Combining interferometry soundings with angle uncertainty and signal to noise ratio into data processing for sea floor charting

In this article, we present a new development with test results on the sonar uncertainty model and the relevant processing of the interferometric bathymetry sounding data. Within a sonar sounding solution, not only an angle and time-delay pair is resolved, but quality factor, angle uncertainty and signal to noise ratio (SNR), with which the angle solution is obtained, are provided as well for the completeness of the data fields required for post processing. The goal is to generate a final sounding with registered uncertainties in three dimensions. As a result, the quality factor and SNR are applied to eliminate the low quality angle solutions due to backscatter from either multiple path or shadow area. The cleaned angle and time delay pair is then used for the estimation of the local depth, registered with the depth uncertainty derived from the angle uncertainty. The resolved depth plus its uncertainty, along with data from other sensors (IMU and SVP) and their related uncertainties, are processed by the Combined Uncertainty and Bathymetry Estimator (CUBE) [1] to generate a final mapping solution registered by a total propagated uncertainty to be checked against IHO S-44 at different order levels. From another view point, the SNR is a measurement of both the signal quality at each spatial sample and the coherence between the data streams from different bathymetric transducer receive staves. The coherence measurement is helpful to determine the valid angle apart from the nadir from which a robust sounding solution can be obtained in a full ping. It is evident that the SNR streams computed from different receiver pairs start to converge from an angle α apart from the nadir. The angle solutions within α would be considered as only fine quality due to the de-correlation introduced by different spacing among the receive staves. Conventionally, sonar manufacturers specify angles ~35 deg to start processing the data. However, with the new capability, α is determined from real-time data and is adaptively adjusted in situ based on the bottom, depth, motion, and other acoustic environments. First, a Sonar Uncertainty Model is presented. It is a Simulink (Matlab) package composed of Sonar Simulation Toolsets (Applied Physics Lab, University of Washington, Seattle), Klein Bathymetry Engine, and statistical processing. Second, in a cooperative effort with the Center for Coastal and Ocean Mapping/Joint Hydrographic Center (CCOM/JHC), sea trial data has been processed to evaluate the uncertainty model. Third, a full data set collected from Portsmouth, N.H. by a Klein HydroChart 5000 has been processed by CARIS through CUBE application. Analysis results are discussed and demonstrated.