To what extent do atmospheric refractivity variations affect the interferograms,
is it possible to determine this influence using additional measurements, and how
should it be modeled stochastically?
The effect of refractivity variations in the atmosphere is a known problem to space geodetic observations such as SLR, VLBI, radar altimetry, and GPS. Nevertheless,
its influence on radar interferometry has been ignored for several years, largely due
to the assumptions of the all-weather capability of radar, which has been one of the
main arguments for using radar instead of optical remote sensing from the beginning. The problem is now recognized as one of the main limitations of repeat-pass
radar interferometry. The specific characteristics of different atmospheric situations
can be analyzed to identify which constituents form the main bottleneck. Both a
deterministic and a stochastic approach can be chosen to address this problem. In
the deterministic approach, the anomalous signal is derived from additional meteorological or radio-geodetic measurements and subtracted from the interferometric
products. In the stochastic approach, the unknown signal is modeled using a parameterization of its magnitudes and its spatial or temporal behavior.
