Parameter list

SEIS_FILO's programs involve a number of tuning parameters. The complete list is given here. These parameters should appear in the main parameter file or data summary file.

A

a_gauss

The parameter for the Gaussian low-pass filter.

  • Type: Double precision

C

cmax

The maximum phase velocity (km/s).

  • Type: Double precision

cmin

The minimum phase velocity (km/s). This value is used as the starting point of the root search for dispersion curves.

  • Type: Double precision

correct_amp

Whether the observed receiver functions (input data) is corrected for the energy loss by the Gaussian low-pass filter. See here for more details.

  • Type: Logical (T or F)

D

damp

Damping factor for water-level deconvolution. Default is 0.

  • Type: Double precision

dc

The phase velocity interval (km/s) used for the root search. This interval is used in the first-step rough search for zero crossings. The second-step binary search follows this rough search.

  • Type: Double precision

deconv_flag

Whether the forward computation accompanies deconvolution (T) or not (F). For deconv_flag=T, receiver functions are computed by the deconvolution of the radial-component deconvolved by the vertical component. For deconv_flag=F, the output is the radial-component impulse response.

  • Type: Logical (T or F)

delta

The sampling interval of receiver functions (sec).

  • Type: Double precision

dev_sig_c

The standard deviation of random perturbation amount added to the standard deviation of phase velocity data noise (km/s). Only used for inversion.

  • Type: Double precision

dev_sig_hv

The standard deviation of random perturbation amount added to the standard deviation of H/V ratio data noise. Only used for inversion.

  • Type: Double precision

dev_sig_ra

The standard deviation of random perturbation amount added to the standard deviation of Rayleigh admittance data noise. Only used for inversion.

  • Type: Double precision

dev_sig_rf

The standard deviation of random perturbation amount added to the standard deviation of receiver function data noise (km/GPa). Only used for inversion.

  • Type: Double precision

dev_sig_u

The standard deviation of random perturbation amount added to the standard deviation of group velocity data noise (km/s). Only used for inversion.

  • Type: Double precision

dev_vp

The standard deviation of random perturbation added to P-wave velocity (km/s)

  • Type: Double precision

dev_vs

The standard deviation of random perturbation added to S-wave velocity (km/s)

  • Type: Double precision

dev_z

The standard deviation of random perturbation added to layer interfaces (km)

  • Type: Double precision

diagnostic_mode

If .true. (or T), it activates diagnostic mode (see diagnostic_mode). Default is .false..

  • Type: logical

disp_phase

The phase-type of surface waves. L: Love-wave wave (not supported yet), R: Rayleigh wave.

  • Type: Character (length=1)

disper_in

The filename of the data summary file for dispersion curves.

  • Type: Character (length=200)

disper_out

The output filename for disper_fwd.

  • Type: Character (length=200)

dvp_sig

The standard deviation of Gaussian prior probability for Vp anomaly. This parameter is valid only if solve_anomaly is true.

  • Type: double precision

dvs_sig

The standard deviation of Gaussian prior probability for Vs anomaly. This parameter is valid only if solve_anomaly is true.

  • Type: double precision

dx

The frequency or period interval of a dispersion curve (Hz or s).

  • Type: Double precision

F

freq_or_period

Input type of dispersion measurements. This parameter must be either 'freq' or 'period' (freq: dispersion curve as a function of frequency; period: as a functon of period).

  • Type: Character(6)

I

i_seed1

The number to initialize pseudo-random numbers. The Xorshift algorithm is used for the random number generation.

  • Type: Integer

i_seed2

The number to initialize pseudo-random numbers. The Xorshift algorithm is used for the random number generation.

  • Type: Integer

i_seed3

The number to initialize pseudo-random numbers. The Xorshift algorithm is used for the random number generation.

  • Type: Integer

i_seed4

The number to initialize pseudo-random numbers. The Xorshift algorithm is used for the random number generation.

  • Type: Integer

is_ocean

Whether the velocity structure has the sea (T) or not(F).

  • Type: Logical (T or F)

is_sphere

Given (or output) velocity model is for spherical Earth (T) or not (F). When it is true, the earth-flattening transformation is applied to the given velocity model before forward computation. Consider setting this parameter to be "T" for studies on Mantle structure (>~ 100 km depth). See also this page

  • Type: logical

K

k_max

The maximum number of layer interfaces (excluding the seawater and bottom half-space).

  • Type: Integer

k_min

The minimum number of layers (excluding the seawater and bottom half-space).

  • Type: Integer

N

nx

The number of data point in a dispersion curve.

  • Type: Integer

n_bin_vp

The number of Vp bins to construct the posterior probability.

  • Type: Integer

n_bin_vs

The number of Vs bins to construct the posterior probability.

  • Type: Integer

n_bin_z

The number of depth bins to construct the posterior probability.

  • Type: Integer

n_burn

The number of iterations in the burn-in period. The sampled model is not saved for this period to forget the initial state.

  • Type: Integer

n_chain

The number of MCMC chains per process.

  • Type: Integer

n_cool

The number of non-tempered MCMC chains per process.

  • Type: Integer

n_corr

The interation interval at whith the sampled model is saved. This is useful to avoid artificial correlation to the previous sample.

