[A Three Band Chlorophyll-a Concentration Estimation Model Based on GOCI Imagery]

Area of Science:

• Oceanography
• Remote Sensing
• Environmental Science

Background:

• Chlorophyll-a concentration is a key indicator of water quality and phytoplankton biomass.
• Traditional remote sensing models for chlorophyll-a estimation have limitations in accuracy and stability.

Purpose of the Study:

• To develop and validate a novel three-band model for estimating chlorophyll-a concentration using Geostationary Ocean Color Imager (GOCI) data.
• To compare the performance of the proposed GOCI-based model against existing MERIS-based and GOCI band ratio models.

Main Methods:

• Development of a GOCI-based three-band model using in-situ chlorophyll-a measurements and simulated GOCI spectra.
• Validation of the model using independent GOCI imagery and in-situ datasets.
• Comparative analysis with MERIS-based three-band model and GOCI band ratio model.

Main Results:

• The GOCI-based three-band model demonstrated strong performance, with R2 values of 0.809 in the modeling dataset and similar accuracy in the validation dataset.
• The proposed model showed significantly better accuracy and stability compared to the widely used GOCI band ratio model (R2 = 0.450).
• GOCI image validation revealed that the band ratio model underestimates chlorophyll-a and lacks spatial detail in Taihu Lake.

Conclusions:

• The GOCI-based three-band model offers superior accuracy, stability, and application potential for chlorophyll-a estimation compared to the GOCI band ratio algorithm.
• This model provides a reliable tool for monitoring water quality and phytoplankton dynamics using GOCI data.
• The findings highlight the advancements in remote sensing techniques for aquatic ecosystem assessment.