A theoretical model that describes the power of a scattered Global Position
ing System (GPS) signal as a function of geometrical and environmental para
meters has been developed. This model is based on a bistatic radar equation
derived using the geometric optics limit of the Kirchhoff approximation, T
he waveform (i.e., the time-delayed power obtained in the delay-mapping tec
hnique) depends on a wave-slope probability density function, which in turn
depends on wind. Waveforms obtained for aircraft altitudes and velocities
indicate that altitudes within the interval 5-15 kin are the best for infer
ring mind speed. In some regimes, an analytical solution for the bistatic r
adar equation is possible. This solution allows converting trailing edges o
f waveforms into a set of straight lines, which could be convenient for min
d retrieval. A transition to satellite altitudes, together with satellite v
elocities, makes the peak power reduction and the Doppler spreading effect
a significant problem for wind retrieval based on the delay-mapping techniq
ue. At the same time, different time delays and different Doppler shifts of
the scattered GPS signal could form relatively small spatial cells on sea
surface, suggesting mapping of the wave-slope probability distribution in a
synthetic-aperture-radar (SAR) fashion. This may allow more accurate measu
rements of wind velocity and wind direction.