地殻変動データに含まれる雪と降雨の影響の除去手法の開発

抄録

Crustal deformation data recorded by strainmeters and tiltmeters installed in snowy areas are often affected by snow （snowmelt and snow loads） in addition to rainfall. Specific characteristics of this effect have not been sufficiently investigated. Here, we analyzed the effect of snow on tilt data by comparing tilt changes with meteorological records at nearby stations. Our results showed that the amount of tilt changes during the snowmelt season is related to the annual amount of snowmelt. Furthermore, the observed tilt changes are synchronized with the period of meltwater infiltration into the ground. Based on these results, we improved a tank model that includes a ʻsnow tankʼ to remove the effects of rainfall and snow from the tilt data and applied this model to tilt data from two sites in snowy areas. Our model can calculate the expected crustal deformation response to snow loads, snowmelt and rainfall based on meteorological data （precipitation, air temperature, and hours of sunshine）. We determined optimal parameters for this improved model on a site-bysite basis from long-term crustal deformation data spanning approximately 5 years. As a proof of concept, we successfully retrieved long-term tilt changes by removing the effects of snow and rainfall of a few micro radians. By applying our model to other crustal deformation data, we can more reliably determine the amount of crustal deformation due to, for example, magmatic intrusions, seismic uplift and subsidence, etc.