Potential Fishing Zone Characterization in the Indian Ocean by Machine Learning Approach

Abstract

Potential fishing zone (PFZ) advisory is an essential guideline for the fishing community of India. Several studies have shown that high sea surface temperature (SST) gradient and high chlorophyll concentration in the ocean are the prospective areas for pelagic fish catch. ESSO-INCOIS provides advisories on PFZ on a daily basis using remotely sensed SST and chlorophyll-a data. The limitation of this advisory is that it does not give any information about the probable quantity of the fish. In this study, a hybrid decision tree model is developed for characterizing PFZ in the Indian Ocean. If SST gradient, persistence of SST gradient and chlorophyll concentration of any PFZ are given as the input variables, this model can classify the corresponding PFZ in terms of low, medium or high category of fish catch. According to this study, low SST gradient persistence and high SST gradient indicate possibility of high fish catch.