Absolute gravity change associated with the March 1997 earthquake swarm in the Izu Peninsula, Japan

Abstract

We carried out both absolute and relative gravity measurements in the Izu Peninsula just before and after the March 1997 earthquake swarm occurred. The measurements revealed significant absolute gravity changes, which we find to be made of three spatial components. The first one is located near Cape Kawana, and would be associated with the volcanic activity that caused the earthquake swarm. The second one would be associated with shallow and localized magma intrusion just beneath Ito. The third one may be due to a change in the deep region beneath the Kita-Izu fault system, which is considered to be a major tectonic line of this region. The gravity changes can be used to detect underground mass movement. For this purpose, we first use crustal movement observations to construct an elastic dislocation model with two tensile faults and a left lateral fault. Then we use the gravity changes to constrain the density of the material which filled the tensile faults. We find that the density is likely to be small, and that the gravity changes of the first component are reproduced well by the fault model. The smallness of the density implies that highly vesiculated magma or water would have injected into the faults.

First detection of absolute gravity change caused by earthquake

“…Yoshida et al [1999] employed a more sophisticated approach combining both absolute and relative gravimetry to detect gravity changes due to an earthquake swarm. However, this gravity change was mostly of volcanic origin because tensile faulting was dominant [Yoshida et al, 1999]. In short, a relation between earthquakes and gravity changes had not been clearly established yet prior to the result presented here, since none Table 1).…”

Groundwater-level Changes Due to Pressure Gradient Induced by Nearby Earthquakes off Izu Peninsula, 1997

“…The fault parameters during the swarm activity are estimated from gravity change (YOSHIDA et al, 1999) and geodetic data (CERVELLI et al, 2001), respectively. They analyzed the accumulated change through the period of swarm activity, but did not give the parameters for E1 or E2.…”