Sounding Data Thinning Tehniques: A Knowledge-based Approach To Modeling

Abstract

The density of sounding data being collected Over coastal waterways has increased dramatically in recent years. The cost to manually model the acean Boor has also increased accordingly in both time and resources. This paper presents the results of a prototyping effort designed to automate the basic sounding selection portion of the modeling process. The sounding selection technique utilizes intelligent data structures and rule-based selection criteria and is based on a Triangulated Irregular Network (TIN) to represent the ocean floor. The rules define significant soundings (e.g. shoals) and the degree of vertical dehnition to be retained in the resultant digital terrain model of the ocean floor. Rules are also provided to till in areas of the model not previously populated. The results of applying the modeling algorithm to data from a coastal area and a comparison of the model to the published chart for the area is described.