An empirical determination of the effects of sea state bias on SEASAT altimetry

Abstract

A simple linear empirical model has been developed to allow the user of SEASAT altimetry data to correct the altitude measurements for sea state bias effects. This bias is the result of two separate effects: one relating to instrumental effects and the other relating to a bias in the electromagnetic mean sea level relative to the geometric mean sea level. The electromagnetic bias is caused by the fact that the troughs of the ocean waves reflect the altimeter signal more strongly than the crests, thus causing the altimeter‐determined mean sea level to shift toward the wave troughs. An additional, but independent, instrument‐related bias is introduced because height corrections applied in the ground processor are not sufficient to compensate for simplifying assumptions made for the processor aboard SEASAT. This altitude bias also is predominately wave height dependent and directed toward the wave trough. The sum of these two effects is referred to as sea state bias. After applying appropriate corrections to the altimetry data, an empirical model for the sea state bias is obtained by differencing significant wave height and height measurements from coincident ground tracks. Ignoring small temporal variations in surface topography, the height differences on a given track should be zero; consequently, any residual height difference is attributed to sea state bias. The height differences are minimized by solving for the coefficient of a linear relationship between height differences and wave height differences that minimize the height differences. It is concluded that, for the altimeter data released by the Jet Propulsion Laboratory, 7% of the value of significant wave height should be subtracted from the SEASAT height measurements to correct for sea state bias. This correction reduced biases in the SEASAT data by more than 50% in the 36 cases examined here.