Enhanced Accuracy of Airborne Volcanic Ash Detection Using the GEOKOMPSAT-2A Satellite

In this study, a technique facilitating the enhanced detection of airborne volcanic ash (VA) has been developed, which is based on the use of visible (VIS), near-infrared (NIR), and infrared (IR) bands by meteorological satellite systems. Channels with NIR and IR bands centered at ~3.8, 7.3, 8.7, 10.5, and 12.3 μm are utilized, which enhances the accuracy of VA detection. The technique is based on two-band brightness temperature differences (BTDs), two-band brightness temperature ratios (BTRs), and background image BTDs. The physical effects of the observed BTDs and BTRs, which can be used to distinguish VA from meteorological clouds based on absorption differences, depend on the channel and time of day. The Advanced Meteorological Imager onboard the GEOKOMPSAT-2A (GK-2A) satellite has several advantages, including the day- and nighttime detection of land and ocean. Based on the GK-2A data on several volcanic eruptions, multispectral data are more sensitive to volcanic clouds than ice and water clouds, ensuring the detection of VA. They can also be used as an input to provide detailed information about volcanoes, such as the height of the VA layer and VA mass. The GK-2A was optimized, and an improved ash algorithm was established by focusing on the volcanic eruptions that occurred in 2020. In particular, the 3.8 μm band was utilized, the threshold was changed by division between day and night, and efforts were made to reduce the effects of clouds and the discontinuity between land and ocean. The GK-2A imagery was used to study volcanic clouds related to the eruptions of Taal, Philippines, on 12 January and Nishinoshima, Japan, from 30 July–2 August to demonstrate the applicability of this product during volcanic events. The improved VA product of GK-2A provides vital information, helping forecasters to locate VA as well as guidance for the aviation industry in preventing dangerous and expensive interactions between aircrafts and VA.