  • Type: Integer

n_disp

The number of dipsersuion curves used for inversion. Need to specify in the first line of the data summary file.

  • Type: Integer

n_iter

The number of total iterations for the inversion, which includes the burn-in period.

  • Type: Integer

n_mode

The number of mode for surface waves. 0: fundamental mode, 1: 1st overtone, ...

  • Type: Integer

Note

If this parameter is set to be less than 0, full search mode is activated. See here for more details.

n_proc

The number of processes. NOTE: this parameter should be passed as command line arugment: mpirun -np [n_proc] bin/joint_int joint_inv.in, for example.

Warning

DONOT put this parameter in any input files.

  • Type: Integer

n_rf

The number of receiver functions use for inverions. Need to specify in the first line of the data summary file.

  • Type: Integer

n_smp

The number of elements in receiver functions.

  • Type: Integer

noise_added

Standard deviation of noise added to forward computation results.

  • Type: Double precision

O

ocean_thick

The thickness of the ocean layer (km)

  • Type: Double precision

R

r_earth

The earth's radius in kilo-meter used for the earth-flattening transformation. The default value is 6371.

  • Type: double precision

rayp

Ray parameter (s*km^-1).

  • Type: Double precision

recv_func_in

The filename of the data summary file for receiver functions.

  • Type: Character (length=200)

recv_func_out

The output SAC filename for recv_func_fwd.

  • Type: Character (length=200)

ref_vmod_in

A filename that contains a reference velocity model. This parameter is valid only if solve_anomaly is true. The file format is here.

  • Type: Character (length=200)

rf_phase

The phase-type of receiver functions: P (P-wave); S (S-wave).

  • Type: Characeter (length=1)

rho_bottom

The density of the bottom half-space layer (km/s). If a negative value is given to this parameter, the density of the bottom layer is determined empirically from its Vp.

  • Type: Double precision

rho_ocean

The density of the ocean layer (g/cm^3). The default value is 1.0 g/cm^3.

  • Type: Double precision

S

sig_c_max

The maximum standard deviation for data noise in phase velocity (km/s). Only used for inversion.

  • Type: Double precision

sig_c_min

The minimum standard deviation for data noise in phase velocity (km/s). Only used for inversion.

  • Type: Double precision

sig_hv_max

The maximum standard deviation for data noise in H/V ratio. Only used for inversion.

  • Type: Double precision

sig_hv_min

The minimum standard deviation for data noise in H/V ratio. Only used for inversion.

  • Type: Double precision

sig_ra_max

The maximum standard deviation for data noise in Rayleigh admittance (km/GPa). Only used for inversion.

  • Type: Double precision

sig_ra_min

The minimum standard deviation for data noise in Ryaleigh admittance (km/GPa). Only used for inversion.

  • Type: Double precision

sig_rf_max

The maximum standard deviation for data noise for receiver function amplitude. Only used for inversion.

  • Type: Double precision

sig_rf_min

The minimum standard deviation for data noise for receiver function amplitude. Only used for inversion.

  • Type: Double precision

sig_u_max

The maximum standard deviation for data noise in group velocity (km/s). Only used for inversion.

  • Type: Double precision

sig_u_min

The minimum standard deviation for data noise in group velocity (km/s). Only used for inversion.

  • Type: Double precision

solve_anomaly

If this parameter is set to ".true." (or "T"), velocity anomalies are solved for instead of the absolute values. These anomalies are defined relative to a reference velocity model given by users. See here for more details.

  • Type: Logical

solve_vp

Whether Vp is solved for (T) or not (F).

  • Type: Logical (T or F).

T

t_pre

Time length (sec) preceding the direct P arrival on receiver functions.

  • Type: Double precision

temp_high

The maximum temperature allowed for parallel tempering scheme.

  • Type: Double precision

V

vmod_in

Input filename that contains a velocity model for forward computations. See this page for the file format.

  • Type: Character (length=200)

vp_bottom

Vp of the bottom half-space layer (km/s). Set this parameter to be negative (and solve_vp to be .true.) if you want to solve for the Vp of this layer.

  • Type: Double precision

vp_ocean

Vp of the ocean layer (km/s). The default value is 1.5 km/s.

  • Type: Double precision

vp_max

The maximum Vp (km/s)

  • Type: Double precision

vp_min

The minimum Vp (km/s)

  • Type: Double precision

v_ref_mod_in

Filename for a reference velocity model. This is used when solve_anomaly = T. A required format for the reference model file is here.

  • Type: Character

vs_bottom

Vs of the bottom half-space layer (km/s). Set this parameter to be negative if you want to solve for the Vs of this layer.

  • Type: Double precision

vs_max

The maximum Vs (km/s).

  • Type: Double precision

vs_min

The minimum Vs (km/s).

  • Type: Double precision

X

xmax

The maximum frequency or period of a dispersion curve (Hz or s).

  • Type: Double precision

xmin

The minimum frequency or period of a dispersion curve (Hz or s).

  • Type: Double precision

Z

z_max

The maximum depth of layer interfaces (km below the sea surface).

  • Type: Double precision

z_min

The minimum depth of layer interfaces (km below the sea surface). NOTE: z_min must be deeper than the sea bottom for ocean setting.

  • Type: Double precision