Statistical analysis of dispersal and deposition patterns of volcanic emissions from Mt. Sakurajima, Japan

Abstract

With the eruption of Eyjafjallajökull (Iceland) in 2010, interest in the transport of volcanic ash after moderate to major eruptions has increased with regards to both the physical and the emergency hazard management aspects. However, there remain significant gaps in the understanding of the long-term behaviour of emissions from volcanoes with long periods of activity. Mt. Sakurajima (Japan) provides us with a rare opportunity to study such activity, due to its eruptive behaviour and dense observation network. In the 6-year period from 2009 to 2015, the volcano was erupting at an almost constant rate introducing approximately 500 kt of ash per month to the atmosphere. The long-term characteristics of the transport and deposition of ash and SO2 in the area surrounding the volcano are studied here using daily surface observations of suspended particulate matter (SPM) and SO2 and monthly ashfall values.Results reveal different dispersal patterns for SO2 and volcanic ash, suggesting volcanic emissions’ separation in the long-term. Peak SO2 concentrations at different locations on the volcano vary up to 2 orders of magnitude and decrease steeply with distance. Airborne volcanic ash increases SPM concentrations uniformly across the area surrounding the volcano, with distance from the vent having a secondary effect. During the period studied here, the influence of volcanic emissions was identifiable both in SO2 and SPM concentrations which were, at times, over the recommended exposure limits defined by the Japanese government, European Union and the World Health Organisation.Depositional patterns of volcanic ash exhibit elements of seasonality, consistent with previous studies. Climatological and topographic effects are suspected to impact the deposition of volcanic ash away from the vent: for sampling stations located close to complex topographical elements, sharp changes in the deposition patterns were observed, with ash deposits for neighbouring stations as close as 5 km differing as much as an order of magnitude. Despite these effects, deposition was sufficiently approximated by an inverse power law relationship, the fidelity of which depended on the distance from the vent: for proximal to intermediate areas (<20 km), errors decrease with longer accumulation periods (tested here for 1–72 months), while the opposite was seen for deposition in distal areas (>20 km).