Precision requirements for a spacecraft tide program

Abstract

A tidal analysis of sea-surface elevations from an orbiting altimeter will include errors due to the instrument and the orbital determinations. Furthermore, the results may be somewhat degraded by the small amount of data and their random distribution in time. These effects have been evaluated in a numerical feasibility investigation with the use of observed hourly heights at two tide stations and a hypothetical satellite orbit with a 30° inclination and a period of about 95 min. For a station at 13°N with a water elevation assumed to be obtained each time the altimeter transits a 1° square centered at the station, 86 observations are obtained in a year. For a station at 22°N, areas of 1° to 5° square were used and the number of observations in a year varied from 112 to 551. Degradation of the signal is studied by introducing sequentially increased levels of white noise to the observations. The harmonic constants for the smaller constituents deteriorate rather quickly as white noise is added, but the largest constituents appear to be reasonably determined even when random numbers with extreme absolute values as large as the amplitude are added to the data. These results indicate that previously stated minimum precision requirements for a spacecraft tide program can be relaxed by roughly an order of magnitude.