In order to do quantitative modeling of the output of ROI_PAC, you will probably want to load the derived products into other programs. Below are some hints and software that will help you do this.

look and incidence angles

There are look angles in degrees for all four corners of the scene in the final geo*.unw.rsc file. For a PALSAR scene example from the most common "34.2" beam, the look angles vary from 32.8 at the near edge to 36.3 degrees on the far edge.

The important thing to remember is that, at least in ROI_pac, the look angle is defined as the angle from the vertical at the satellite, which is different from the angle at the ground (called the incidence angle) due to the curvature of the Earth. The difference is about 4-6 degrees, and you should use the incidence angle for any calculations of the LOS projection at the surface. The PALSAR "34.2" beam has a look angle at the center of 34.2 degrees and incidence angle around 39 degrees at the center. If you want a map of the exact incidence angle and LOS azimuth, you can run the "" script described below.

other hints

GOOGLE EARTH Perl scripts: and will convert a geocoded interferogram to a Google Earth kml file and png file Download the scripts here. This code requires that you have imagemagick installed. It's free and the "convert" command line utility is fantastically useful. This code also requires mdx.

GMT Various GMT scripts are available in the Contributed software section, particularly roipac2grdfile

MATLAB Example Matlab script for loading data in a variety of formats. Available in the Contributed software section

GIS For GIS software, use .rsc file to create metadata file and perhaps rmg2mag_phs to make a single binary file from the Richard M. Goldstein (rmg) format (real*4: including .unw, .hgt, .msk and .cor): each line has twice the number of columns specified. For each line, first there is a column width of "amplitude" values, then a column's width of "phase" values.

MAKE_LOS There is a script in ROI_PAC called that will generate a file that gives the incidence angle (the magnitude portion of the rmg file) and satellite heading (the phase portion) for each pixel. This information is essential for use in converted modeled ground displacements in x, y, and z to the projection in the satellite LOS.

VARRESPY: Variable Resolution Sampler attempts to reduce the redundancy in an unwrapped interferogram by identifying data samples that sufficiently represent the change in curvature of the surface deformation signal. This code implements the variable resolution resampler originally developed by Simons et al (2001) in Numpy/Scipy and includes some new extensions like an approximate covariance model. Package can be downloaded from GoogleCode.

Modeling (last edited 2013-01-03 02:19:42 by user-vcauugs)