10:45 AM - 12:15 PM
[STT39-P02] Monitoring ground deformation in Kawasaki City, Kanagawa Prefecture, Japan, using interferometric SAR time series analysis of Sentinel-1 data
Keywords:InSAR time series analysis, Sentinel-1, large-scale embankments, Kawasaki City, Kanagara Prefecture
Landslides have occurred frequently in recent years due to frequent heavy rains. In July 2021, a landslide occurred in Atami City, Shizuoka Prefecture, resulting in the deaths and disappearances of many people.Artificial embankments exist in both mountainous and plain areas, and their soft soil can cause damage to residential areas in urban areas due to sliding and collapsing during earthquakes and other disasters.Therefore, displacements in embankment areas should be monitored carefully to protect the lives and property of those living there.
For example, there are more than 1,000 large-scale embankment areas in Kawasaki City, Kanagawa Prefecture. Therefore, it is difficult to monitor all of them by ground observation and survey. On the other hand, the interferometric analysis of satellite-based synthetic aperture radar (InSAR), which has recently been usedin various fields such as earthquakes and volcanic eruptions, makes it possible to monitor surface displacement as if scanning without ground observations. Therefore, it is expected to be useful for extracting displacements of objects that are distributed locally and countlessly, such as embankment areas. In this study, we performed an InSAR time series analysis using Sentinel-1 data to extract surface displacements in a large-scale embankment area in Kawasaki City as a model area.
We performed PS-InSAR time series analysis by using 142 scenes of ascending and 153 scenes of descending orbits, excluding scenes that are considered to be strongly affected by atmospheric delays, from data observed by Sentinel-1 from 2014 to the end of 2021. The obtained velocities in each satellite line-of-sight direction were combined to estimate the quasi-EW and quasi-UD components. In order to evaluate whether an embankment area is displaced or not, it is necessary to estimate the displacement relative to the surrounding area, which includes regional displacement. Therefore, in this study, the long-wavelength component was removed from the quasi-UD component to extract the subsided areas relative to their surroundings. Moreover, we compared the analysis result with the large-scale embankment areas that Kawasaki City detected.
The results showed that some of the large-scale embankment areas in the inland area of Kawasaki City subsided relatively more than the surrounding areas. Of 1093 large-scale embankment areas, 74 areas showed a subsidence rate of 2 mm/yr or more, and 16 areas showed a subsidence rate of 3 mm/yr or more. In order to detect the signs of sliding collapse, it is necessary to detect 3D ground displacement in detail on embankments to understand the existence of sliding displacement. In the future, we plan to conduct GNSS observations in areas where significant subsidence was observed and conduct integrated analysis with these observations.
For example, there are more than 1,000 large-scale embankment areas in Kawasaki City, Kanagawa Prefecture. Therefore, it is difficult to monitor all of them by ground observation and survey. On the other hand, the interferometric analysis of satellite-based synthetic aperture radar (InSAR), which has recently been usedin various fields such as earthquakes and volcanic eruptions, makes it possible to monitor surface displacement as if scanning without ground observations. Therefore, it is expected to be useful for extracting displacements of objects that are distributed locally and countlessly, such as embankment areas. In this study, we performed an InSAR time series analysis using Sentinel-1 data to extract surface displacements in a large-scale embankment area in Kawasaki City as a model area.
We performed PS-InSAR time series analysis by using 142 scenes of ascending and 153 scenes of descending orbits, excluding scenes that are considered to be strongly affected by atmospheric delays, from data observed by Sentinel-1 from 2014 to the end of 2021. The obtained velocities in each satellite line-of-sight direction were combined to estimate the quasi-EW and quasi-UD components. In order to evaluate whether an embankment area is displaced or not, it is necessary to estimate the displacement relative to the surrounding area, which includes regional displacement. Therefore, in this study, the long-wavelength component was removed from the quasi-UD component to extract the subsided areas relative to their surroundings. Moreover, we compared the analysis result with the large-scale embankment areas that Kawasaki City detected.
The results showed that some of the large-scale embankment areas in the inland area of Kawasaki City subsided relatively more than the surrounding areas. Of 1093 large-scale embankment areas, 74 areas showed a subsidence rate of 2 mm/yr or more, and 16 areas showed a subsidence rate of 3 mm/yr or more. In order to detect the signs of sliding collapse, it is necessary to detect 3D ground displacement in detail on embankments to understand the existence of sliding displacement. In the future, we plan to conduct GNSS observations in areas where significant subsidence was observed and conduct integrated analysis with these observations.