Accurate deep-learning estimation of chlorophyll-a concentration from the spectral particulate beam-attenuation coefficient

Opt Express. 2020 Aug 3;28(16):24214-24228. doi: 10.1364/OE.397863.

Abstract

Different techniques exist for determining chlorophyll-a concentration as a proxy of phytoplankton abundance. In this study, a novel method based on the spectral particulate beam-attenuation coefficient (cp) was developed to estimate chlorophyll-a concentrations in oceanic waters. A multi-layer perceptron deep neural network was trained to exploit the spectral features present in cp around the chlorophyll-a absorption peak in the red spectral region. Results show that the model was successful at accurately retrieving chlorophyll-a concentrations using cp in three red spectral bands, irrespective of time or location and over a wide range of chlorophyll-a concentrations.

MeSH terms

  • Bias
  • Chlorophyll A / analysis*
  • Databases as Topic
  • Deep Learning*
  • Neural Networks, Computer
  • Reproducibility of Results
  • Spectrum Analysis*
  • Time Factors

Substances

  • Chlorophyll A