A Linear Decomposition of the Southern Ocean Thermohaline Structure

Pauthenet, Etienne; Roquet, Fabien; Madec, Gurvan; Nérini, David

doi:10.1175/jpo-d-16-0083.1

Cited by 31 publications

(49 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An alternative approach is to use functional principal component analysis (fPCA), in which PCA is performed on functions instead of the original data. In Pauthenet et al (), the authors represent vertical temperature and salinity profiles from the Southern Ocean State Estimate (Mazloff et al, ) as linear combinations of B‐spline basis functions and apply fPCA to the resulting spline functions. They use the PCs to examine large‐scale structures such as fronts in the SO.…”

Section: Discussionmentioning

confidence: 99%

Unsupervised Clustering of Southern Ocean Argo Float Temperature Profiles

et al. 2019

View full text Add to dashboard Cite

The Southern Ocean has complex spatial variability, characterized by sharp fronts, steeply tilted isopycnals, and deep seasonal mixed layers. Methods of defining Southern Ocean spatial structures traditionally rely on somewhat ad hoc combinations of physical, chemical, and dynamic properties. As a step toward an alternative approach for describing spatial variability in temperature, here we apply an unsupervised classification technique (i.e., Gaussian mixture modeling or GMM) to Southern Ocean Argo float temperature profiles. GMM, without using any latitude or longitude information, automatically identifies several spatially coherent circumpolar classes influenced by the Antarctic Circumpolar Current. In addition, GMM identifies classes that bear the imprint of mode/intermediate water formation and export, large-scale gyre circulation, and the Agulhas Current, among others. Because GMM is robust, standardized, and automated, it can potentially be used to identify structures (such as fronts) in both observational and model data sets, possibly making it a useful complement to existing classification techniques. Plain Language SummaryThe Southern Ocean is an important part of the climate system, in part because it absorbs a large fraction of the heat and carbon that is added to the atmosphere/ocean system by human-driven fossil fuel burning. In this work, we use a machine learning technique to automatically sort Southern Ocean temperature measurements into groups based on how those temperature measurements change with depth. Different groups have the fingerprints of different large-scale circulation patterns, such as the powerful Antarctic Circumpolar Current that flows around Antarctica. The groups that we identify are consistent with our understanding of the Southern Ocean, which gives us confidence that our machine learning technique may be useful for automatically grouping measurements and computer model data in the future. This matters because the climate science community needs a new set of tools, possibly including the machine learning technique that we use in this paper, to deal with a very large, ever-increasing volume of observational and computer model data.

show abstract

Section: Discussionmentioning

confidence: 99%

Unsupervised Clustering of Southern Ocean Argo Float Temperature Profiles

et al. 2019

View full text Add to dashboard Cite

show abstract

“…We used an objective method developed in Pauthenet et al (2017) to decompose in modes the vertical structure of T and S profiles and locate the PF in the upper 300 m of the Southern Indian Ocean, between 2007 and Figure 4. The transport between Heard and Kerguelen Islands (orange) computed from GLORYS is anticorrelated with the average PF latitude between 60 and 70 ∘ E (black).…”

Section: Discussionmentioning

confidence: 99%

“…However, in the Antarctic Zone, PC1 is almost constant and PC2 reveals additional structures such as the intense Fawn Trough current ) across the Kerguelen Plateau and contours following the SACCF. Pauthenet et al, 2017). (d) Two-dimensional kernel density estimate of the mean PC1 and PC2 values, with a 0.2 contour interval.…”

Section: Vertical Modesmentioning

confidence: 99%

Seasonal Meandering of the Polar Front Upstream of the Kerguelen Plateau

Pauthenet

Roquet

Madec

et al. 2018

Geophysical Research Letters

View full text Add to dashboard Cite

The location of the Antarctic Polar Front (PF) is mapped in the Southern Indian Ocean by decomposing the shape of temperature and salinity profiles into vertical modes using a functional Principal Component Analysis. We define the PF as the northernmost minimum of temperature at the subsurface and represent it as a linear combination of the first three modes. This method is applied on an ocean reanalysis data set and on in situ observations, revealing a seasonal variability of the PF latitudinal position that is most pronounced between the Conrad Rise and the Kerguelen Plateau. This shift coincides with variations in the transport across the Northern Kerguelen Plateau. We suggest that seasonal changes of the upper stratification may drive the observed variability of the PF, with potentially large implications for the pathways and residence time of water masses over the plateau and the phytoplankton bloom extending southeast of the Kerguelen Islands. Plain Language SummaryThe Antarctic Polar Front (PF) is a water mass boundary that flows around Antarctica between approximately 48 ∘ S and 56 ∘ S in the Southern Indian Ocean. The position of the PF in space and time is important to understand the oceanic circulation, the heat and salt exchanges, and also marine ecosystems. In the Indian sector the PF has to cross the Kerguelen Plateau, a major bottom topography feature. The present study develops and then applies a novel method for mapping the PF taking into account the whole hydrographic structure in the upper 300 m of the ocean. We are able to map the PF position and find that it presents large seasonal variations that are more intense just west of the Kerguelen Plateau. Between the Conrad Rise and the Kerguelen Plateau, the PF is essentially zonally orientated in September and found farther south by up to 4 ∘ latitude in March. Shifts in the PF position are shown to correlate with a seasonal variation in volume transport between Kerguelen and Heard Islands. We discuss how these seasonal variations in circulation pathways could have an impact on the local marine ecosystems.

show abstract

“…At RT of T = 293 K, the net M S of the sample is mainly contributed by the Co/Ni bilayer since the magnetic moments of TbCo layer are nearly compensated. As T decreases, because the magnetic moment of Tb atom increases more significantly with T decreasing than that of Co(Ni), the switching field in AHE loops rises with the reduction of the net M S and reaches the maximum value at T near the T M . When T is further decreased, the AHE loop polarity of the sample reverses and its switching field starts to decrease, indicating the net M S of the sample turns to Tb dominant at T < T M .…”

Section: Resultsmentioning

confidence: 98%