Program: MKLSLOCATIONDEMO
mklslocationdemo...........demonstrate LOCSMITH location with artifical data
call:
[lsopt,mygrid,seisdat]=mklslocationdemo;
[lsopt,mygrid,seisdat]=mklslocationdemo('demo1');
[lsopt,mygrid,seisdat]=mklslocationdemo('demo2');
[lsopt,mygrid,seisdat]=mklslocationdemo(slon);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat,hdep);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat,hdep,dbaz);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat,hdep,dbaz,bazbias);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat,hdep,dbaz,bazbias,dt,tbias,phase);
[lsopt,mygrid,seisdat]=mklslocationdemo(slon,slat,selv,hlon,hlat,hdep,dbaz,bazbias,dt,tbias,phase,initwdt);
(sorry for not having a property name / property value technology here)
'demo1': a predefined demo, using the lunar apollo network
configuration and an arbitrary surface event.
This demonstrates the "interior" search method.
Only Azimuth data is used.
'demo2': a predefined demo, using a tetrahedron configuration and a
surface event close to one of the stations. This demo shows the
"Attention Creep Effect": The initial grid points to an area
which is far from the real solution, but using the
neighborhood iteration method, the attention of the
program creeps into right direction.
I found this configuration by trying some random inputs,
therefore it is not a perfect example (takes a bit long to
compute).
This also demonstrates the boundary search method.
Only Azimuth data is used.
'demo3': a predefined demo, using the Apollo network configuration
and an arbitrary deep event.
Only Travel Time data is used.
slon: (numeric array) [deg]
vector of station longitudes in 0...360deg
slat: (numeric array) [deg]
vector of station latitudes
selv: (numeric array) [km]
vector of station elevations, positive upwards
hlon: (number) [deg]
Hypocenter longitude in 0...360deg
hlat: (number) [deg]
Hypocenter latitude
hdep: (number) [km]
Hypocenter depth, positive downwards
dbaz: (numeric array) [deg]
back azimuth uncertainty of each station, in degrees
This defines how wide the belt around the true BAZ is, which
is used in location. Simulates BAZ measurement error.
Use the empty martrix [] to switch off azimuth evaluation
bazbias: (numeric array) [deg]
A station specific bias which is added to the theoretical BAZ
values. Also simluates BAZ measurement error.
dt: (numeric array) [s]
arrival time uncertainty of each station, in seconds
tbias: (numeric array) [s]
arrival time biases of all stations, in seconds
phase: (string array)
seismic phases assumed to be detected at all stations
arrival times are computed for these phases.
initwdt: (number) [km]
initial mesh width on the outermost layer of the grid
DEFAULT values are returned by MKLSCONSTANTS('LocationDemo1').
result: lsopt: (struct)
The LSOPT structure used for location
mygrid: (struct)
The grid strucure resulting from the location.
seisdat: (struct)
The LSD structure used for location.
This routine computes azimuth and arrival time data for the given
hypocenter and station network and then runs MKLOCSMITH to locate the
hypocenter using the given data.
If you don't want azimuth data to be generated, use the empty matrix []
as input for DBAZ.
This is a demonstration and test environment for MKLOCSMITH.
The routine generates
figure(1): a map of the Epicenter, station network and
Azimuthal gap (which is a widely used proxy for achievable
location accuracy)
figure(2): view of the search grid evolution during the location process
figure(3): 3D view of the final grid
Martin Knapmeyer, 17.03.2005, 08.03.2007
Read M-File Source Code
eof.