Airborne thermography and ground geophysical investigation for detecting shallow ground disturbance under vegetation
This thesis discusses the potential of airborne thermal prospection for detecting shallow ground disturbance beneath vegetation based on images acquired by the NERC Airborne Thematic Mapper (ATM) at thermal infrared wavelengths. Shallow ground disturbance creates a differential heat flux due to a variation in the thermal properties between disturbed and undisturbed soils. When observed above a canopy, the effect of vegetation growth on the thermal regime of the underlying soils is poorly understood. The research extends current understanding by examining areas where ground disturbance is known to exist under variable vegetation cover at an archaeological site at Bosworth, Leicestershire and areas of abandoned mine activity on Baildon Moor, W. Yorkshire and in the N. Pennine Orefield, Weardale. The investigation focuses on qualitative image interpretation techniques, where anomalies on day and night thermal images are compared with those manifest on the multispectral images, and a more quantitative approach of Apparent Thermal Inertia (ATI) modelling. Physical thermal inertia is a parameter that is sensitive to volumetric variations in the soil, but cannot be measured directly using remote sensing techniques. However, an apparent thermal inertia is determined by examining the day and night temperature contrast of the surface, where spatial variations can signify potential features buried in the near-surface environment. Ground temperature profiling at the Bosworth site indicates that diurnal heat dissipates between 0.20-0.50m at an early stage in vegetation development with progressively lower diurnal amplitudes observed at 0.20m as the vegetation develops. Results also show that the time of diurnal maximum temperature occurs progressively later as vegetation develops, implying an importance for thermal image acquisition. The quantitative investigation concentrates on the Bosworth site where extensive ground geophysical prospection was performed and vertical soil samples extracted across features of variable multispectral, thermal and ATI response to enable comparison of the observed airborne thermal response with physical soil properties. Results suggest that there is a high correlation between ATI and soil moisture properties at 0.15-0.25m depth (R(^2)=0.99) at an early stage in cereal crop development but has a high correlation at a wider depth range (0.10-0.30m) at a later stage in development (R(^2)=0.98). The high correlation between physical ground disturbance and the thermal response is also corroborated qualitatively with the results of the resistivity surveys. The ATI modelling reveals similar features to those evident on day or night thermal images at an early stage in vegetation growth, suggesting that thermal imaging during the day at an early stage in vegetation growth may supply sufficient information on features buried in the near-surface environment. Airborne thermal imaging therefore provides a useful complementary prospection tool for archaeological and geological applications for surfaces covered by vegetation.