Key points are not available for this paper at this time.
Bathymetry has been traditionally charted via shipboard echo sounding. Alhough able to generate accurate depth measurements at points or along transects, this method is constrained by its high operating cost, inefficiency, and inapplicability to shallow waters. By comparison, remote sensing methods offer more flexible, efficient and cost-effective means of mapping bathymetry over broad areas. Remote sensing of bathymetry falls into two broad categories: non-imaging and imaging methods. The non-imaging method (as typified by LiDAR) is able to produce accurate bathymetric information over clear waters at a depth up to 70 m. However, this method is limited by the coarse bathymetric sampling interval and high cost. The imaging method can be implemented either analytically or empirically, or by a combination of both. Analytical or semi-analytical implementation is based on the manner of light transmission in water. It requires inputs of a number of parameters related to the properties of the atmosphere, water column, and bottom material. Thus, it is rather complex and difficult to use. By comparison, empirical implementation is much simpler and requires the input of fewer parameters. Both implementations can produce fine-detailed bathymetric maps over extensive turbid coastal and inland lake waters quickly, even though concurrent depth samples are essential. The detectable depth is usually limited to 20 m. The accuracy of the retrieved bathymetry varies with water depth, with the accuracy substantially lower at a depth beyond 12 m. Other influential factors include water turbidity and bottom materials, as well as image properties.
Jay Gao (Sun,) studied this question.