Exploring the 2022-2023 eruption sequence of the Northern group of volcanoes in Kamchatka with the seismic background level (SBL) technique and satellite images

Episodes of volcanic unrest and eruptions are accompanied by various seismic signals, so seismic observations have become one of the most effective methods for monitoring volcanoes. Among them, to examine the potential of the seismic background level (SBL) technique for monitoring exceptionally active volcanoes with a limited number of stations, we applied it to seismic data recorded at the Northern group of volcanoes (NGV) in Kamchatka. NGV is an area with highly active and diverse volcanism formed by a dense cluster of active volcanoes, the Klyuchevskoy volcano group (KVG), and Shiveluch, the northernmost active volcano of Kamchatka.
For this work, we chose four stations located within a few tens of kilometers of the active volcanoes and calculated the SBL during the 2022--2023 eruptive sequence. We combined the results with more conventional Real-time Seismic Energy Measurement (RSEM) and the thermal anomalies detected by the Himawari-8/9 satellite. Because the Bezymianny and Klyuchevskoy volcanoes, separated only by 10 km, were erupting in the same period, it was impossible to distinguish between the two volcanoes using only a single SBL time series. By comparing the SBL amplitudes at the three stations at the KVG in various frequency bands, we were able to separate the unrest of Klyuchevskoy and Bezymianny. The results suggest continuous low-frequency tremor from a deep source beneath the KVG, which rose to the shallow depths beneath Klyuchevskoy before its eruptions. Also, we identified high-frequency continuous tremor at shallow depths beneath Bezymianny, indicating sustained unrest during the five major eruptions over 18 months. 
On the other hand, the SBL variations at Shiveluch did not reflect the surface eruptive activity but a potential inflation event of this volcano after its catastrophic eruption. Based on these observations, it appears that SBL can detect both eruptive and non-eruptive processes in magmatic systems.
We also demonstrated that growing features of eruption precursors, consisting of volcanic earthquake events or discrete tremors, can be captured by the RSEM but not by the SBL.  We emphasize that combining SBL and traditional approaches will better capture the precursors and significance of volcanic unrest.