Modeling fish movement in sonar beam

Abstract

Enumerating salmon in the Yukon River drainage allows for assessment of annual harvest management guidelines and prediction of long-term salmon population trends in Alaska. Sonar is currently used to enumerate migrating salmon and determine salmon location in the river. To understand these results, a model of fish movement is required. This thesis analyzes the existing sonar data on fish movement to construct a model that predicts typical spatial and temporal distribution of fish. A model of the sonar measurement system, which includes target strength, transmission loss, transducer beam pattern, time delay, and noise is developed. This system will simulate a sonar signature for an arbitrary distribution of fish by making several simplifying assumptions. This thesis compares the simulated system sonar signature with assumed fish distribution to predict the accuracy of the sonar fish counting system.