Supervisory team: Tom Jordan (BAS), Chris Green (Leeds), Teal Riley (BAS), Peter Fretwell (BAS), Andrew Fleming (BAS), Alex Burton-Johnson (BAS), Dr Sabine Chabrillat (gfz-potsdam).
Antarctica’s extreme environment and vast scale are a challenge for traditional geological mapping. The new hyperspectral EnMAP satellite provides a unique opportunity to build on decades of fieldwork and airborne geophysical observations to provide the first large-scale remotely sensed map of the continent’s geology. Previous BAS research using the ASTER multi-spectral satellite data  demonstrated that such techniques work for geological mapping, but were limited to mapping alteration minerals. Subsequent BAS research using airborne hyperspectral thermal infrared data  was able to map bedrock but was limited in coverage. With its full spectrum hyperspectral data (420 nm to 2450 nm), the new Environmental Mapping and Analysis Program (EnMAP) satellite offers both coverage and accuracy, enabling this PhD project to carry out the first automated large-scale geological mapping of Antarctica.
The aim of this project is to enhance both the coverage and the confidence in Antarctic Peninsula geological mapping based on EnMAP satellite data, combined with developing a spectral library from Antarctic samples and ground truthing from existing geological field observations. This will be integrated with airborne magnetic and gravity data and laser/radar derived digital elevation models to map across areas with no outcrop.
The first objective will be to derive appropriate calibration for the EnMAP satellite data set in Antarctica. This will focus on regions with well-known geology and build on previous atmospheric calibration tools developed by BAS for high resolution airborne hyperspectral datasets . The second objective will be to derive a robust physical model for the observed spectra, utilising the rock samples held in the extensive BAS rock collection. The final objective will be to integrate hyperspectral, geophysical, and elevation datasets to map the geology of less well-studied regions.
As well as EnMAP data, airborne hyperspectral data and airborne geophysics, other data sets such as multi-spectral satellite data (e.g. ASTER) and surface roughness results can add useful information. Identification of geological provinces in the study area can potentially be enhanced by a combination of geological modelling, statistical analysis, and classification analysis using machine-learning techniques.
 Haselwimmer, C.E., T.R. Riley, and J.G. Liu, Assessing the potential of multispectral remote sensing for lithological mapping on the Antarctic Peninsula: case study from eastern Adelaide Island, Graham Land. Antarctic Science, 2010. 22(03): p. 299-318. https://doi.org/10.1017/S0954102010000015 (https://tinyurl.com/txpkjv22)
 M. Black, T.R. Riley, G. Ferrier, A.H. Fleming, P.T. Fretwell, Automated lithological mapping using airborne hyperspectral thermal infrared data: A case study from Anchorage Island, Antarctica. Remote Sensing of Environment, 176, p225–241 https://doi.org/10.1016/j.rse.2016.01.022 (https://tinyurl.com/6b75nsch)