Monitoring and forecasting of green tide biomass based on multi-source remote sensing

In this study, we sought to develop a method for the accurate monitoring and prediction of green tide biomass in the Yellow Sea. To this end, we obtained multi-source remote sensing data (including GF-1 and HJ-2 A/B, etc.) and green tide salvage data from 2023 and constructed a segmented exponential model based on the mean normalized difference vegetation index and green tide biomass per unit area to estimate the green tide biomass in 2025. The unit-area model validation presents R² = 0.8590, RMSE = 0.0888 kg/m². The results indicated that green tides initially appeared off the coast of Yancheng, Jiangsu Province, on May 7th, reached a peak of 1499.8 kt off the coast of Lianyungang on June 22nd, and dissipated off the Rizhao coastline, Shandong Province, after August 7th. In addition, we established that the distribution of green tide biomass in the Yellow Sea was characterized by a clustered pattern, which declined from the center to the periphery. Moreover, we developed a method for predicting cumulative green tide biomass. Using a combination of the Logistic and Gompertz growth curve models, we monitored the cumulative green tide biomass in 2025, with the Logistic model showing stronger overall explanatory power for the data (R² > 0.99). The segmented exponential estimation model and biomass prediction method proposed in this study will provide key technical support and a scientific decision-making basis for the early warning, emergency response, and refined management of the marine ecological environment regarding green tide disasters in the Yellow Sea.