ERS SAR Tropical
Atmospheric Phenomena Introduction
Mesoscale and submesoscale atmospheric phenomena become visible on SAR images because they are associated with variations of the wind stress at the sea surface. The wind stress depends on the wind speed at the sea surface and on the stability of the air-sea interface which is a function of the temperature difference between water and air (Keller et al., 1989). When the water is warmer than the air, then the air-sea interface is unstable.
Variations of the wind speed at the sea surface disturb the small-scale sea surface roughness and thus give rise to "imprints" on the sea surface which are visible on the radar image as variations of the SAR image intensity.
From SAR images of the sea surface also quantitative information on the wind speed can be extracted. However, this is only possible when the wind direction is known. The wind direction can often be inferred from the SAR images itself, e.g., from the direction of the wind streaks or from the direction of the wind shadows behind mountainous island, or from weather charts. With this knowledge of the wind direction the wind speed can be extracted from the NRCS values by applying a wind scatterometer model like the CMOD4 model of ESA (Stoffelen et al., 1993) or the CMOD-IFR2 model of the Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER) at Brest, France (Quilfen et al., 1993). Care has to be taken when converting ERS SAR image intensities into NRCS values because of nonlinearities in the ERS SAR analogue-to-digital converter (see Laur et al., 1997, Lehner et al, 1998). For a discussion of the accuracy of wind retrieval from ERS SAR images using the above mentioned C-band wind scatterometer models, the reader is referred to the papers of Vachon and Dobson (1996), Scoon et al. (1996), Furevik and Korsbakken (2000) and Horstmann et al. (2000).
The SAR onboard the ERS satellites, which operate at C-band and
VV polarization, are much more sensitive for the detection of
atmospheric phenomena in the marine boundary layer than the Seasat
SAR, which operated at L-band and HH polarization. This is because
the Bragg waves responsible for the radar backscattering at the sea
surface have for ERS SAR a wavelength of 7 cm and for Seasat SAR a
wavelength of 31 cm. Obviously, the 7 cm Bragg waves are much more
responsive to wind variations than the 31 cm Bragg waves.