Matlab code for Machine Learning-based Chlorophyll-a Size Distribution (CSD) model in the Pacific Arctic

This repository contains the MATLAB implementation of the Machine Learning (ML)-based Chlorophyll-a Size Distribution (CSD) model. The model is developed to estimate the CSD slope (η)—a key indicator of phytoplankton size structure—from satellite ocean color imagery in the optically complex waters of the Pacific Arctic.

The code includes implementations for:

• A multivariable linear regression (ML) model using remote sensing reflectance (Rrs(λ)).

• A support vector machine learning model using phytoplankton absorption coefficient (aph(λ)).

This software supports the results presented in the manuscript: Machine learning for estimating phytoplankton size structure from satellite ocean color imagery in optically complex Pacific Arctic waters (Waga et al., accepted).