Analytical Model of the Electromagnetic Bias Using the Physical Optics Scattering Theory

The electromagnetic bias (EM) is a critical error term in sea surface height estimation from satellite radar altimetry. At present the EM bias models used in current altimetry missions are empirically based and globally-averaged functions of the altimetermeasured signi cant wave height and wind speed alone. Recent studies have shown that a reduction in the EM bias error variance can be achieved by incorporating ancillary wave model data into the EM bias model. This motivates an improved understanding of the physical mechanisms of the EM bias, so that an optimal means for incorporating ancillary data can be developed. While the electromagnetic bias has been studied extensively, most previous studies (e.g. [1],[2]) have resorted to low-order hydrodynamic and electromagnetic models in order to express the backscattering radar cross section as a function of surface height. Recently, an alternate approach for EM bias studies has been proposed by Naenna and Johnson based on Monte Carlo simulations of altimeter pulse returns [4]. It has shown that, under the Brown model [3], the EM bias as obtained by Jackson [1] can also be expressed in terms of the normalized rst moment of the altimeter time-domain pulse return. The simulation produces a deterministic set of sea surface pro les and the corresponding altimeter pulse returns, thus allows the impact of various physical effects to be investigated by varying the method used to produce the sea surfaces simulated. The Monte Carlo results are reasonable, but hard to interpret. This is because the EM bias is so small that a large number of surface realizations are required for good convergence. As a result, an analytical model is is developed in this presentation to provide physical insight into the EM bias mechanism.