Tilt change recorded by broadband seismometer prior to small phreatic explosion of Meakan-dake volcano, Hokkaido, Japan

We discovered an anomalous very long period (VLP) signal (duration: about 60 s) in the broadband seismic record of the precursory swarm of a small phreatic explosion at the Meakan‐dake volcano, Hokkaido, Japan, in 2006; this signal was masked by a short‐period tremor in the original velocity waveform, but a considerable DC step appeared in the displacement trace. Since such a step results from a tilt change, we interpreted the observed trace as being a combination of rotational (tilt) and translational (displacement) motions. The apparent displacement step in the VLP signal indicates subsidence of the summit crater. Owing to limited data, we attempted deflating the Mogi source and simulating the seismic record including both motions. When only rotational motion is considered, the simulated waveform explains well the general features of the observed trace. However, the combined trace has incompatible phases, suggesting that a more complicated source process should be considered.