Related Experiment Video
Updated: Sep 5, 2025
10:28
Investigating the Relationship between Sea Surface Chlorophyll and Major Features of the South China Sea with Satellite Information
Published on: June 13, 2020
5.9K
Modeling ocean surface chlorophyll-a concentration from ocean color remote sensing reflectance in global waters using
Srinivas Kolluru1, Surya Prakash Tiwari2
1Harbor Branch Oceanographic Institute, Florida Atlantic University, FL 34946, USA; Centre of Studies in Resources Engineering, Indian Institute of Technology Bombay, Bombay 400076, India.
The Science of the Total Environment
|July 10, 2022
Summary
This study introduces a new machine learning method using a neural network to accurately estimate Chlorophyll-a (Chl-a) from ocean color data. This approach improves marine health assessments in various waters.
Area of Science:
- Oceanography
- Remote Sensing
- Marine Ecology
Background:
- Chlorophyll-a (Chl-a) variations are key indicators of marine ecosystem health.
- Traditional models struggle with complex Chl-a estimations in diverse waters.
- Machine learning offers enhanced accuracy for deriving oceanographic parameters.
Purpose of the Study:
- To develop a novel Multi-layer Perceptron Neural Network (MLPNN) approach for Chlorophyll-a (Chl-a) retrieval.
- To utilize four ocean color bands for improved Chl-a estimation accuracy.
- To assess the model's performance across various datasets and conditions.
Main Methods:
- Employed a Multi-layer Perceptron Neural Network (MLPNN) with Resilient backpropagation.
- Trained the NN on NASA's bio-optical Marine Algorithm Dataset (NOMAD).
- Validated the model using SeaWiFS, MODIS Aqua, and simulated Red Sea datasets.
Main Results:
- The MLPNN model demonstrated promising accuracy in estimating Chlorophyll-a (Chl-a).
- Hyperparameter tuning optimized the NN configuration for Chl-a retrieval.
- The developed algorithm outperformed traditional blue-green band ratio methods.
Conclusions:
- The study presents a robust approach for improved Chlorophyll-a (Chl-a) retrieval in global waters.
- The algorithm can accurately map ocean color features, phytoplankton blooms, and physical processes.
- Findings advance ocean color remote sensing and understanding of marine biogeochemical cycles.
