Monitoring deformation due to volcanic activity is important for understanding volcanic behavior and for predicting eruptions. The Global Navigation Satellite System (GNSS) is sometimes used for such purposes. However, the system requires the installation of sensors at the site, and the installation and maintenance of these sensors are usually difficult and dangerous when volcanoes in the area are very active. On the other hand, Deferential Interferometry Synthetic Aperture Radar (DInSAR) can be used to measure the surface deformation of volcanoes without the installation of any devices on the ground. Moreover, the time-series DInSAR analysis has the potential to reveal the deformation behavior of volcanoes at the pre- and post-eruption stages. However, the DInSAR results can be greatly affected by the temporal variation in the refractivity in the tropospheric layer, especially when there are large differences in height in the target area of the measurements. Although it is a fundamental issue of DInSAR, it is still difficult and there are only a few methods for removing the errors caused by the tropospheric delays of radar pulse waves. This paper discusses this issue and proposes two simple methods for improving the measurement accuracy. The proposed methods were applied to Sakurajima Volcano in Japan. Ground surface deformation was observed for three years, from November 2014 to August 2017, by a time-series DInSAR analysis using Sentinel-1 data (operated by the European Space Agency). In this way, the effect of the tropospheric delays was successfully reduced.
As a country that is located in the "ring of fire", Japan has dozens of the active volcanoes. A volcano can be disastrous when it erupts, especially one that lies near a city. Therefore, the monitoring of the volcano's behavior is very important for safety and risk management. The activities of a volcano usually lead to ground surface deformation. Thus, a volcano's behavior can be monitored through the observation of its surface deformation.
Monitoring of the ground surface deformation can be conducted using a traditional measurement method, i.e., leveling or modern technologies such as GPS. Using the proper GPS method (i.e., the method developed by Shimizu et al., 2014) for surface displacement monitoring enables the yielding of continuous and highly accurate results. However, the method necessitates the installation of GPS sensors at the monitoring site. The installation and maintenance of the devices on the ground involve high costs and a great labor force. In addition, the installation and maintenance of the devices under tough and harsh conditions, like active volcanoes, can be dangerous.