Python codes to compare the retrieval of impulse response functions with different methods (Viens et al., 2017, GJI)
This repository contains the functions used in Viens et al. (2017) along with an example.
This repository contains:
- Codes folder:
- Functions_GJI_2017.py: Functions to compute cross-correlation, deconvolution, coherency of raw data, and cross-correlation of 1-bit data.
- Interferometry.py: runs the different techniques on 1 day of data recorded at "station_1" and "station_2".
- Data folder:
- station_1_1d.sac: 1 day of vertical data with a sampling rate of 4 Hz at station 1.
- station_2_1d.sac: 1 day of vertical data with a sampling rate of 4 Hz at station 2.
- Figures folder:
- Foo.png: Figure generated by Interferometry.py.
In this example, Station 1 and Station 2 are two MeSO-net stations located in the Kanto region, Japan. The distance between the two stations is 8.26 km. Therefore, the waves observed in the causal (positive) part between ~10 and 25 s in the figure below are likely Rayleigh waves. Note that the anti-causal (negative) and causal parts are strongly asymmetric due to the station locations, with Station 1 being closer to the Pacific Ocean than Station 2. Spurious arrivals near the zero time-lag can also be observed and disappear when the correlation functions are stacked over a longer time period.
