ELECTROMAGNETIC BIAS IN SEA-SURFACE RANGE MEASUREMENTS AT FREQUENCIESOF THE TOPEX POSEIDON SATELLITE

Citation
Lg. Hevizi et al., ELECTROMAGNETIC BIAS IN SEA-SURFACE RANGE MEASUREMENTS AT FREQUENCIESOF THE TOPEX POSEIDON SATELLITE, IEEE transactions on geoscience and remote sensing, 31(2), 1993, pp. 376-388
Citations number
17
Categorie Soggetti
Engineering, Eletrical & Electronic","Geosciences, Interdisciplinary
ISSN journal
01962892
Volume
31
Issue
2
Year of publication
1993
Pages
376 - 388
Database
ISI
SICI code
0196-2892(1993)31:2<376:EBISRM>2.0.ZU;2-W
Abstract
Range measurements made by satellite radar altimeters experience an el ectromagnetic (EM) bias toward the troughs of ocean waves. This EM bia s was measured at the 13.6 GHz and 5.3 GHz operating frequencies of th e NASA altimeter on the TOPEX/POSEIDON satellite in a series of 11 air craft flights from January 17, 1991, through March 4, 1991, during the Surface WAve Dynamics Experiment (SWADE). The data are consistent wit h an earlier set of airborne measurements and indicate that EM bias is slightly higher at 5.3 GHz than at 13.6 GHz, and that the magnitudes of both biases increase with increasing wind speed, as does their diff erence. With some exceptions, EM bias shows little variation over a me soscale region on a given day or within 1 or 2 h, but it can change si gnificantly over a 6-h period. An examination of recent tower, airborn e, and satellite measurements shows a consistency in the characteristi cs of the wind speed dependence but suggests that there may be a heigh t dependence in the determinations, with the bias decreasing with incr easing altitude. The airborne measurements appear to be the most reaso nable basis for correcting the NASA altimeter range data from the TOPE X/POSEIDON satellite. In addition, a preliminary analysis of a very li mited amount of the data acquired at 20.3 m s-1 in Southern Ocean Wave s Experiment (SOWEX) has given confidence that the quadratic models fo r the prelaunch EM bias corrections are more appropriate for wind spee d dependence than are the linear models initially considered.