Introduction to InSAR

Interferometric synthetic aperture radar (InSAR) is a satellite remote sensing technique capable of measuring mm to cm scale surface displacements. The technique is explained here briefly, but the interested reader can find more information in Refs 2, 7, and 9.

Several satellites are on polar orbits around the earth constantly taking C- and L- band radar images. Their antennas have an inherent spatial resolution on the ground of several kilometers, far too large to be of much use in many studies. However, by post processing the data to take into account the timing at which various pulses are returned and the Doppler shift of these radar pulses the spatial resolution can be greatly improved. With this so call synthetic aperture radar (SAR) the spatial resolution can be reduced to several meters, which make this a practical measuring tool for many geologic applications.

In additions to recording the amplitude data of the returning pulses (which is useful for creating imagery, probing surface properties, and other applications) the phase of the pulses are also recorded. By taking two images from successive satellite passes of a particular region the two sets of phase information may be combined to form an interferogram that would map the changes in surface, along the look direction of the satellite, to a fraction of the radar wavelength (several mm).

There are several complications to this method. Firstly, the topography of the scene needs to be removed to help flatten the phase. A digital elevation modeled is created from the shuttle radar topography mission (SRTM) data to subtract out any effects from elevation. Secondly if the satellite is in different positions while it is acquiring the two scenes it can create artificial phase signal that must be subtracted. To minimize this effect the distance between the two positions in the perpendicular direction to the look direction, called the perpendicular baseline, should be kept small. Additionally other factors such as atmospheric effects, electrical noise, and vegetation can cause decorrelation in the phase signal.