The geochemistry and evolution of the Lizard Complex, Cornwall
The Lizard Complex consists of a poorly exposed assemblage of serpentinites, gabbros, amphibolites, basic dykes, metasediments and gneisses. As an alternative approach to conventional mapping of the inland area, the geochemistry of the residual soils of the area are used to identify the underlying lithologies. Using multi-element geochemical data obtained by X-ray fluorescence spectrometry, an identification scheme is established which allows the recognition of over a dozen mappable units with a high rate of success. This scheme evolved from a training set of samples collected during an orientation survey from areas of undisputed geology immediately adjacent to coastal exposure. At approximately 250 sites it has been possible to confirm the geological predictions directly against material obtained from the base of power auger holes. A non-hierarchical K-means clustering technique is developed and used to recluster data from over 800 samples into gabbroic and ultrabasic groups. The results of this classification produced more sensitive discriminatory parameters which in turn are used on 1500+ samples collected during the project. The soil identification procedure combines the use of a variety of numerical techniques and the algorithm designed to use them operates in a sequential manner. The most distinct units are identified at an early stage whilst those samples which are less distinct chemically take longer to isolate and as a final stage of treatment are separated using a modified K-means technique. This approach to mapping has allowed the recognition of units not found outcropping on coastal sections and for an improved geological map of the Lizard Complex to be drawn. In particular the Crousa gabbro is shown to be a smaller body than previously considered and to comprise several different bodies. As a result of this work the evolutionary history of the Complex is revised.