GNSS is strong and common observation tool for understanding the spatial and temporal development of crustal deformation. The one of the advantage of GNSS is the high dynamic range with broadband characteristic. Based on these characteristics, various studies developed the quasi realtime coseismic fault determination method deduced from the real-time GNSS data. Crustal deformation due to coseismic slips at an earthquake fault detected by real-time GNSS analysis is quite useful in estimating fault expansion and the amount of its slip. It in turn contributes to rapid tsunami warning for near field coast, even though the huge interplate earthquakes such as larger than the magnitude 9 event. This paper summarizes the author’s past research on improvement and its comprehensive noise assessment of real-time kinematic GNSS analysis and development of rapid coseismic fault determination technique based on those data. The author also reviews recent development status on crustal deformation monitoring system in Japan, which have been jointly developed by Geospatial Information Authority of Japan and Tohoku University. Furthermore, the author also discusses a future direction of the real-time crustal deformation monitoring and near field tsunami early warning system based on the GNSS data.

Behavior and instability of a GEONET pillar under strong seismic ground motion was evaluated. Coseismic displacement at the 0513 GEONET site due to a M5.0 local earthquake with 5- intensity showed four times larger value than predicted one from theoretical estimation. Inclination of the pillar due to strong seismic ground shaking was a candidate of the tilt change because little tilt change was expected from earthquake faulting. Corrected coseismic displacement was well consistent with that from theoretical value, and suggested unstable behavior of this pillar by strong ground shaking. High correlation between temperature and coordinate fluctuations for three years data also might request correlation of pillar movement for GNSS coordinate time series using tiltmeter data.

A Doppler frequency has been used to track a spacecraft and to know the orbit. FFT （Fast Fourier Transforms） gives a mean spectrum for a period of Fourier transforms so that it is a useful tool to estimate a Doppler frequency from a received signal. Phase variations of the signal in the period would be derived if we need to know the frequency variation of the signal in the period. FFT enables us to estimate phase variations of the signal in the period, when the signal is frequency-converted by software. The software for phase estimation of S- and X-band carrier signals with very narrow bandwidth （e.g. several ten Hz） has been developed, using time series data sampled at two hundred ksps （kilo-sample per second） after frequency conversions by hardware. This software can be applied to detect the lunar ionosphere with very small TEC variations by tracking phases of two frequency signals from a lunar orbiter. It can be widely used to measure phases of a carrier signal from a satellite, for example, GNSS or others with high accuracy.