A Raspberry Pi-Powered Ultraviolet Camera System Seeks to Monitor Volcanoes to Forecast Eruptions

A team of researchers from the Universities of Sheffield and Cambridge, the Universidad Católica del Norte, the Millennium Institute on Volcanic Risk Research, and the US Geological Survey (USGS) has developed a new, lower-cost camera that they say could help forecast volcano eruptions with enough warning to get affected populations to safety.

"Our instrument uses a sensor not dissimilar to smartphone camera sensors," explains lead author Thomas Wilkes, PhD, of the camera the team have developed. "It is modified to make it sensitive to ultraviolet light, therefore enabling SO₂ [sulfur dioxide] detection.

"Before now, only three volcanoes have had permanent SO2 cameras installed on them. Discrete field campaigns have been carried out, and whilst they can be invaluable for a range of research questions, it is important to be able to measure volcanic activity continuously, since it can vary substantially from minutes to decades to centuries and beyond."

The main problem with deploying sulfur dioxide-monitoring cameras widely is their cost, with each unit costing upwards of $20,000. The team's version, by contrast, costs just $5,000 — meaning it's possible to deploy four cameras for every one of the previous models. As well as costing less, the new design also draws less power — averaging at 3.75 watts, roughly half that of competing designs.

The secret sauce at the heart of the camera: modified Raspberry Pi camera modules, tweaked to be sensitive to ultraviolet light through the removal of their Bayer filters. Following the filter removal, an optical system which transmits UV light is fitted followed by UV bandpass filters. Elsewhere in the triple-camera unit is a family of Raspberry Pi single-board computers, connected to a network switch and with power scheduling via WittyPi Hardware Attached on Top (HAT) accessories, a GPS module, a spectrometer for camera calibration, and a 1TB SSD for data storage, all housed within a Peli case.

"Wherever possible we 3D print parts too, to keep costs as low as we can," Wilkes adds. "We also introduce a user-friendly, freely available software for controlling the instrument and processing the acquired data in a robust manner. The SO₂ camera can provide higher time- and spatial-resolution data [than other sensor types,] which could facilitate new volcanological research when installed permanently."

While the team's design has been field tested, gathering data from Chile's Lascar stratovolcano and Hawaii's Kilauea shield volcano, there are limitations to its use — common to all SO₂ camera sensors. "They are dependent on meteorological conditions," Wilkes admits, "and work best under clear blue skies when the volcanic gas plume moves in a 90-degree angle to the viewing direction of the camera." They also can't operate at night, due to a lack of UV radiation.

The team's work has been published in the journal Frontiers in Earth Science under open-access terms.