Thermal noise estimation improvement: Point Target Response deconvolution

Cédric Tourain (CNES, France)


Thomas Grelier (CNES, France); Arnaud Jouzeau (CNES, France); Céline Tison (CNES, France)

Event: 2016 Ocean Surface Topography Science Team Meeting

Session: Instrument Processing: Measurement and retracking (SAR and LRM)

Presentation type: Type Poster

The SWIM (Surface Wave Instrument Monitoring) is a wave conical scanning scatterometer in Ku-band with 6 incidence angles (from 0° to 10°) which will be embarked on the CFOSAT mission [1]. CFOSAT is a Chinese-French oceanographic mission for the joint observation of the wave and the wind vectors at the oceanic surface.
The estimation of thermal noise is a key issue for the SWIM sigma0 estimation performance and the inversion of the signal to the 2D wave spectrum.
The thermal noise is estimated on nadir and 2° beam waveforms, where the first gates are free of ground signal It is then extrapolated to the 4 to 10° beams, as there is no access to a noise floor for those beams. This implies to have a very accurate estimation at 0° or 2° as the SNR of the beams is very different and the propagation of errors can have a dramatic impact. In this context, the simple averaging of first gates is not suitable. The contribution of the Point Target Response (PTR) and of the quantification noise on the noise floor cannot be neglected (impact of PTR secondary lobs). A PTR deconvolution method has been specified and implemented at CNES. The thermal noise obtained with this method is compliant with the accuracy required for SWIM high incidence beams data processing.
In addition, as SWIM products are planned to be delivered in Near Real Time (NRT), the computation time is a major issue. The SWIM nadir processing relies on a so called ‘numerical retracking’, which implies to convolve many times a model echo with the measured PTR. In order to reduce the number of operations, an evolution of the retracking has been performed: it consists in initially deconvoluting the nadir signal from the PTR, the number of convolution operations is then reduced to one.
This paper will describe the PTR deconvolution method and its results on thermal noise estimation. It will also describe results on the SWIM nadir retracking and propose further possible evolutions.

[1] Hauser D., C. Tison, J.-M. Lefevre, J. Lambin, T.Amiot, L. Aouf, F. Collard, and P. Castillan, Measuring ocean waves from space: Objectives and characteristics of the China-France Oceanography SATellite (CFOSAT), Proceedings of the ASME 2010 29th International Conference on Ocean, Offshore and Arctic engineering, 2010


Poster show times:

RoomStart DateEnd Date
Grande Halle Thu, Nov 03 2016,11:00 Thu, Nov 03 2016,18:00
Cédric Tourain