AS Multivariate Risk Analysis of Echotoxic Chemicals of Ballast Water Chemicals Based on PCA and DSS Using ECOTOX GISIS Data

We examine the ecotoxic risk classification of ballast water treatment chemicals by integrating global datasets—ECOTOX from the U.S. EPA and GISIS from the IMO—using a multivariate approach. The study aims to provide a practical decision-support tool to assist in ecological compliance, early warning, and regulatory prioritization. Grounded in ecological risk theory and precautionary policy frameworks, our work bridges empirical toxicological thresholds (e.g., LC50, NOEC, BCF) with practical decision-making systems for maritime chemical regulation. We apply Principal Component Analysis (PCA) to a curated dataset of 37 ballast water substances across three key dimensions: acute toxicity (LC50), chronic toxicity (NOEC), and bioaccumulation potential (BCF). A scoring matrix is then developed to construct a chemical risk heatmap supported by decision support system (DSS) logic. Notably, we observe that 43.24% of chemicals fall into the high-risk category, while only 8.11% qualify as low-risk. PCA explains 73.63% of the data variance with just two components, supporting dimensionality reduction without significant loss of ecological insight. A closer look at substances like Dibromoacetic acid and Dichloroacetonitrile reveals alarming risk profiles with high BCF and low NOEC values—chemicals that current regulations may be overlooking. Surprisingly, some widely used oxidizing agents demonstrate latent chronic toxicity previously underestimated. This finding may reflect policy inertia in updating risk evaluation metrics post-BWM Convention. Our findings contribute to both theoretical refinement in multivariate ecotoxicology modeling and policy design by suggesting a data-driven classification framework for Port State Control (PSC) and IMO risk flags. The visual heatmap tool further enables early-warning applications in smart port ecosystems. This study introduces a novel integration of ECOTOX–GISIS datasets using PCA for ecological risk visualization. Unlike prior works relying solely on laboratory thresholds or qualitative review, our approach enables real-time chemical risk ranking and supports cross-jurisdictional ballast water compliance. The methodological blend of theory and application provides a replicable model for other maritime pollutants in ASEAN and beyond.