Koji Suzuki

In response to recent advances in satellite ocean color remote sensing, we have developed a chlorophyll a size distribution (CSD) model using machine learning (ML) approaches for optically complex Pacific Arctic waters. Previous CSD models have used principal component analysis (PCA) to retrieve spectral features from satellite-estimated phytoplankton absorption coefficient (aph(λ)) by assuming a strong correlation between the spectral features and phytoplankton size structure determined from the exponent of the CSD (η). A weakness of such approach is that it relies on satellite retrievals of aph(λ), which can be highly uncertain due to the optical effects of water constituents other than phytoplankton. In this study, we have developed a method based on ML to use remote sensing reflectance (Rrs(λ)) for directly retrieving η, thus avoiding uncertainties due to the inversion of aph(λ) from Rrs(λ). Results show superior performance of the ML-based CSD models compared to the PCA-based model utilizing both Rrs(λ) and aph(λ) as predictors of η. For direct Rrs(λ)-based retrievals, a CSD model based on multivariable linear regression produced the best performance among all models considered. Nevertheless, models using in-situ aph(λ) yielded better accuracy, reflecting a closer optical linkage between η and aph(λ) than between η and Rrs(λ). Our choice of an Rrs(λ)-based model for satellite application is therefore practical, motivated by the limitations and uncertainty of aph(λ) inversions in optically complex waters. Another key finding is that more complex ML approaches do not always produce more effective models than standard linear regression. Indeed, multivariable linear regression outperformed other ML approaches for retrieving η directly from Rrs(λ), whereas support vector machine performed the best among diverse ML approaches in the case of aph(λ). Overall, this study found benefits in using Rrs(λ) with ML to improve the retrieval accuracy of η for Pacific Arctic waters.

The Sea of Okhotsk is one of the most biologically productive regions, where primary production during spring phytoplankton blooms after sea ice melting/retreat has the potential to contribute to the sea-to-air emission flux of atmospheric organic aerosols (OAs). To elucidate the effect of oceanic biological activity during blooms on the formation process of OAs, aerosol samples and surface seawater were collected during the bloom period of April 2021. Organic matter (OM) was the dominant component of submicrometer aerosols during both the bloom (53 ± 16 %) and bloom-decay periods (44 ± 12 %), with OM being highly water-soluble during the bloom. Stable carbon isotope ratios of aerosol organic carbon (OC) showed that 73 %–82 % of the observed aerosols were of marine origin. Relations between water-soluble OC (WSOC) and molecular tracers suggested that aerosol WSOC of marine origin was likely affected by secondary formation from precursors such as α-pinene and DMS-relevant compounds rather than primary emissions of sea spray aerosols. The amounts of water-soluble organic nitrogen (WSON) in aerosol and dissolved organic nitrogen (DON) in seawater during the bloom were larger than those during bloom-decay period, suggesting the preferential formation of N-containing water-soluble OAs of marine origin during the bloom. The increase in the amount of DON during the bloom was likely associated with the predominant diatoms, Thalassiosira spp. and Fragilariopsis spp. This study highlights the significant contribution of the secondary formation of marine biogenic OAs with increased N-containing components during the bloom after sea ice melting/retreat in the subarctic ocean.

The Sea of Okhotsk is one of the most biologically productive regions, where primary production during spring phytoplankton blooms after sea ice melting/retreat has the potential to contribute to the sea-to-air emission flux of atmospheric organic aerosols (OAs). To elucidate the effect of oceanic biological activity during blooms on the formation process of OAs, aerosol samples and surface seawater were collected during the bloom period of April 2021. Organic matter (OM) was the dominant component of submicrometer aerosols during both the bloom (53±16 %) and bloom-decay periods (44±12 %), with OM being highly water-soluble during the bloom. Stable carbon isotope ratios of aerosol organic carbon (OC) showed that 73–82 % of the observed aerosols were of marine origin. Relations between water-soluble OC (WSOC) and molecular tracers suggested that the majority of WSOC of marine origin was affected by secondary formation from precursors such as α-pinene and DMS-relevant compounds instead of primary emissions of sea spray aerosols. The amounts of water-soluble organic nitrogen (WSON) in aerosol and dissolved organic nitrogen (DON) in seawater during the bloom were larger than those during bloom-decay period, suggesting the preferential formation of N-containing water-soluble OAs of marine origin during the bloom. The increase in the amount of DON during the bloom was likely associated with the predominant diatoms, Thalassiosira spp. and Fragilariopsis spp. This study highlights the significant contribution of the secondary formation of marine biogenic OAs with increased N-containing components during the bloom after sea ice melting/retreat in the subarctic ocean.

