A Simple Red Tide Monitoring Method using Sentinel-2 Data for Sustainable Management of Brackish Lake Koyama-ike, Japan

Abstract

We proposed and validated a method for monitoring red tides in the brackish Lake Koyama-ike, Japan, using Sentinel-2 Multispectral Instrument (MSI) data with a 10 m spatial resolution. To achieve this objective, we acquired 36 spectral reflectance/Chla data points in the field from 2012 to 2018. We obtained a high correlation of Chla (R2 = 0.83) using the proposed red tide model (RIKY = [MSI Band 5 − MSI Band 4]/[MSI Band 5 + MSI Band 4]) and field data. Based on our results, the proposed model was also validated using five Sentinel-2/Chla datasets from April to August 2017. Chla and red tide distribution characteristics estimated from Sentinel-2 data hardly appeared from April to July, and then spread rapidly throughout the lake (more than 70%) in August. Thus, Sentinel-2 data proved to be a very powerful tool in monitoring red tides in Lake Koyama-ike.

Keywords

• red tide
• chlorophyll
• Sentinel-2
• brackish lake
• atmospheric correction