Development of an L-Band Geophysical Model Function and Comparison with Measured Cross Section Data
Donald R. Thompson(1), Frank M. Monaldo(1), Jochen Horstmann(2) and Merete Bruun Christiansen(3)
(1) Johns Hopkins University/APL, 11100 Johns Hopkins Road, Laurel, Maryland 20723, United States
(2) GKSS Research Center, Max-Planck-Straße 1, 21502 Geesthacht, Germany
(3) Technical University of Denmark, Frederiksborgvej 399, 4000 Roskilde, Denmark
The possibility for collecting SAR imagery over the world oceans at different radar frequencies on a routine basis has existed only since the launch of the Japanese ALOS SAR (L-band; ~1.2 GHz) in January 2006 and the German TerraSAR-X (X-band; ~10 GHz) in June 2007. These two new SAR satellites complement the Canadian RADARSAT-1 SAR launched in November 1995 and the European EnviSAT ASAR system in March 2002 (both operating at C-band; ~5.3 GHz). It is quite probable that quantitative characterization of differences in SAR imagery of a particular ocean process collected at different radar frequencies can yield significant improvement in the extraction of geophysical parameters relevant to that process. In particular, such multi-frequency imagery is expected to be especially useful for the study of high wind events such as hurricanes, where signatures observed by (roughly concurrent) SAR imagery collected by the L-, C-, and X-band satellites presently in orbit could provide new insight into hurricane morphology and quantitative high-resolution estimates of such parameters as sea state, rain rate, and foam coverage, as well as wind velocity. If such multi-frequency observations are collected at multiple-polarization states - already available from three of the four SAR systems discussed above - it may be possible to further improve the accuracy of these retrieved parameters.
Work by several international research groups to address multi-frequency (and multi-polarization) SAR retrieval techniques is currently underway. At present, one of the major deficiencies in this research is the lack of a reliable L-band Geophysical Model Function (GMF) that relates the L-band normalized radar cross section (NRCS) to the surface wind vector. We report here our construction of a simple, physics-based L-band GMF based on wind-dependent models of the ocean surface roughness spectrum and our attempts to validate this function using measurements from airborne SAR imagery as well as the ALOS sensor. GMF predictions were compared against simultaneous dual-polarization airborne SAR data from the Danish EMISAR and the German E-SAR systems. We find that the Elfouhaily roughness spectrum shows generally good agreement with the V-Pol data, but is low for H-Pol. The GMFs using the Romeiser or Apel spectral models agree somewhat better at H-Pol, but are significantly higher than the measurements at V-Pol. As expected, all of the L-band GMFs (at both V- and H-Pol) are less sensitive to wind than the commonly used C-band GMFs. A major difficulty, for which we have no explanation as yet, is that the ocean surface NRCS values from ALOS wide-swath (H-Pol) imagery not only appear to be significantly higher than corresponding predictions from our GMFs using any of the three spectral models, but are also higher than measurements collected by the airborne SAR systems for similar environmental conditions and radar geometries.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,