Bottom reverberation predictions with the acoustic model Lybin - a comparison between three bottom backscattering algorithms

Comparing the three bottom backscattering algorithms we observe variations in the resulting bottom reverberation levels from 0 to more than 20 dB in some cases. Concentrating on the variations at shorter to medium ranges where detection is limited by bottom reverberation we find that levels predicted by the McKinney & Anderson and del Balzo algorithms tend to produce equivalent results in many cases whereas the Lambert algorithm tend to produce higher levels at least for the harder bottoms. Turning to the cases where real sonar data are available, we observe that the McKinney & Anderson and del Balzo algorithms provide good agreement with the data in many cases whereas the Lambert algorithm generally predicts too high levels. It is observed that the agreement between model predictions and sonar data is clearly dependent on the correctness of the model assumptions, particularly the depth profile (flat bottom in our case) and the oceanographic conditions. Based on the simulations and the comparisons with real data it appears that for the cases studied the McKinney & Anderson algorithm is no less valid for the CAPTAS and Spherion frequencies than are the two other algorithms. We have seen that the McKinney & Anderson and del Balzo algorithms generate equivalent results in many cases, particularly when comparing with measured sonar data. As far as this study goes, we have found no reason to replace the McKinney & Anderson algorithm with neither the del Balzo algorithm nor the Lambert/TUS algorithm.