it
83.
jf
In
fo
4,
SIMULATION OF SAR OCEAN WAVE IMAGERY FROM SPACEBORNE PLATFORMS
D.R. Lyzenga
Radar Science Laboratory
Environmental Research Institute of Michigan
Ann Arbor, MI 48107
ABSTRACT
A numerical model for predicting the synthetic aperture radar (SAR)
image of a moving ocean surface is described, and results are presented to
illustrate the effects of surface motions on the SAR imaging process. The
relationship between the SAR image spectrum and the wave height and wave
slope spectrum is examined using this model, and the possibility of invert-
ing this relationship is discussed.
1. INTRODUCTION
The ability of synthetic aperture radar systems to image ocean waves
is dependent on system parameters as well as environmental conditions.
This dependence has been extensively investigated over the past decade,
and although there are still areas of disagreement regarding the theoreti-
cal basis for this imaging, models have been developed which appear to
explain most of the trends observable in the experimental data base.
In the present paper, a conceptually simple model for the SAR imaging
process is discussed and the numerical implementation of this model is
described. Predictions of the SAR image spectra for various environmental
conditions and SAR system parameters are presented, some comparisons with
actual SAR data are discussed, and the possibility of inverting the process
to obtain wave height or slope spectra is evaluated on the basis of this
model.
2. MODEL DESCRIPTION
The microwave scattering properties of the ocean surface, including
the phase changes due to surface motions, may be described in terms of a
complex reflectivity r(x, t). The signal received by a synthetic aperture
radar within a given range cell may then be written as
-2jkR(x". t)
s(t) - frs, t)a, (x' - Vt) e dx' (1)
where aj(x' - Vt) represents the antenna gain pattern, k is the electro-
magnetic wavenumber, and R(x', t) is the instantaneous range distance from
the scattering surface to the SAR platform, which is located at x' = Vt.
Assuming that the antenna pattern limits the illumination to an along-track
distance interval which is much smaller than the range distance, this range
distance may be approximated by
: 2
R(x!, 8) = gt UE (2)
427