A Clustering Centered on Spatio Temporal HDBSCAN and its implementation to Classify Potential Fishing Areas using Oceanographic Parameters

Abstract

Potential Fishing Zones (PFZ) advisories furnish with the understanding of the places where possible fish shoals may present. The proposed work focuses on predicting the Potential Fishing Zone(PFZ) in the landing location Nagapattinam, Arcottu Thurai, Pazhayar, and Sethur. The daily fish catch data were collected from  National Marine Fisheries Data Center, Cochin and the Oceanographic satellite data such as Sea Surface Temperature(SST), Sea Surface Chlorophyll(SSC) and Sea Surface Height(SSH) were collected from National Oceanic and Atmospheric Administration's (NOAA's). Spatio-temporal HDBSCAN clustering method used to identify clusters of zones. The average of fish catch in each cluster was compared with the average of total fish catch data and labeling the data for training were used as training data.KNN (K-Nearest Neighbour)Classification is used and the outcome of this technique results the information to predict the Potential Fishing Zone. The experimental results have shown that the landing location between the range of 10 -13 "N to 79-81 "E   are influenced by different oceanographic features and fish data yields a good accuracy of PFZ prediction that reaches 91.1458%.