Example:

Data directories with SAC traces:
---------------------------------
2010-E085a/SAC_TRACES
2010-E085b/SAC_TRACES
2010-E085c/SAC_TRACES
2010-E085d/SAC_TRACES
Each of these directories contains 10 days of 3-C data for the
Spanish station E085 (Cuenca, E-Spain). The data were cut into
2h segments and are stored in daily subdirectories.

Script to process SAC_TRACES:
-----------------------------
run_polfre.cmd
Script to determine polarization attributes for all data in
the defined data directory (e.g., 2010-E085a/SAC_TRACES).
Results are written to a "Results" directory (e.g., 
2010-E085a/Results). Edit the script
to change the data directory and/or the input parameters
of the polarization program.

Directories for data analyses and figures:
------------------------------------------
Fig_baz_rose
Fig_freq_baz
Fig_pol_spec
These 3 directories contain scripts to extract the different
information from the "Results/" directories and to plot this 
information. See my example figures! Look also at Schimmel et 
al. (2011) to see what else one can extract from these files. 

How to proceed ?
----------------
(1) Understand the directory structure, look at the data, ... . 
(2) Understand/adapt/execute run_polfre.cmd.
    You will have to specify the data directory. Right now
    "cd 2010-E085a/SAC_TRACES" is being used.
(3) Look at the results, understand data format.
(4) Go to each Figure directory, understand example figure and 
    script, execute script. 

Practical:
----------
The practical consists in reproducing these results, computing the
polarizations for the remaining 2 time spans and to start changing
the parameters to further understand their meaning and influence.
The absolute number of detected signals will change for the different
settings and we therefore consider only relative occurrences. In a
next step, the script should be adapted for more data and other stations
to analyze the seasonal variability of ambient noise.

Some published results:
-----------------------
The polfre software has been used in different studies to characterize
noise and to monitor their sources (Schimmel et al. 2011;
Obrebski et al., 2012; Sergeant et al., 2013; Davy et al., 2015).
In Agurto-Detzel et al. (2016), the detection of a tailings dam
failure has been shown. These studies show different applications
which may inspire new and interesting studies.

Martin Schimmel (schimmel@ictja.csic.es)