Spatiotemporal Topic Modeling Reveals Storm‐Driven Advection and Stirring Control Plankton Community Variability in an Open Ocean Eddy. San Soucie, J. E., Girdhar, Y., Johnson, L., Peacock, E. E., Shalapyonok, A., & Sosik, H. M. Journal of Geophysical Research: Oceans, 129(11):e2024JC020907, November, 2024. ![Spatiotemporal Topic Modeling Reveals Storm‐Driven Advection and Stirring Control Plankton Community Variability in an Open Ocean Eddy [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex 1 download Abstract Phytoplankton communities in the open ocean are high‐dimensional, sparse, and spatiotemporally heterogeneous. The advent of automated imaging systems has enabled high‐resolution observation of these communities, but the amounts of data and their statistical properties make analysis with traditional approaches challenging. Spatiotemporal topic models offer an unsupervised and interpretable approach to dimensionality reduction of sparse, high‐dimensional categorical data. Here we use topic modeling to analyze neural‐network‐classified phytoplankton imagery taken in and around a retentive eddy during the 2021 North Atlantic EXport Processes in the Ocean from Remote Sensing (EXPORTS) field campaign. We investigate the role physical‐biological interactions play in altering plankton community composition within the eddy. Analysis of a water mass mixing framework suggests that storm‐driven surface advection and stirring were major drivers of the progression of the eddy plankton community away from a diatom bloom over the course of the cruise. , Plain Language Summary Plankton communities in the ocean can have many different species, with large differences in their abundance and patchy distributions in space. Automated imaging systems allow for high‐resolution observation of these plankton communities, but many traditional statistical techniques fail to capture their full complexity. Spatiotemporal topic models, a kind of statistical model designed to work directly with categorical data, can effectively represent this kind of information. In this work, we use topic models to analyze plankton images taken near an eddy in the spring of 2021 and classified into 50 different kinds of plankton with an automated algorithm. We investigate how interactions between ocean physics and biology can change the plankton community inside the eddy. Analysis suggests that storms in the area moved surface water carrying a different plankton community into the eddy. , Key Points Topic models provide a robust alternative to traditional statistical techniques for analysis of sparse, high‐dimensional categorical data We perform topic model analyses of machine‐classified plankton images taken near a retentive eddy during the 2021 EXPORTS North Atlantic field campaign Surface advection and stirring during storms controlled the surface plankton community of the eddy as it transitioned out of a diatom bloom
@article{san_soucie_spatiotemporal_2024,
title = {Spatiotemporal {Topic} {Modeling} {Reveals} {Storm}‐{Driven} {Advection} and {Stirring} {Control} {Plankton} {Community} {Variability} in an {Open} {Ocean} {Eddy}},
volume = {129},
copyright = {All rights reserved},
issn = {2169-9275, 2169-9291},
url = {https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024JC020907},
doi = {10.1029/2024JC020907},
abstract = {Abstract
Phytoplankton communities in the open ocean are high‐dimensional, sparse, and spatiotemporally heterogeneous. The advent of automated imaging systems has enabled high‐resolution observation of these communities, but the amounts of data and their statistical properties make analysis with traditional approaches challenging. Spatiotemporal topic models offer an unsupervised and interpretable approach to dimensionality reduction of sparse, high‐dimensional categorical data. Here we use topic modeling to analyze neural‐network‐classified phytoplankton imagery taken in and around a retentive eddy during the 2021 North Atlantic EXport Processes in the Ocean from Remote Sensing (EXPORTS) field campaign. We investigate the role physical‐biological interactions play in altering plankton community composition within the eddy. Analysis of a water mass mixing framework suggests that storm‐driven surface advection and stirring were major drivers of the progression of the eddy plankton community away from a diatom bloom over the course of the cruise.
,
Plain Language Summary
Plankton communities in the ocean can have many different species, with large differences in their abundance and patchy distributions in space. Automated imaging systems allow for high‐resolution observation of these plankton communities, but many traditional statistical techniques fail to capture their full complexity. Spatiotemporal topic models, a kind of statistical model designed to work directly with categorical data, can effectively represent this kind of information. In this work, we use topic models to analyze plankton images taken near an eddy in the spring of 2021 and classified into 50 different kinds of plankton with an automated algorithm. We investigate how interactions between ocean physics and biology can change the plankton community inside the eddy. Analysis suggests that storms in the area moved surface water carrying a different plankton community into the eddy.
