Why is the grazing angle of the acoustic pulse on the bottom so important in Multibeam backscatter, and even MORE important when generating multispectral backscatter imagery? Look at the road surface in the two images below.
The two pictures above were taken within moments of each other, looking in opposite directions. Take note of the colour of the road surface in the two photos. Why is it so different?
GRAZING ANGLE!
In these two pictures, the grazing angle of the transmitter (the sun) is constant at ~60 degrees, but the grazing angle of the receiver is 130 degrees in the left-hand image and ~40 degrees in the right-hand image. The road surface’s colour (backscatter) changes with the grazing angles.
If I wanted to create a composite image from the two, the road colours’ mismatch would confuse the new image. It is the same for multispectral backscatter.
To successfully create a coherent RGB Composite multispectral backscatter image, the grazing angle of the transmitter (and receiver) must be equal for each frequency. Specifically, you must use a multibeam sonar capable of more than one frequency (interleaved or simultaneous) as you survey the bottom.
Three passes at three different frequencies (or n passes at n different frequencies) change the grazing angles, thus changing the backscatter response, making a composite image confusing to the eye AND any algorithm trying to classify the bottom.
Here are some peer-reviewed resources about multispectral backscatter:
- Multispectral Multibeam Echo Sounder Backscatter as a Tool for Improved Seafloor Characterization, Brown et al.
- Harmonizing Multi-Source Sonar Backscatter Datasets for Seabed Mapping Using Bulk Shift Approaches, Misiuk et al.
- Multiple imputation of multibeam angular response data for high resolution full coverage seabed mapping, Misiuk & Brown
