Remote sensing of water temperature using Raman spectroscopy

The measurement of water temperature, both at and below the surface in bodies of water, is of great importance for research fields including climate change, marine biology, hydrologic modelling and marine engineering. While sea surface temperature can be obtained with passive satellite sensing, sub–surface measurements are limited to suspended arrays of detectors, expendable probes and autonomous vehicles. None of these methods are compatible with rapid systematic mapping of sub–surface water temperature. The aim of the research presented in this thesis is to develop methods for rapid, depth–resolved remote sensing of water temperature using Raman spectroscopy, and to explore the potential to incorporate measurement of salinity. An optical, depth–resolved temperature sensing solution would provide a capability for rapid, three–dimensional sub–surface measurement (tens of metres) within bodies of water from airborne, surface or sub–surface platforms. Raman scattered light from vibrational transitions of water have temperature dependent properties which may be exploited for temperature determination. In order to achieve these goals, a detailed laboratory study of the polarisedand unpolarised Raman spectra of water was conducted. This identified spectral features and measurement parameters which gave the greatest temperature determination accuracy. Multiple linear regression was employed to combine these parameters in models for determination of water temperature. An assessment was made of the temperature accuracy which might be expected for practical sensing using these parameters. Raman spectra of samples from natural water sources were analysed to assess the effect of water constitution on temperature determination accuracy and the application of baseline correction to improve this. The effect of changing both temperature and salinity on Raman spectra was examined using multivariate analysis, and the potential measurement accuracy of both quantities was determined. Configurations for simple multi–channel sensing instruments were designed based on the Raman temperature and salinity determination methods investigated, and these will be evaluated in future field studies.