Volcanic eruptions pose significant risk to local populations, as well as having the potential to cause global-scale atmospheric impacts. A range of tools are used to research and monitor volcanoes, to better understand the phenomena and, ultimately, to attempt to forecast volcanic eruptions. Remote sensing techniques are critically important to these efforts, providing invaluable data on a range of volcanic manifestations whilst mitigating the risk to volcanologists gathering the data.
This thesis presents on the development and testing of three novel low-cost instruments for ultraviolet (UV) and near-infrared (NIR) remote sensing applications in volcanology. The instruments were all based on the Raspberry Pi Camera Module (v1.3), which houses a sensor primarily designed for the smartphone market. Firstly, this work details the modification of this sensor, by chemical removal of its Bayer filter, to attain an improved UV/NIR sensitivity. From this modified sensor, three instruments were designed and built: the PiCam, a UV imaging camera for SO2 camera retrievals; the PiSpec, a UV spectrometer for differential optical absorption spectroscopy retrievals of SO2; a NIR PiCam, for thermal imaging of high-temperature sources. Each instrument was built from the modified sensor, 3D-printed optical mounts, and off-the-shelf optical components. All instruments were rigorously tested, through laboratory and field tests, to evaluate their performances. This included data acquired from deployments on a number of active volcanoes: Mount Etna (Italy), Masaya volcano (Nicaragua), Kilauea (Hawai’i, US). In general, the instruments were found to perform at least adequately for
their desired applications.
These low-cost instruments could promote widespread dissemination of important scientific equipment, facilitating the expansion of data acquisition for volcano monitoring and research. This could be of particular importance in developing countries, where a large proportion of the world’s active volcanoes are located, but where volcanology funding is often quite limited.
