Skip to content

Latest commit

 

History

History
155 lines (130 loc) · 5.9 KB

README.md

File metadata and controls

155 lines (130 loc) · 5.9 KB
Raw

NoiseCC.jl: A framwork for large ambient noise cross-correlation (CC) based on SeisNoise.jl.

🚨 The package is still undergoing development. 🚨

NoiseCC.jl use Multi-process and Multi-thread to accelerate noise cross-correlation (CC) for large dataset. All the core code for CC comes from SeisNoise.jl, and this package is an example of SeisNoise.jl application. We designed a generic CC framework which is suitable for large-scale Dense-Stations, Large-N and DAS data sets. It is convenient for the subsequent study of noise-imaging and noise-monitor.

Dependencies

The package NoiseCC.jl runs on Unix-like systems including Mac and Linux. Package need Julia 1.6 or greater. It depends on the following Julia modules:

Installation

Install SeisNoise firstly. Use the Julia package manager (Press ] to enter pkg):

julia>]
(@v1.6) pkg> add SeisNoise

Then install NoiseCC:

(@v1.6) pkg> add NoiseCC

Framwork

Prepare database

Before using NoiseCC.jl, you should prepare the database. Make sure the name of the file contains the exact network, station and component information. The following figure shows the format of other paths in the database.

  • Format: Year | Yearday | Time-Chunck | Data(sac/mseed)

database_format

Notice that the name of file 080_2020275000002.064_CHN.sac do not include Network name, which is not recommended.

Processing flow

Process_flow

Usage

NoiseCC.jl includes 2 modes of cross-correlation for seismic data.

  • FFT --> CC --> STACK
  • FFT_CC --> STACK

1-CC:

NoiseCC.FFT: Use Multi-process to accelerate FFT from raw-data (sac/mseed) to fft-data (FFTData). Parameters description here.

NoiseCC.FFT(
    # processes parameters
    cores,flag,precompile,
    # absolute path parameters
    FFTDIR,
    # select sac/mseed files or not
    select_file,
    # when select_file = ture, 'TMPTXT' works
    FILETXT,
    # when select_file = false, 'DATADIR' 'select_sta' 'STATXT' and 'ncomp' work.
    DATADIR,select_sta,STATXT,channel_regular,
    # instrument parameters
    rm_resp,respdir,
    # some control parameters
    input_fmt,time_norm,freq_norm,cc_method,
    # pre-processing parameters
    freqmin,freqmax,cc_len,cc_step,
    # criteria for data selection in time_norm
    factor_clip_std,factor_mute,time_half_win,
    # criteria for data selection in freq_norm
    freq_half_win,
)

NoiseCC.CC: Use Multi-process and Multi-thread to accelerate cross-correlation from fft-data (FFTData) to corr-data (CORRData). Parameters description here.

NoiseCC.CC(
    # processes and threads parameters
    cores,threads,flag,precompile,
    # absolute path parameters
    FFTDIR,CCFDIR,
    # select time-chunck folder for cross-correlate
    select_time,TIMETXT,
    # select station for cross
    select_sta,STATXT,SRTXT,channel_regular,comp,
    # some control parameters
    cc_method,auto_corr,cross_corr,
    # cross-correlation parameters
    maxlag,substack,substack_method,
)

NoiseCC.FFT_CC: Use Multi-process and Multi-thread to accelerate FFT and CC from raw-data (sac/mseed) to corr-data (CORRData). Parameters description here.

NoiseCC.FFT_CC(
    # processes and threads parameters
    cores,threads,flag,precompile,
    # absolute path parameters
    DATADIR,CCFDIR,
    # select time-chunck folder for cross-correlate
    select_time,TIMETXT,
    # select station for cross
    select_sta,STATXT,SRTXT,channel_regular,
    # instrument parameters
    rm_resp,respdir,
    # some control parameters
    input_fmt,time_norm,freq_norm,cc_method,auto_corr,cross_corr,
    # pre-processing parameters
    freqmin,freqmax,cc_len,cc_step,
    # cross-correlation parameters
    maxlag,substack,substack_method,
    # criteria for data selection in time_norm
    factor_clip_std,factor_mute,time_half_win,
    # criteria for data selection in freq_norm
    freq_half_win,
)

2-STACK:

NoiseCC.STACK: Use Multi-process and Multi-thread to accelerate STACK with corr-data (CORRData). Parameters description here.

NoiseCC.STACK(
    # process and threads parameters
    cores,threads,flag,precompile,
    # absolute path parameters
    CCFDIR,STACKDIR,
    # select time-chunck folder for stack
    select_time,TIMETXT,
    # select station for stack
    select_sta,STATXT,SRTXT,channel_regular,comp,
    # stack time interval
    stack_all,time_interval,min_stack_chunck,
    # stack method
    substack_method,stack_method,auto_stack,cross_stack,
    # filter 
    select_filter,freqmin,freqmax,corners,zerophase,
    # criteria for corr-data selection
    select_corr,median_high,median_low,
)

3-PLOT:

S0_read_jld2.py: Use Matplotlib for visualization, HDF5 files can be read by h5py package.

S1_plot_waveform.py:

waveform

S2_plot_moveout.py: working...

S3_plot_dailychange.py: working...

References

[1] Clements, T., and M. A. Denolle (2020). SeisNoise.jl: Ambient Seismic Noise Cross Correlation on the CPU and GPU in Julia, Seismol. Res. Lett. 92, 517–527, doi: 10.1785/ 0220200192.

[2] Jones, J. P., K. Okubo, T. Clements, and M. A. Denolle (2020). SeisIO: A Fast, Efficient Geophysical Data Architecture for the Julia Language, Seismol. Res. Lett. 91, 2368–2377, doi: 10.1785/0220190295.