The useful information is a wave’s phase
A complex SAR image contains both echo amplitude and phase. Amplitude describes return strength; phase records position within a repeating wave cycle. InSAR aligns corresponding pixels from two acquisitions and compares their phases. If a scattering surface changes its distance from the satellite, the round-trip radar path changes and can shift that phase difference. Orbit geometry and topographic relief also contribute. The initial interferogram therefore contains several effects, and a digital elevation model and orbital information are used to remove predicted contributions before interpreting the remaining phase as deformation.
Sources: insar-gsi
One color cycle measures half a wavelength
A wrapped interferogram repeats after a full 360-degree phase cycle. Because radar makes a round trip, one full cycle corresponds to a one-way range change of half the radar wavelength. GSI’s explanation uses approximately 24-centimeter ALOS-2 radar waves, giving roughly 12 centimeters per cycle. That is an instrument-specific example, not a universal fringe spacing. A point that returns to the same color could be separated by one or several cycles. Phase unwrapping attempts to recover those missing whole-cycle differences across the image.
Sources: insar-gsi
Toward the satellite is not necessarily upward
The displacement obtained from one viewing geometry is along the radar line of sight. Both horizontal and vertical movement can shorten that distance. A single line-of-sight result does not determine their separate contributions. The Alaska Satellite Facility explains that its vertical-displacement conversion assumes all movement is vertical, an assumption that must be justified for the process being studied. A colored region moving toward the sensor may consequently include sideways motion. The look direction and sign convention are part of the measurement, not optional decorations on the map.
Sources: insar-asf
Unwrapped displacement still needs a reference
Unwrapping restores relative cycle counts but does not supply an independent absolute displacement everywhere. In the ASF product, values are tied to a selected reference pixel whose phase is set to zero. If that pixel moves, the rest of the image is expressed relative to its movement. Coherent areas separated by unreliable phase can also be difficult to connect correctly. A smoothly colored displacement map therefore needs its reference location and acquisition dates to describe a meaningful change, rather than merely a continuous-looking set of numbers.
Sources: insar-asf
Changing vegetation can erase a phase comparison
The method needs similar scattering behavior in both observations. Leaves and branches move and grow, changing the collection of objects that returned the radar signal within a pixel. GSI explains that this can prevent useful interference, particularly for shorter-wavelength radar over forest. Longer wavelengths can retain useful coherence under some vegetation conditions, but major vegetation changes still affect the result. Losing a stable phase relationship is different from measuring zero displacement: the pair may no longer reliably compare that patch of ground.
Sources: insar-errors
Water vapor changes the radar path too
The atmosphere can delay radar signals differently on the two acquisition dates. GSI distinguishes water-vapor variations with height, which can create patterns correlated with terrain, from horizontal variations that produce other spatial patterns. A mountain-shaped phase feature need not be uplift. Weather-model corrections can reduce these effects, and repeated interferograms help test whether a feature behaves consistently across acquisitions. The physical distinction is between a changed ground-to-satellite distance and a changed propagation delay; both can alter phase, so interpreting a deformation pattern requires separating their contributions.
Sources: insar-errors