In response to recent advances in satellite ocean color remote sensing, the present study developed a chlorophyll-a size distribution (CSD) model using machine learning (ML) approaches for optically complex Pacific Arctic waters. Previous CSD models capture the spectral features of satellite-estimated phytoplankton absorption coefficient (a_ph(λ)) through principal component analysis (PCA) and assume a strong correlation between the spectral features and phytoplankton size structure determined from the exponent of the CSD (η). A weakness of this approach is that relies on satellite retrievals of a_ph(λ), which can be highly uncertain due to the optical effects of water constituents other than phytoplankton. Therefore, we tested the utility of remote sensing reflectance (R_rs(λ)) for directly deriving η and ML methods to identify other viable algorithm formulations besides PCA. Results show superior performance of the ML-based CSD models compared to the PCA-based model utilizing both R_rs(λ) and a_ph(λ) as predictors of η. Considering the large uncertainties in the inversion of a_ph(λ) from R_rs(λ), the CSD model with R_rs(λ) based on multivariable linear regression produced the best performance among all models considered. Another key finding is that more complex ML approaches do not always produce more effective models than standard linear regression. Indeed, simple linear regression outperformed other ML approaches for retrieving η directly from R_rs(λ), whereas support vector machine performed the best among diverse ML approaches in the case of a_ph(λ). Overall, this study found benefits in using R_rs(λ) with ML to improve the retrieval accuracy of η for Pacific Arctic waters.

In response to recent advances in satellite ocean color remote sensing, the present study developed a chlorophyll-a size distribution (CSD) model using machine learning (ML) approaches for optically complex Pacific Arctic waters. Previous CSD models capture the spectral features of satellite-estimated phytoplankton absorption coefficient (a_ph(λ)) through principal component analysis (PCA) and assume a strong correlation between the spectral features and phytoplankton size structure determined from the exponent of the CSD (η). A weakness of this approach is that relies on satellite retrievals of a_ph(λ), which can be highly uncertain due to the optical effects of water constituents other than phytoplankton. Therefore, we tested the utility of remote sensing reflectance (R_rs(λ)) for directly deriving η and ML methods to identify other viable algorithm formulations besides PCA. Results show superior performance of the ML-based CSD models compared to the PCA-based model utilizing both R_rs(λ) and a_ph(λ) as predictors of η. Considering the large uncertainties in the inversion of a_ph(λ) from R_rs(λ), the CSD model with R_rs(λ) based on multivariable linear regression produced the best performance among all models considered. Another key finding is that more complex ML approaches do not always produce more effective models than standard linear regression. Indeed, simple linear regression outperformed other ML approaches for retrieving η directly from R_rs(λ), whereas support vector machine performed the best among diverse ML approaches in the case of a_ph(λ). Overall, this study found benefits in using R_rs(λ) with ML to improve the retrieval accuracy of η for Pacific Arctic waters.

In response to recent advances in satellite ocean color remote sensing, the present study developed a chlorophyll-a size distribution (CSD) model using machine learning (ML) approaches for optically complex Pacific Arctic waters. Previous CSD models capture the spectral features of satellite-estimated phytoplankton absorption coefficient (a_ph(λ)) through principal component analysis (PCA) and assume a strong correlation between the spectral features and phytoplankton size structure determined from the exponent of the CSD (η). A weakness of this approach is that relies on satellite retrievals of a_ph(λ), which can be highly uncertain due to the optical effects of water constituents other than phytoplankton. Therefore, we tested the utility of remote sensing reflectance (R_rs(λ)) for directly deriving η and ML methods to identify other viable algorithm formulations besides PCA. Results show superior performance of the ML-based CSD models compared to the PCA-based model utilizing both R_rs(λ) and a_ph(λ) as predictors of η. Considering the large uncertainties in the inversion of a_ph(λ) from R_rs(λ), the CSD model with R_rs(λ) based on multivariable linear regression produced the best performance among all models considered. Another key finding is that more complex ML approaches do not always produce more effective models than standard linear regression. Indeed, simple linear regression outperformed other ML approaches for retrieving η directly from R_rs(λ), whereas support vector machine performed the best among diverse ML approaches in the case of a_ph(λ). Overall, this study found benefits in using R_rs(λ) with ML to improve the retrieval accuracy of η for Pacific Arctic waters.

