Generate an interferogram for your chosen area of interest in Western US.

Steps:

1) Select your area and check the data availability.

 1a. Create a subdirectory <orb_info> in your work directory on /sword/usr

 1b. Copy files <bounds, reference.gmt, and base.m> from  
     /sword/usr/insar/yuri/orb_info to your orb_info dir

 1c. Edit file <bounds> to specify coordinates of your geographic area
     (W and S are negative)
 
 1d. Run <ers_pick.pl bounds g> to extract all available data from 
     the ERS catalog. <g> stands for the ERS operational phase (there
     were several phases, a, b, c, ...etc., and g is all-inclusive).

 1e. <cut_col.pl g> and <ers_subpick.pl g> to generate coordinates of 
     available ERS scenes (each scene, or frame, is a radar swath about 
     100 km x 100 km). 

 1f. Make a map of your area with satellite frames on top. Specify
     lat/lon bounds for your area in file reference.gmt. Then run
     <./reference.gmt g 1 s> (all of the above scripts will print 
     a usage statement if no arguments are given). 

 1g. Look at the map and select the scene(s) (track# and frame#) you 
     want to process. Edit file des_tf.g to delete all other scenes.
     Re-run <ers_subpick.pl g>.

 1h. Edit <base.m> to specify the name of your area (optional). Start
     Matlab and run <base> to produce a baseline plot. Data available
     in our WInSAR archive will be shown in red. Select your
     interferometric pair (satellite number, A=ERS1, B=ERS2, and orbit
     number).

2) Copy and pre-process the raw archive data. 

 2a. Go up to you working directory, and copy the raw data files for 
     the master image from
     /net/winsar/T###/####/, where T### is the track number, and ####
     is the frame number. The file names are e#_#####*, where e# is the 
     ERS satellite number (i.e., e1 or e2), and ##### is the orbit number. 
     For example, to copy data from track 127, frame 2907, and ERS-1 
     orbit 23662, 
     <cp /net/winsar/T124/2907/e1_23662* .>
     Each scene should have 3 files (extensions .raw.gz, .vdf, .ldr).
     Repeat for the slave image. 

 2b. Uncompress and prepare the raw data files for processing.
     Run <sio2roi.pl> in the same dir. This should produce 
     subdirectories <yymmdd> (yy=year, mm=month, dd=day) for each 
     acquisition. 

 2c. Create the input process file int_date2_date1.proc, where
     date2 and date1 correspond to the radar acquisition dates
     (in yymmdd format); either copy and modify your existing
     *proc file, or run <cp_proc.pl "your_existing".proc date2 date1>

3) If your area is not in Southern California, you need to make your
   own Digital Elevation Model (DEM). 
 3a. Create a directory called DEM, go there, and run 
     <mosaic_srtm.pl -Rw/e/s/n "YourDEMname">, where w, e, s, and n
     denote the western, eastern, southern, and northern boundaries of your
     DEM. Make sure that the satellite frames are covered by the DEM (but
     don't be too generous - large DEMS take a lot of space, and slow down
     processing).

 3b. Edit your int_*proc file to include the location of your new DEM.  

4) Process the data using ROI_PAC (you've done this before!)
  <process_2pass.pl int_date2_date1.proc &>
  If there are no problems with the data, you should have a geocoded
  unwrapped interferogram in about 1 hour. If the program fails, most
  likely reasons are: the focused SLC images were not successfully
  aligned, or the images did not have enough correlation to allow
  interferometry (e.g., changes in surface reflectivity are too
  dramatic, baseline is too large, etc.). Show me your intermediate
  results, or choose another area.
  A map showing WInSAR acquisitions is available at
   http://topex.ucsd.edu/winsar/figures.pdf

Your results should include: 
1. A map of your area showing satellite frame(s) used.
2. Baseline plot(s). 
3. Image of radar correlation (date2-date1.cor). 
4. Georeferenced unwrapped interferogram. 
Either print-outs, or Post-Script/jpeg files would be fine.