14:00 〜 14:15
[ACG41-20] Estimation of the horizontal distribution of sea-surface natural slicks based on multispectral satellite imaging
キーワード：海表面、スリック、海面マイクロレイヤー、界面活性物質、Sentinel-2
The ocean surface is covered by a thin boundary layer of less than 1 mm thick, called the sea surface micro layer (SML), which controls the vital exchange of heat, momentum, gases, and various substances between the atmosphere and the ocean. SML is characterized by the accumulation of biogenic surfactants at extremely high concentrations compared to underlying waters in natural conditions, and for this reason, recently, its impacts on the global climate and biogeochemical cycles through the suppression of air-sea exchange processes have been brought to attention. During calm conditions, the horizontal distribution of SML is perceptible visually as a smooth, patchy membrane structure called a slick, due to the suppressive effect of SML on the development of ripples and wind waves. However, different from slicks caused by artificial sources such as oil slicks, the actual horizontal distribution of natural slicks remains unknown, especially in the open ocean.
As one of the representative studies on the impact of sea-surface slicks on the global environment, the rate of CO2 absorption by the global ocean is estimated to be 20-50% lower than the actual rate due to the surfactant action of slicks. However, the uncertainty raises concerns about the reliability of this estimation in the current situation where the actual horizontal distribution of slicks is still unknown. Development of a method for accurately estimating the horizontal distribution of natural slicks is an urgent issue, and recently, Nichol et al. (2023) tried to estimate the slick distribution through high-resolution satellite imaging, over a wide area never covered in previous studies. Although they revealed the horizontal distribution of slicks in the English Channel using the multispectral imagery of the Sentinel-2 satellites, it was restricted to a single snapshot, and the limitation of the method and the temporal variability of the slick distribution are unknown.
Therefore, we aimed to estimate the horizontal distribution of natural slicks in open oceans, including its temporal variation. We applied the method of Nichol et al. (2023) to the surrounding seas of Sanriku Coast, Japan, which is characterized by a series of small inlets typical of a rias coast, and the topography creates a diverse marine environment. The target area also has accumulated knowledge about the physical, meteorological and biogeochemical aspects of the ocean, based on a large amount of field monitoring.
The horizontal distribution of slicks was identified by calculating the Surfactant Index (SI) from the sea surface reflectance of the respective wavelengths for each pixel of the optical image, applying the algorithm of Nichol et al. (2023) that determines a pixel as slick if the SI value is lower than a certain threshold value. Analysis over a two-year period from January 2023 to December 2024 revealed that the existing SI-based identification method can only be applied in the summer, when the solar altitude is sufficiently high (~50 ° ). Based on this knowledge, we estimated the frequency of slick occurrence from June to September in the period from 2020 to 2024, and found that the occurrence frequency differs by several tens of percent on a horizontal scale of a few kilometers. The factors that characterize this horizontal distribution include the possibility that the microtopography unique to the rias coast may cause differences in the development of local winds and the initial wind waves on a horizontal scale of a few kilometers. This study is the first to show the limitations of the existing SI-based method for estimating the horizontal distribution of natural slicks, and it is the first to identify the horizontal distribution of the occurrence frequency of slicks within a season over a wide area that extends from the coastal to offshore areas. The future challenge is to expand the availability of this method and to clarify the factors that form the horizontal distribution of slicks.
As one of the representative studies on the impact of sea-surface slicks on the global environment, the rate of CO2 absorption by the global ocean is estimated to be 20-50% lower than the actual rate due to the surfactant action of slicks. However, the uncertainty raises concerns about the reliability of this estimation in the current situation where the actual horizontal distribution of slicks is still unknown. Development of a method for accurately estimating the horizontal distribution of natural slicks is an urgent issue, and recently, Nichol et al. (2023) tried to estimate the slick distribution through high-resolution satellite imaging, over a wide area never covered in previous studies. Although they revealed the horizontal distribution of slicks in the English Channel using the multispectral imagery of the Sentinel-2 satellites, it was restricted to a single snapshot, and the limitation of the method and the temporal variability of the slick distribution are unknown.
Therefore, we aimed to estimate the horizontal distribution of natural slicks in open oceans, including its temporal variation. We applied the method of Nichol et al. (2023) to the surrounding seas of Sanriku Coast, Japan, which is characterized by a series of small inlets typical of a rias coast, and the topography creates a diverse marine environment. The target area also has accumulated knowledge about the physical, meteorological and biogeochemical aspects of the ocean, based on a large amount of field monitoring.
The horizontal distribution of slicks was identified by calculating the Surfactant Index (SI) from the sea surface reflectance of the respective wavelengths for each pixel of the optical image, applying the algorithm of Nichol et al. (2023) that determines a pixel as slick if the SI value is lower than a certain threshold value. Analysis over a two-year period from January 2023 to December 2024 revealed that the existing SI-based identification method can only be applied in the summer, when the solar altitude is sufficiently high (~50 ° ). Based on this knowledge, we estimated the frequency of slick occurrence from June to September in the period from 2020 to 2024, and found that the occurrence frequency differs by several tens of percent on a horizontal scale of a few kilometers. The factors that characterize this horizontal distribution include the possibility that the microtopography unique to the rias coast may cause differences in the development of local winds and the initial wind waves on a horizontal scale of a few kilometers. This study is the first to show the limitations of the existing SI-based method for estimating the horizontal distribution of natural slicks, and it is the first to identify the horizontal distribution of the occurrence frequency of slicks within a season over a wide area that extends from the coastal to offshore areas. The future challenge is to expand the availability of this method and to clarify the factors that form the horizontal distribution of slicks.