NAME
dbdepest - estimates depth to Moho from a receiver function
SYNOPSIS
dbdepest z1 z2 dz Vcrust Fmin Fmax Delta Tfit Zbasin VpVs tshift inwf_file subset_rf [depest.pf]
SUPPORT
Contributed code: NO BRTT support.
THIS PIECE OF SOFTWARE WAS CONTRIBUTED BY THE ANTELOPE USER COMMUNITY. BRTT DISCLAIMS ALL OWNERSHIP, LIABILITY, AND SUPPORT FOR THIS PIECE OF SOFTWARE.
FOR HELP WITH THIS PIECE OF SOFTWARE, PLEASE CONTACT THE CONTRIBUTING AUTHOR.
DESCRIPTION
Reads a receiver function stored in a css3.0-database, and using a grid
search method estimates the crustal thickness that best fits that receiver
function in a sum-of-squares sense. The crustal thickness is varied in a
simple crustal model to find that which best predicts the observed waveform. See
Sheehan et al. [1995, JGR] for longer explanation and tests of an earlier
version of this program.
Model receiver functions are calculated, for each set of descriptive parameters,
using a propagator-matrix estimate of crustal response to incident plane waves
at its base. The model includes a single crustal layer, of specified
velocity, and optionally a shallow basin. Assumed parameters include mantle
P and S velocities of 8.0 and 4.62 km/s, mantle density of 3.3 g/cc, basin
P velocity of 3.9 km/s, basin density of 2.4 g/cc. Vp/Vs is specified if
there is no basin, or set at 1.73 if there is (the specified Vp/Vs then
applies to the basin). Crustal density is estimated from a linear fit to
Nafe-Drake empirical curves.
COMMAND LINE ARGUMENTS
-
z1
Shallowest depth to Moho to test, km.
-
z2
Deepest depth to Moho to test, km.
-
dz
Depth increment, km, for grid search over Moho depth.
-
Vcrust
Assumed crustal P-wave velocity, km/s.
-
Fmin
Minimum frequency or High-pass filter frequency, Hz. Both the observed and
calculated receiver functions are filtered before they are compared. Filter
is 3-pole zero-phase Butterworth.
-
Fmax
Maximum frequency of Low-pass filter frequency, Hz. See Fmin.
-
Delta
Distance, degrees, from source to receiver. Used for calculating ray parameter,
using iasp91.
-
Tfit
Length of the time window to evaluate fits, s, from start of the record.
-
Zbasin
Depth to bottom of a sedimentary basin, in km (Vp=3.9 km/s, rho = 2.4 g/cc).
-
VpVs
If Zbasin=0, the P-to-S ratio assumed for the crust. If Zbasin>0, the
P-to-S velocity ratio assumed for the basin sediments; in that case
Vp/Vs=sqrt(3). Also, if VpVs = 0 then Vp/Vs is set to sqrt(3) everywhere.
-
tshift
Time shift of zero lag from start of record, used in generating receiver
function. Should match "phaseshift" in dbrfcn or "tshift" in
dbtimerf.
-
inwf_file
Name of wfdisc-table with data of interest; should include (.wfdisc).
-
subset_rf
A logical string used to select from the database the receiver function of interest.
For example, a subset on station/channel/time.
-
[depest.pf]
Parameter file prefix -- can control inversion parameters here. Parameters
include:
(vpm, vsm, rhom) for Mantle;
(vpb, vsb, rhob) for Basin;
vpvs for crustal velocity ratio;
vpvssed for basin velocity ratio.
These parameters override defaults and command-line arguments.
OUTPUT
-
stdout
A 4-column table; columns are DEPTH, RMS1, NRMS1, RMS2. RMS1 is the Rms
residual between calculated and observed records; NRMS1 is RMS1 normalized
to power in the observed receiver function; RMS2 is the rms residual calculated
after BOTH the observed AND calculated receiver functions are normalized by
their RMS amplitudes.
-
syn.filt
An ASCII file with the best-fitting receiver function calculated at minimum RMS2.
the columns are Time, Amplitude. Record is filtered via Fmin, Fmax.
-
dat.filt
An ASCII file with the filtered observed receiver function. Format as syn.filt.
EXAMPLE
dbdepest 25. 65. 2. 6.5 .04 .2 35. 30. 4. 2.0 0. lmndb.wfdisc 0
Estimates crustal thickness every 2 km from 25 to 60 km depth. Assume a
crustal thickness of 6.5 km/s, bandpass signals from .04 to .2 Hz, assume
a 35 degree incidence angle, fit the first 30 s of record, assume a 4 km
thick basin with Vp/Vs = 2.0. No phase shift was placed in the records;
the first wfdisc record (0) is taken out of the database lmndb.
LIBRARY
-ltr -ldb -lstock -lcoords -ltttaup -lm -lteles
BUGS AND CAVEATS
Most of the details and parameter conventions were done to finish the
Rocky Mt. paper and should be made more general. As well, a better scheme
needs to be worked out to allow more complicated crustal models, grid
searches over other parameters, etc.
Filtering, especially low-pass filtering, is very important as at high
frequencies the synthetic seismogram calculation is unstable.
AUTHOR
Geoff Abers, Boston University;
some help from Anne Sheehan, CU, on an earlier version.
The origin of the
propagator matrix routines are lost to history.
Antelope User Group Contributed Software