An Experimental Investigation Of The Importance Of The Rar Modulation Transfer Function In Sar Ocean Wave Image Spectra

Abstract

This work has been carried out as a part of the NORCSEX’88 ERS-1 pre launbh experiment. The main objective of the wave part of experiment was to investigate the Ocean wave imaging capabilities of the C-band ERS- 1 SAR. The experiment took place on Haltenbanken off the coast of Norway during spring 1988, and provided SAR data obtained under extreme conditions with regard to both waveheight and multimode sea states. On March 11. seven SAR flights at two different altitudes were performed. The flights were in ai starpattern over a directional metocean buoy. The buoy showed a mmodal wave spectrum with a significant waveheight of 4-5 meters. On March 20. two passes differing approximately 1800 in flight direction were performed. The passes crossed an atmosphwic front. Across the front the wind changed from 6 to 12 m/s and shifted nearly 180 degrees in direction. This data sets give an unique opportunity to study SAR imaging of complex seastates under different imaging geometries and atmospheric conditions. SAR image spectra from all seven passes on March 11. and both passes on March 20. were computed for different incidence angles and ranges. Using the wavescan buoy spectra as insitu references a simulation study was performed to compare the result of the SAR imaging with the linear imaging theory. The buoy spectrum was first modified with scanning distortion and then multiplied with the smearingfilter and the linear modulation transfer function. The width of the filter was computed from the SAR image spectra using the ratio technique. Different models for the RAR (tilt- and hydrodynamic-) modulation transfer function were used, and the hydrodynamic modulation transfer function was varied using wind dependent parameters. The amplitude and phase of the RAR modulation transfer function are found to be important when explaining the spectral peak energy density in the SAR image spectra. It is observed that using a wind dependent hydrodynamic modulation transfer function in general gives the closest agreement between the simulated spectra and the observed SAR image spectra. It is also observed that the splitting in the SAR image spectra is changing dramatically when crossing the atmospheric front. It is concluded that this is due to wind input changing the fase and amplitude of the hydrodynamic modulation transfer function.