,
Key Points
Topic models provide a robust alternative to traditional statistical techniques for analysis of sparse, high‐dimensional categorical data
We perform topic model analyses of machine‐classified plankton images taken near a retentive eddy during the 2021 EXPORTS North Atlantic field campaign
Surface advection and stirring during storms controlled the surface plankton community of the eddy as it transitioned out of a diatom bloom},
language = {en},
number = {11},
urldate = {2025-01-02},
journal = {Journal of Geophysical Research: Oceans},
author = {San Soucie, John E. and Girdhar, Yogesh and Johnson, Leah and Peacock, Emily E. and Shalapyonok, Alexi and Sosik, Heidi M.},
month = nov,
year = {2024},
pages = {e2024JC020907},
}
Downloads: 1
{"_id":"b3QARZ8MprzoavdMN","bibbaseid":"sansoucie-girdhar-johnson-peacock-shalapyonok-sosik-spatiotemporaltopicmodelingrevealsstormdrivenadvectionandstirringcontrolplanktoncommunityvariabilityinanopenoceaneddy-2024","author_short":["San Soucie, J. E.","Girdhar, Y.","Johnson, L.","Peacock, E. E.","Shalapyonok, A.","Sosik, H. M."],"bibdata":{"bibtype":"article","type":"article","title":"Spatiotemporal Topic Modeling Reveals Storm‐Driven Advection and Stirring Control Plankton Community Variability in an Open Ocean Eddy","volume":"129","copyright":"All rights reserved","issn":"2169-9275, 2169-9291","url":"https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024JC020907","doi":"10.1029/2024JC020907","abstract":"Abstract Phytoplankton communities in the open ocean are high‐dimensional, sparse, and spatiotemporally heterogeneous. The advent of automated imaging systems has enabled high‐resolution observation of these communities, but the amounts of data and their statistical properties make analysis with traditional approaches challenging. Spatiotemporal topic models offer an unsupervised and interpretable approach to dimensionality reduction of sparse, high‐dimensional categorical data. Here we use topic modeling to analyze neural‐network‐classified phytoplankton imagery taken in and around a retentive eddy during the 2021 North Atlantic EXport Processes in the Ocean from Remote Sensing (EXPORTS) field campaign. We investigate the role physical‐biological interactions play in altering plankton community composition within the eddy. Analysis of a water mass mixing framework suggests that storm‐driven surface advection and stirring were major drivers of the progression of the eddy plankton community away from a diatom bloom over the course of the cruise. , Plain Language Summary Plankton communities in the ocean can have many different species, with large differences in their abundance and patchy distributions in space. Automated imaging systems allow for high‐resolution observation of these plankton communities, but many traditional statistical techniques fail to capture their full complexity. Spatiotemporal topic models, a kind of statistical model designed to work directly with categorical data, can effectively represent this kind of information. In this work, we use topic models to analyze plankton images taken near an eddy in the spring of 2021 and classified into 50 different kinds of plankton with an automated algorithm. We investigate how interactions between ocean physics and biology can change the plankton community inside the eddy. Analysis suggests that storms in the area moved surface water carrying a different plankton community into the eddy. , Key Points Topic models provide a robust alternative to traditional statistical techniques for analysis of sparse, high‐dimensional categorical data We perform topic model analyses of machine‐classified plankton images taken near a retentive eddy during the 2021 EXPORTS North Atlantic field campaign Surface advection and stirring during storms controlled the surface plankton community of the eddy as it transitioned out of a diatom bloom","language":"en","number":"11","urldate":"2025-01-02","journal":"Journal of Geophysical Research: Oceans","author":[{"propositions":[],"lastnames":["San","Soucie"],"firstnames":["John","E."],"suffixes":[]},{"propositions":[],"lastnames":["Girdhar"],"firstnames":["Yogesh"],"suffixes":[]},{"propositions":[],"lastnames":["Johnson"],"firstnames":["Leah"],"suffixes":[]},{"propositions":[],"lastnames":["Peacock"],"firstnames":["Emily","E."],"suffixes":[]},{"propositions":[],"lastnames":["Shalapyonok"],"firstnames":["Alexi"],"suffixes":[]},{"propositions":[],"lastnames":["Sosik"],"firstnames":["Heidi","M."],"suffixes":[]}],"month":"November","year":"2024","pages":"e2024JC020907","bibtex":"@article{san_soucie_spatiotemporal_2024,\n\ttitle = {Spatiotemporal {Topic} {Modeling} {Reveals} {Storm}‐{Driven} {Advection} and {Stirring} {Control} {Plankton} {Community} {Variability} in an {Open} {Ocean} {Eddy}},\n\tvolume = {129},\n\tcopyright = {All rights reserved},\n\tissn = {2169-9275, 2169-9291},\n\turl = {https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024JC020907},\n\tdoi = {10.1029/2024JC020907},\n\tabstract = {Abstract\n Phytoplankton communities in the open ocean are high‐dimensional, sparse, and spatiotemporally heterogeneous. The advent of automated imaging systems has enabled high‐resolution observation of these communities, but the amounts of data and their statistical properties make analysis with traditional approaches challenging. Spatiotemporal topic models offer an unsupervised and interpretable approach to dimensionality reduction of sparse, high‐dimensional categorical data. Here we use topic modeling to analyze neural‐network‐classified phytoplankton imagery taken in and around a retentive eddy during the 2021 North Atlantic EXport Processes in the Ocean from Remote Sensing (EXPORTS) field campaign. We investigate the role physical‐biological interactions play in altering plankton community composition within the eddy. Analysis of a water mass mixing framework suggests that storm‐driven surface advection and stirring were major drivers of the progression of the eddy plankton community away from a diatom bloom over the course of the cruise.\n , \n Plain Language Summary\n Plankton communities in the ocean can have many different species, with large differences in their abundance and patchy distributions in space. Automated imaging systems allow for high‐resolution observation of these plankton communities, but many traditional statistical techniques fail to capture their full complexity. Spatiotemporal topic models, a kind of statistical model designed to work directly with categorical data, can effectively represent this kind of information. In this work, we use topic models to analyze plankton images taken near an eddy in the spring of 2021 and classified into 50 different kinds of plankton with an automated algorithm. We investigate how interactions between ocean physics and biology can change the plankton community inside the eddy. Analysis suggests that storms in the area moved surface water carrying a different plankton community into the eddy.\n , \n Key Points\n \n \n \n Topic models provide a robust alternative to traditional statistical techniques for analysis of sparse, high‐dimensional categorical data\n \n \n We perform topic model analyses of machine‐classified plankton images taken near a retentive eddy during the 2021 EXPORTS North Atlantic field campaign\n \n \n Surface advection and stirring during storms controlled the surface plankton community of the eddy as it transitioned out of a diatom bloom},\n\tlanguage = {en},\n\tnumber = {11},\n\turldate = {2025-01-02},\n\tjournal = {Journal of Geophysical Research: Oceans},\n\tauthor = {San Soucie, John E. and Girdhar, Yogesh and Johnson, Leah and Peacock, Emily E. and Shalapyonok, Alexi and Sosik, Heidi M.},\n\tmonth = nov,\n\tyear = {2024},\n\tpages = {e2024JC020907},\n}\n\n\n\n","author_short":["San Soucie, J. E.","Girdhar, Y.","Johnson, L.","Peacock, E. E.","Shalapyonok, A.","Sosik, H. M."],"key":"san_soucie_spatiotemporal_2024","id":"san_soucie_spatiotemporal_2024","bibbaseid":"sansoucie-girdhar-johnson-peacock-shalapyonok-sosik-spatiotemporaltopicmodelingrevealsstormdrivenadvectionandstirringcontrolplanktoncommunityvariabilityinanopenoceaneddy-2024","role":"author","urls":{"Paper":"https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2024JC020907"},"metadata":{"authorlinks":{}},"downloads":1},"bibtype":"article","biburl":"https://bibbase.org/zotero-mypublications/girdhar","dataSources":["Aba7SDZ3i2zffh3rY"],"keywords":[],"search_terms":["spatiotemporal","topic","modeling","reveals","storm","driven","advection","stirring","control","plankton","community","variability","open","ocean","eddy","san soucie","girdhar","johnson","peacock","shalapyonok","sosik"],"title":"Spatiotemporal Topic Modeling Reveals Storm‐Driven Advection and Stirring Control Plankton Community Variability in an Open Ocean Eddy","year":2024,"downloads":2}