11:45 AM - 12:00 PM
[S03-06] Duration of Transient Deformation in Northeast Japan as Observed by Geodetic Data from 1890 to 2010
Crustal deformation has been transformed in the past decades by the availability of continuous GPS data. Time series analysis of GPS data often assumes strain accumulation at a steady rate. In the case of NE Japan, analysis of horizontal GPS data from 1996-2010 revealed a decrease in trench perpendicular strain-rate in the decade prior to the 2011 Tohoku-oki earthquake (Mavrommatis et al., 2014). However, the limited duration of the observations makes it difficult to assess how these transients relate to the full interseismic period. Importantly, we do not know whether the strain-rate at the beginning of the GPS era (1996-1998) is typical of the preceding decades or not. It could be that strain-rate during this period was anomalously high.
Conventional geodetic data (triangulation and trilateration) are available in Japan since the end of the 19th century. The data consist of two first-order triangulation surveys (1883-1901; 1946-1972) and two trilateration campaigns in the latter half of the 20th century (1973-1985; 1977-1994), including second-order surveys. Examination of historical data in combination with GPS data can provide a way to evaluate the velocity and strain distributions in the last 100 years and elucidate whether the acceleration observed prior to the 2011 Tohoku-oki earthquake represents a short-term transient deformation or not. Thus, we analyze horizontal deformation in NE Japan utilizing all the geodetic sets available.
We estimate horizontal displacement fields in NE Japan for the conventional geodetic data using the method developed by Yu and Segall (1996). In this, a denuisancing procedure is performed so that the dependency of the data on station coordinates is projected out of the system of equations and the horizontal displacement fields can be estimated directly. We solve for the inner coordinate or minimum norm estimate, representing the essential displacement after removing rigid motion and scale factor (in the case of triangulation) which fall in the null space of the observations. These results are transformed into average velocities by assuming steady deformation over the survey period and average shear strain rates are calculated for the region. The null space inherent in trilateration and triangulation data is constrained by GPS velocities in either the early GPS period or the period prior to the 2011 Tohoku-oki earthquake.
We found that the maximum shear strain rate in South-Central Tohoku from 1884-1984 was not higher than those observed during the GPS era. Thus, suggesting that the transient deformation described by Mavrommatis et al. (2014) did not begin prior to 1996. Furthermore, the maximum shear strain rates from conventional geodetic data are consistent with those prior to the 2011 Tohoku-oki earthquake, while the shear strain rate in the early stage of the GPS data may represent an anomalously high strain rate from the average interseismic situation, having geophysical implications related to precursory phenomena of the giant earthquake.
Conventional geodetic data (triangulation and trilateration) are available in Japan since the end of the 19th century. The data consist of two first-order triangulation surveys (1883-1901; 1946-1972) and two trilateration campaigns in the latter half of the 20th century (1973-1985; 1977-1994), including second-order surveys. Examination of historical data in combination with GPS data can provide a way to evaluate the velocity and strain distributions in the last 100 years and elucidate whether the acceleration observed prior to the 2011 Tohoku-oki earthquake represents a short-term transient deformation or not. Thus, we analyze horizontal deformation in NE Japan utilizing all the geodetic sets available.
We estimate horizontal displacement fields in NE Japan for the conventional geodetic data using the method developed by Yu and Segall (1996). In this, a denuisancing procedure is performed so that the dependency of the data on station coordinates is projected out of the system of equations and the horizontal displacement fields can be estimated directly. We solve for the inner coordinate or minimum norm estimate, representing the essential displacement after removing rigid motion and scale factor (in the case of triangulation) which fall in the null space of the observations. These results are transformed into average velocities by assuming steady deformation over the survey period and average shear strain rates are calculated for the region. The null space inherent in trilateration and triangulation data is constrained by GPS velocities in either the early GPS period or the period prior to the 2011 Tohoku-oki earthquake.
We found that the maximum shear strain rate in South-Central Tohoku from 1884-1984 was not higher than those observed during the GPS era. Thus, suggesting that the transient deformation described by Mavrommatis et al. (2014) did not begin prior to 1996. Furthermore, the maximum shear strain rates from conventional geodetic data are consistent with those prior to the 2011 Tohoku-oki earthquake, while the shear strain rate in the early stage of the GPS data may represent an anomalously high strain rate from the average interseismic situation, having geophysical implications related to precursory phenomena of the giant earthquake.