Abstract's details

Surface Films: Is it possible to detect them using Ku/C band sigmaO relationship

Jean Tournadre (IFREMER, France)

CoAuthors

Douglas Vandemark (University New Hampshire, USA)

Event: 2019 Ocean Surface Topography Science Team Meeting

Session: Science Keynotes Session

Presentation type: Type Keynote/invited

Contribution: PDF file

Abstract:

Most altimeters operate at dual frequencies, Ku (13.5GHz) and C (5.3GHz) bands. The relationship between the two backscatters has been used to detect altimeter samples affected by rain using the differential attenuation of the signal by liquid water. The KU-C band relationship contains much more information than the rain attenuation. The analysis of Jason1 and 2 and Envisat data during the Deep Water Horizon oil spill and the comparison with oil spill thickness and extent data showed that the presence of oil spills on the sea surface can distort the altimeter waveforms and cause 'blooms' in the radar backscatter cross-section signal, but compresses the surface capillary wave and reduce the sigma0 in SAR imagery.
The comparison of high resolution surface sigma0 obtained by waveform inversion method and ERMA oil cover fields showed that locally the Jason-1/2 Ku band sigma0 increased up to 10 dB in low wind speed. At low wind speed (<3 m.s^{-1}), the mean sigma0 in Ku and C bands increased by 1.0 to 3.5 dB for thick oil and 0.9 to 2.9 dB for thin oil while the waveforms are strongly distorted. At medium winds (up to 6 m.s^{-1}) the mean sigma0 bloom and waveform distortion in both Ku and C bands weakened for both thick and thin oil. For larger winds (> 6 m.s^{-1}) only does the Ku band sigma0 slightly increase by 0.2-0.5 dB for thick oil.
A further analysis reveals that surface films have a differential impact at Ku and C band depending on the existing surface roughness (due to wind) and film thickness. This differential impact reflects in the Ku-C band relationship by modifying the departure from the "normal" Ku-C band relation. The presence of a film tend to increase the Ku band backscatter more than the C band one, i.e. the Ku band roughness (2cm) is more suppressed than the C (6cm) band one. The difference (dS) between the measured Ku band sigma0 and the Ku sigma0 estimated from the C band using the mean KU/C band relationship becomes significantly larger than that without film. In general it can exceed 1.8 rms of the relationship. At low wind speed, the Ku/C band relation, i.e. dS above 1.8 rms, could then allow to discriminate between normal low wind conditions and the presence of surface film either biogenic or anthropogenic. Such occurrences are called high dS in the following.
We use the whole Jason2 archive to test the possibility of the detection of surface films on a global scale using Ku/C band relationship. The departure dS at 1Hz has thus been analyzed over the global ocean. Six special regions corresponding to different environmental conditions and number of high dS events have been selected for which the 20 Hz data are analyzed. The distributions of the events are estimated and analyzed in terms of wind speed (from ECMWF), off-nadir angles (to estimate the waveform distortion), Chlorophyll (using coincident daily CHL estimates), SWH ..
In some regions such as the Argentinian basin, there is a clear indication that high dS are associated to high CHL content while other regions like the China Sea high dS can be associated to very low CHL content. As these zone are close to main shipping lanes they might result from anthropogenic film.
The global and regional analysis of the distribution of the relationship departure dS as a function of CHL reveals a clear impact of the presence of biomass on the the dS pdf confirming that the KU/C band can be used to detect surface films at least at low winds.
The global spatio-temporal distribution of high dS on a global will be presented showing the regions of high biomass productivity as well as the impact of human activities in some specific regions.
 

Keynote/invited presentation show times:

Room Start Date End Date
The Forum Mon, Oct 21 2019,14:50 Mon, Oct 21 2019,15:15
Jean Tournadre
IFREMER
France
jean.tournadre@ifremer.fr