selection
- Shrinkage Methods (e.g. LASSO, Elastic Net, Bayesian LASSO)
CRCSI Biomass Business (4.12): PhD Project 3.1 – Subset selection and spatial inference for soil carbon.
The Earth’s soils contain more carbon than all land plants and the atmosphere combined. With a third of the dry land devoted to agriculture, seeking to understand the distribution and evolution of soil carbon stocks in agricultural soils is an important component of seeking to understand global carbon dynamics. Soil carbon has high spatial heterogeneity even at relatively local scales understanding which will aid the formation of a more holistic understanding. My research is focused on methods for quantifying soil carbon stocks at an individual paddock scale. Soil carbon is typically measured by laboratory analysis of soil core samples which is both time and resource intensive. My research focuses on methods to extract maximum value from limited soil core survey data. To this end I am exploring methods to model soil carbon with a broad ensemble of environmental variables all available as either high resolution geostatistical surveys or high resolution full cover rasters. The aim of this modelling being to use these high resolution environmental variables to assist in interpolating soil carbon levels between the locations of the individual soil cores and provide full cover inference for the soil organic carbon distribution.
Fitzpatrick, B. R., Lamb, D. W., & Mengersen, K. (2016). Ultrahigh Dimensional Variable Selection for Interpolation of Point Referenced Spatial Data: A Digital Soil Mapping Case Study . PLoS ONE 11(9): e0162489.
Langlois, T. J., Fitzpatrick, B. R., Fairclough, D. V., Wakefield, C. B., Hesp, S. A., McLean, D. L., Harvey, E. S., Meeuwig, Jessica J. (2012). Similarities between Line Fishing and Baited Stereo-Video Estimations of Length-Frequency: Novel Application of Kernel Density Estimates. (D. M. Bailey, Ed.) PLoS ONE, 7(11).
Group R Tips Blog: One weiRd tip
Thesis in 3 talk at the CRCSI 2014 Conference.