The Sea of Okhotsk is one of the most biologically productive regions, where primary production during spring phytoplankton blooms after sea ice melting/retreat has the potential to contribute to the sea-to-air emission flux of atmospheric organic aerosols (OAs). To elucidate the effect of oceanic biological activity during blooms on the formation process of OAs, aerosol samples and surface seawater were collected during the bloom period of April 2021. Organic matter (OM) was the dominant component of submicrometer aerosols during both the bloom (53±16 %) and bloom-decay periods (44±12 %), with OM being highly water-soluble during the bloom. Stable carbon isotope ratios of aerosol organic carbon (OC) showed that 73–82 % of the observed aerosols were of marine origin. Relations between water-soluble OC (WSOC) and molecular tracers suggested that the majority of WSOC of marine origin was affected by secondary formation from precursors such as α-pinene and DMS-relevant compounds instead of primary emissions of sea spray aerosols. The amounts of water-soluble organic nitrogen (WSON) in aerosol and dissolved organic nitrogen (DON) in seawater during the bloom were larger than those during bloom-decay period, suggesting the preferential formation of N-containing water-soluble OAs of marine origin during the bloom. The increase in the amount of DON during the bloom was likely associated with the predominant diatoms, Thalassiosira spp. and Fragilariopsis spp. This study highlights the significant contribution of the secondary formation of marine biogenic OAs with increased N-containing components during the bloom after sea ice melting/retreat in the subarctic ocean.

In response to recent advances in satellite ocean color remote sensing, the present study developed a chlorophyll-a size distribution (CSD) model using machine learning (ML) approaches for optically complex Pacific Arctic waters. Previous CSD models capture the spectral features of satellite-estimated phytoplankton absorption coefficient (a_ph(λ)) through principal component analysis (PCA) and assume a strong correlation between the spectral features and phytoplankton size structure determined from the exponent of the CSD (η). A weakness of this approach is that relies on satellite retrievals of a_ph(λ), which can be highly uncertain due to the optical effects of water constituents other than phytoplankton. Therefore, we tested the utility of remote sensing reflectance (R_rs(λ)) for directly deriving η and ML methods to identify other viable algorithm formulations besides PCA. Results show superior performance of the ML-based CSD models compared to the PCA-based model utilizing both R_rs(λ) and a_ph(λ) as predictors of η. Considering the large uncertainties in the inversion of a_ph(λ) from R_rs(λ), the CSD model with R_rs(λ) based on multivariable linear regression produced the best performance among all models considered. Another key finding is that more complex ML approaches do not always produce more effective models than standard linear regression. Indeed, simple linear regression outperformed other ML approaches for retrieving η directly from R_rs(λ), whereas support vector machine performed the best among diverse ML approaches in the case of a_ph(λ). Overall, this study found benefits in using R_rs(λ) with ML to improve the retrieval accuracy of η for Pacific Arctic waters.

The Sea of Okhotsk is one of the most biologically productive regions, where primary production during spring phytoplankton blooms after sea ice melting/retreat has the potential to contribute to the sea-to-air emission flux of atmospheric organic aerosols (OAs). To elucidate the effect of oceanic biological activity during blooms on the formation process of OAs, aerosol samples and surface seawater were collected during the bloom period of April 2021. Organic matter (OM) was the dominant component of submicrometer aerosols during both the bloom (53±16 %) and bloom-decay periods (44±12 %), with OM being highly water-soluble during the bloom. Stable carbon isotope ratios of aerosol organic carbon (OC) showed that 73–82 % of the observed aerosols were of marine origin. Relations between water-soluble OC (WSOC) and molecular tracers suggested that the majority of WSOC of marine origin was affected by secondary formation from precursors such as α-pinene and DMS-relevant compounds instead of primary emissions of sea spray aerosols. The amounts of water-soluble organic nitrogen (WSON) in aerosol and dissolved organic nitrogen (DON) in seawater during the bloom were larger than those during bloom-decay period, suggesting the preferential formation of N-containing water-soluble OAs of marine origin during the bloom. The increase in the amount of DON during the bloom was likely associated with the predominant diatoms, Thalassiosira spp. and Fragilariopsis spp. This study highlights the significant contribution of the secondary formation of marine biogenic OAs with increased N-containing components during the bloom after sea ice melting/retreat in the subarctic ocean.