- Géosciences Rennes
Bât.15 campus Beaulieu
Université de Rennes 1
Adresse courriel : kerry [dot] gallagher [at] univ-rennes1 [dot] fr
Téléphone : +33 223236081
Numéro de bureau : Bât.15, P119
BAT 15, Pièce 119, Campus Beaulieu
Université de Rennes 1
35042 RENNES cedex
Thèmes de recherche
I have worked with geologists, geophysicists, geochemists and geomorphologists in a diverse range of projects. Currently, I am in collaboration with researchers in France, Scotland, England, Denmark, Brazil, and Australia, and I am co-investigator for projects funded in Greenland, South Africa and Brazil. My main research topics consist of the quantification of geological phenomena, for example, in understanding the long-term landscape evolution and chemical fractionation during the formation of magmas. My scientific approach is based on a quantitative interpretation of data, rather than a purely theoretical or empirical/qualitative approach. Consequently, much of my research has been oriented to the development and applications of advanced statistical methods in the field of inference and uncertainty from modeling (focussing on Bayesian transdimensional inverse methods now being more widely commonly used in the Earth Sciences).
In recent years, I have put some effort into developing software based on some of these modelling approaches for use by the research community. The motivation was to produce user-friendy software, that would run on desktop machines (both Windows and Macintosh). To this end, I wrapped the inverse modelling algorithms with a GUI written in Qt. In particular I have produced the following freely available software :
For the inference of discrete mixtures, as often required in detrital geochronology for example. The approach is based on that presented in Jasra, A., Stephens, D.A. Gallagher, K. and Holmes, C.C., (2006) Analysis of geochronological data with measurement error using Bayesian mixtures, Mathematical Geology, 38(3), 269-300.)
For changepoint modelling, that is the inference of discontinuities in time- or depth- data series, allowing for unknown data noise, and combinations of different data sets with a common underlying changepoint structure. This has been used to identify and correlate climate change related variations in chemistry of peat bogs (natural archives of atmospheric composition). The approach is based on that presented in Gallagher, K., Bodin, T. Sambridge, M, Weiss, D, Kylander, M, and Large, D. (2011) Inference of abrupt changes in noisy geochemical records using Bayesian transdimensional changepoint models, Earth Planet. Sci. Letts., 311, 182-194.
For modelling thermal histories from thermochronological data (apatite and zircon (U-Th)/He, apatite and zircon fission track, 40Ar/39Ar (multidomain model) and 4He/3He spectra, U-Pb in apatite, K-Ar/Ar-Ar in mica/hornblende/etc, vitrinite reflectance. The software allows multiple samples to be modelled jointly (e.g. a vertical profile or borehole) with up to 50 samples in a profile. It allows for parent concentration zoning in single crystals, and can deal with age profiles across single crystals. The approach is based on that presented in Gallagher, K., (2012) Transdimensional inverse thermal history modelling for quantitative thermochronology, J. Geophys Res. 117, B02408, doi:10.1029/2011JB00882.
1988-1990 National Research Fellow co-located with Fission Track Research Group, La Trobe University, and Broken Hill Proprietary Research Ltd, Melbourne, Australia.
1990-1992 Research Fellow, The Open University, England (and visiting Research Fellow, University College London).
1992-1993 ELF funded Research Fellow, Fission Track Research Group, University College London, England.
1993-1994 Research Lecturer, Earth Sciences, Kingston University, England
1994-1999 Governor’s Lecturer, Imperial College London, England
1999-2006 Reader in Geophysics, Imperial College London, England
CURRENT POSITION : Professor des Universités, 1e classe University of Rennes 1, France.