Retrieval of Coloured Dissolved Organic Matter with Machine Learning Methods

TL;DR

This paper compares four machine learning methods for estimating coloured dissolved organic matter (CDOM) in water from remote sensing data, highlighting the effectiveness of regularized linear regression and kernel ridge regression.

Contribution

It introduces a comparative analysis of machine learning techniques for CDOM retrieval, demonstrating the efficiency of RLR and KRR over traditional polynomial regression.

Findings

RLR is the simplest and most efficient method.

KRR performs closely behind RLR.

Both outperform polynomial regression methods.

Abstract

The coloured dissolved organic matter (CDOM) concentration is the standard measure of humic substance in natural waters. CDOM measurements by remote sensing is calculated using the absorption coefficient (a) at a certain wavelength (e.g. 440nm). This paper presents a comparison of four machine learning methods for the retrieval of CDOM from remote sensing signals: regularized linear regression (RLR), random forest (RF), kernel ridge regression (KRR) and Gaussian process regression (GPR). Results are compared with the established polynomial regression algorithms. RLR is revealed as the simplest and most efficient method, followed closely by its nonlinear counterpart KRR.