Conference Agenda

Over the past two decades, large-scale blooms of floating macroalgae have increasingly affected human activities and the marine ecosystem. Typical examples of these blooms include the outbreaks of alga Ulva proslifera (green tide) in the China’s Yellow sea and sargassum in the Atlantic. Satellite remote sensing has been widely and effectively used in monitoring the scale of algal blooms, including Sargassum and green macroalgae events. However, our understanding of the temporal and spatial characteristics of floating algae blooms remains to be further investigated, especially for multiscale analysis. At the field scale, algal particles or patches of varying sizes, modulated by oceanic turbulence, exhibit distinct statistical properties. In our study, we use the HY-1 series satellite data to extract the distribution vector of the macroalgal blooms (green tides) in the Yellow Sea of China, propose a novel approach that applies statistical techniques inspired by turbulence and statistical physics to characterize the geometry of macroalgal blooms, generate gridded ocean dynamic parameters with multiple satellites along track data to obtain the ocean dynamic environment information, and develop a Machine-Learning-Based Optical Flow (MLBOF) to retrieve the ocean surface velocity.

A green tide extraction method for green tide using Coastal Zone Imager (CZI) data from the HY-1 series satellites is presented. By combining spectral indices such as NDVI with adaptive threshold segmentation and cloud–sun glint removal, the algorithm enhances the spectral contrast between floating macroalgae and seawater. It effectively suppresses interferences and realizes accurate identification and area estimation of green tide, supporting satellite monitoring in the Yellow Sea.

We aim to characterize the spatial patterns and temporal evolution of macroalgal blooms in Yellow sea using metrics borrowed from statistical physics and approaches used to investigate particle clustering in various fields of science. We represent different indicators throughout the bloom, from its initiation and development to its decline. This provides a better characterization of bloom heterogeneity under the influence of various factors, including turbulence driven by currents and winds.

Based on the along track data from multiple sources of marine dynamic environmental satellites, including the HY-2 series from China, the Sentinel series from ESA, the Jason series from France, the MetOp from EUMETSAT etc, ocean gridded dynamic environment data such as Sea Surface Temperature(SST), Sea Surface Wind(SSW), Significant Wave Height(SWH) and Sea Surface Height(SSH) are generated, which will provide a valuable basis for subsequent data analyses and theoretical investigations.

We also use a Machine-Learning-Based optical flow method to extract ocean surface velocity fields from passive turbulence remote sensing data obtained from geostationary and polar orbit satellites. These satellites includes FY-2, FY-4, Sentinel-4,GOCI,Himawari-8, etc. The ocean surface velocity data will be also useful for the climate change study.