A Multiband Uncooled Infrared Camera for Measuring Volcanic SO2 Gas Concentration and Temperature Distributions

Measuring the surface phenomena of a volcano is useful for predicting its short-term activity. To this end, methods for remotely sensing SO
2
gas emitted from volcanoes have been developed. One such method is an imaging sensor for observing infrared radiation in the longwave infrared (LWIR) region. This method allows for day and night observation of infrared radiation. In this study, we develop a multiband LWIR camera named SPIC-UC/4VGA prototype for the quantitative imaging of volcanic surface phenomena, such as surface temperature and gas distribution. The SPIC-UC/4VGA prototype consists of four uncooled infrared cameras equipped with an internal optical filter (Cameras 1, 2, 3, and 4). All four cameras employ focal plane array (FPA)-type amorphous silicon microbolometer elements. Cameras 1–3 measure the 7500–14000, 9000–14000, and 7950–9300 nm regions, which match the SO
2
absorption bands, and Camera 4 measures the 11785–12785 nm region. These four cameras can acquire frame-synchronized data at 30 fps. The SPIC-UC/4VGA prototype is realized as an environment-resistant type in consideration of outdoor use. Using a shutterless operation, the performance evaluations indicate that all cameras achieve an absolute temperature accuracy of ± 2 K. For SO
2
detection, Camera 3 can achieve an NETD (noise equivalent temperature difference) of approximately 0.39 K. The evaluation of SO
2
measurement accuracy by simulation using the same S02 gas concentration distribution conditions as those under which our original airborne hyperspectral sensor (ARTS) actual observations were made on April 8, 2008 reveals that the developed Camera 3 can detect SO
2
gas concentration distributions under background conditions at a temperature of 50°C with errors of ± 1 ppmv. These results indicate that the developed SPIC-UC/4VGA prototype can be used for volcanic SO
2
gas quantitative detection.