El-Major Cucapah earthquake in Baja California, Apr 4, 2010.

In this example, we give the data, track command files, and processing 
scripts which can be used to generate results similar to those shown in 
the track kinematic examples.

The complete solution can be run with the  shell script.
./Run_BAJA.cmd
will regenerate the results and make GMT plots with the same time range 
as shown in the slides.

All needed files are in the directory and the results generated at MIT 
are in the MIT_results folder.
The rinex data are in the rinex directory with data spanning 3-hours 
from the day before the earthquake (093) and the day of the earthquake (094).  

There are two track command files  track_5Hz.cmd and track_1Hz.cmd.  
The 5Hz command file uses P725 as the base station (affected by surface 
wave arrivals about 220 seconds after the earthquake) and the 1Hz versions 
uses the 1Hz data (no 5Hz are available) from the P030 site which did not 
see any surface waves.  Both command files use the -s options to set options 
for the run in track.

The csh script sh_extts extracts the position estimates into a simplified 
format (that can be used in a graphics program called KaleidaGraph) with 
the time given as seconds after the event time.  While looking at this script 
note the conversion from UTC to GPS time for the event.  In high rate analysis, 
the difference between UTC and GPS time can be critical.  The shell script also make plots using GMT5.  These commands can be modified to make your own plots e.g., all stations could be plotted together with y-axis  offsets applied to separate the different stations. 

The SP3 orbit file supplied can generated with a sh_get_orbits call for 
days 093 and 094 and the cat unix command used to concatenate the files.  
The ionospheric delay model files can be down loaded with sh_get_ion. 

The Run_BAJA.cmd scripts shows an example of how earthquake data can be 
processed.  Due to the long baselines, the idea is to resolve ambiguities 
with the kinematic site treated as non-stochastic outside of the time when 
co-seismic offsets and surfaces waves are occurring.  These ambiguities 
are then used (with the -a option) in the analysis with the kinematic site 
fully stochastic.  

The Run_BAJA.cmd script gives you a starting point for exploration of the 
nature of track solutions on long baselines. Some of the questions to 
explore are:
(1) What do the results look like for day 093?  Can these data be used 
    to reduce multipath effects but shifting them by ~400 seconds (so called 
    sidereal filtering).
(2) What happens if the ionospheric model is not used (option 5 in the 
    -s options, anything other than the word space will stop the ionospheric 
    delay model being used.
(3) What if the float_type command options are modified to reduce the 
    weight given to the EX-WL values (reducing the dependence on the ionospheric 
    delay modeling error being small)
(4) What is the ambiguities are not resolved first with the quasi-static 
    solution?