seisgo.monitoring

The monitoring module measures relative seismic velocity changes (dv/v) from ambient-noise cross-correlations using time-domain stretching or wavelet-based methods.


Function Summary


get_dvv Parameter Reference

Parameter

Default

Description

corrdata

required

seisgo.types.CorrData object containing the cross-correlation time series.

freq

required

[fmin, fmax] — frequency band for measurement (Hz).

win_len

required

Coda window length in seconds. Accepts a scalar or list for multiple windows.

ref

None

Reference trace. If None, computed by stacking all windows with stack_method.

stack_method

"linear"

Stacking method to compute the reference. See seisgo.stacking.

offset

1.0

Lag offset (s) from the main arrival (xcorr) or from zero (autocorr). Accepts a scalar or list aligned with win_len.

resolution

None

Temporal sub-stack resolution (s). Resamples the data before measuring dv/v.

vmin

1.0

Minimum group velocity (km/s) used to set the start of the coda window.

normalize

True

Normalize traces before measuring.

whiten

"no"

Spectral whitening: "no", "rma", or "phase_only".

method

"wts"

dv/v method: "wts" (wavelet stretching) or "ts" (time-domain stretching).

dvmax

0.05

Maximum search range for dv/v (5% by default).

subfreq

True

If True (wts only), return frequency-dependent dv/v.

nproc

None

Number of parallel processes. None = serial.

save

False

If True, write seisgo.types.DvvData to outdir.

format

None

Output format: "asdf" or "pickle". Auto-detected from outfile extension.


dv/v Methods

Wavelet-transform stretching ("wts")

The default method. Decomposes each coda window into the time–frequency domain, applies a stretching operator at each scale, and finds the stretch factor that maximises cross-correlation with the reference. Returns a dv/v value per frequency band.

Credit: adapted from Congcong Yuan (Harvard), originally by Chengxin Jiang and Marine Denolle.

Time-domain stretching ("ts")

Applies a uniform time-axis stretch to the full coda window and maximises the cross-correlation coefficient against the reference. Returns a single broadband dv/v value. Set subfreq=False automatically when method="ts".


Output: DvvData

get_dvv() returns a seisgo.types.DvvData object (or saves it to disk).

Key attributes:

Attribute

Description

dvv

dv/v array, shape (n_windows,) or (n_windows, n_freqs)

err

Measurement uncertainty array (same shape as dvv)

cc

Cross-correlation coefficient of each measurement

time

Timestamps of each measurement window

freq

Frequency vector (for wts) or scalar band

method

Method used ("wts" or "ts")

cc_comp

Component pair (e.g. "ZZ")

dvvdata.plot()
dvvdata.to_asdf("dvv_output.h5")