Intro to Radar Interferometry

There is a good paper.

Radar, an acronym for "radio detection and ranging", is installed on the aircraft or satellite to image the earth surface. The first scientific application started from Seasat satellite in 1978, (artist figure to the right) though it only flew for a few months. The physics in radar imaging yield a special coordinate system in two direction: range and azimuth. Range axis is defined by the round-trip travel time of the electromagnetic (EM) echoes, while azimuth axis is defined by the Doppler shift. The satellite transmit and receive EM waves with wavelength in the range of C band (6 cm), X band (3 cm), L band (24 cm). These different frequency give rise to various characteristics in resolution and other effect, such as atmospheric effect, influence on vegetation. The operation on such long-wavelength band leads to all-weather and night time imaging capability. Nowadays, satellites that carried or are carrying geophysical missions include: ERS by European Space Agency, JERS by Japanese Aerospace Exploration Agency, RADARSAT by Canada.

They are elegant with amazing appearance and useful function.

1.1.1 synthesis and geometric properties
The ratio of the EM wavelength to the aperture scale give us the angular resolution, so a radar with 23cm wavelength, 10m aperture gives about 0.01 angular resolution and approximately 10 km on the ground. The resolution is improved by an advanced technique called focusing. The radar transmits and receive thousands of echoes along the track, then synthesize them together, instead of analyzing each signal individually. The synthetic aperture increase to 5km or so rather than 10m. To achieve this long aperture, it must emit signal every 5m along the orbit, consider the flying speed is about 6km/s to the ground, it means thousands pulse repetition frequency is required. This reconstruction of the radar image set up a different coordinate system (range and azimuth) which is defined by the position and velocity vector of the satellite. We can visualize this by imagining taking pictures of the ground from aircraft using a wierd camera, which give us an image similar to ordinary air photo, but in a differet unit. One of the drawbacks worthy to notice is the "layover" effect. The radar distinguish object by echo travel time, hence two objects that have the same travel time but in different location may cause artifact in the image.

1.1.2 Properties of image amplitude
The amplitude of the image relates to the roughness of the surface, EM properties of the material. If the wavelength is far bigger than the scale of the roughness, the situation is similar to seeing somebody from a mirror. If the wavelength is smaller than that, it will like scattering: the radiation energy spread out in all direction. If there is a calm water surface, no much energy will reflect backward, so we have low amplitude, while if there is rough waves swept over by the wind, or a tree standing in the water, acting as corner reflector, then we will have high amplitude. More, the longer wavelength EM waves can penetrate more vegetation, so it's prone to get to the bottom of the ground rather than reflected by the canopy. This kind of property, plus the conductivity of the material influence the amplitude image. In general, this sort of image is just like a black-white phote.

1.1.3 Property of the phase image
The data is a 2D complex array, because the radar echoes carry both the amplitude and the phase information. The phase data is a combination of all kinds of varios effect so it looks like random noise with phase uniform distributed from 0 to 2*Pi. It can become useful if we substract two phase images to remove the random part.

1.2 Principles of radar phase and interferometry
The phase image with a different position, or with a different time can be compared with proper image registration. The phase differece form a interferometric patten with fringes denoting the geophysical signal.

read the following paper if you get interested.

Reference: Radar Interferometry and its Application to Changes in the Earth's Surface
D. Massonnet, et al. Review of Geophy. 1998.