10:45 AM - 12:15 PM
[SGD02-P15] Attempt to detect anomalies in crustal movement across Japan
Keywords:GNSS
We aim to index the state of crustal activity based on the analysis of earthquake and crustal movement data. More specifically, the aim is to investigate the regional characteristics and temporal changes of individual analysis results, and finally integrate them to determine the anomaly of current crustal activity, evaluate earthquakes that have occurred, and estimate future transitions. We examined three indexes, such as the maximum shear strain rate, for the strain rate field of Japan estimated from GNSS data as analytical indexes for crustal movement data, referring to Savage and Simpson (1997) (Kimura et al., 2022, Seismological Society of Japan). Although strain rates in GNSS data have errors due to various factors, we have confirmed that they can be used to detect anomalies in crustal deformation. In this poster, we report on the status of subsequent studies.
We found that our method can detect known slow slips, but there are many cases where large strain rates are detected in coastal areas and specific areas (Kimura et al., 2022). The reasons for these large strain rates are thought to be related to the problem of estimation accuracy due to the distribution of the observation points and the temporary degradation of the environment at some observation points.
In order to eliminate these factors that adversely affect anomaly detection, we examined the following quality control methods:
・Eliminating velocity data that changes differently from surrounding observation points.
・Eliminating strain rate data with a large bias in observation point placement.
The former aims to exclude changes in appearance due to the deterioration of the observation environment, as the trends will be different from those at surrounding observation points. As a result, very small crustal deformations that change at only one point are excluded, but what we want to extract in this study are crustal deformations that appear at least at several points. In the latter case, we believe that the reason for the anomalous values in the coastal area is that the distribution of the observation points strongly affects the extrapolation and the estimation accuracy deteriorates.
These quality controls have reduced the abnormal strain rate values decreased by about 20%, making it easier to detect changes due to physical phenomena such as slow slip. However, there is a possibility that the excluded crustal deformation associated with seismic activity is included, and it is necessary to confirm the cause of the large strain rate and adjust the parameters when excluding it in the future.
We found that our method can detect known slow slips, but there are many cases where large strain rates are detected in coastal areas and specific areas (Kimura et al., 2022). The reasons for these large strain rates are thought to be related to the problem of estimation accuracy due to the distribution of the observation points and the temporary degradation of the environment at some observation points.
In order to eliminate these factors that adversely affect anomaly detection, we examined the following quality control methods:
・Eliminating velocity data that changes differently from surrounding observation points.
・Eliminating strain rate data with a large bias in observation point placement.
The former aims to exclude changes in appearance due to the deterioration of the observation environment, as the trends will be different from those at surrounding observation points. As a result, very small crustal deformations that change at only one point are excluded, but what we want to extract in this study are crustal deformations that appear at least at several points. In the latter case, we believe that the reason for the anomalous values in the coastal area is that the distribution of the observation points strongly affects the extrapolation and the estimation accuracy deteriorates.
These quality controls have reduced the abnormal strain rate values decreased by about 20%, making it easier to detect changes due to physical phenomena such as slow slip. However, there is a possibility that the excluded crustal deformation associated with seismic activity is included, and it is necessary to confirm the cause of the large strain rate and adjust the parameters when excluding it in the future.