A global Lagrangian eddy dataset based on satellite altimetry

Abstract

Abstract. The methods used to identify coherent ocean eddies are either Eulerian or Lagrangian in nature, and nearly all existing eddy dataset are based on the Eulerian method. In this study, millions of Lagrangian particles are advected by satellite-derived surface geostrophic velocities over the period of 1993–2019. Using the method of Lagrangian-averaged vorticity deviation (LAVD), we present a global Lagrangian eddy dataset (GLED v1.0, Liu and Abernathey, 2022, https://doi.org/10.5281/zenodo.7349753). This open-source dataset contains not only the general features (eddy center position, equivalent radius, rotation property, etc.) of eddies with lifetimes of 30, 90, and 180 days but also the trajectories of particles trapped by coherent eddies over the lifetime. We present the statistical features of Lagrangian eddies and compare them with those of the most widely used sea surface height (SSH) eddies, focusing on generation sites, size, and propagation speed. A remarkable feature is that Lagrangian eddies is generally smaller than SSH eddies, with a radius ratio of about 0.5. Also, the estimated mass transport by Lagrangian eddies is nearly an order of magnitude smaller than that by the Eulerian calculation, indicating that the coherent contribution to the total eddy transport is very limited. Our eddy dataset provides an additional option for oceanographers to understand the interaction between coherent eddies and other physical or biochemical processes in the Earth system.

The Sensitive Area for Targeting Observations of Paired Mesoscale Eddies Associated With Sea Surface Height Anomaly Forecasts

“…Thus, it is anticipated that these algorithms can be leveraged to compute CNOPs for more realistic models, paving the way for further exploration of paired eddies sensitivity in future studies. Furthermore, recent works have noted the significance of filamentary structures at the periphery of eddies in material transport (Liu & Abernathey, 2023; Liu et al., 2019). The findings of this study may hold important value for predicting material transport caused by eddies, presenting an interesting avenue for further research.…”

A Lagrangian coherent eddy atlas for biogeochemical applications in the North Pacific Subtropical Gyre

“…Besides the direct applications for interpreting biogeochemical responses, there are several open questions that a Lagrangian coherent eddy atlas would address. Previous studies found that fewer and smaller structures maintain coherency for longer timescales (Abernathey and Haller, 2018;Xia et al, 2022b;Liu and Abernathey, 2023). However, the average lifespan of a coherent structure is unknown.…”

“…Here we define Lagrangian eddies (or RCLVs) using the LAVD method owing to its objectivity, computational efficiency, and mathematical foundation describing material rotation. RCLVs are ubiquitous in satellite altimetry observations (Abernathey and Haller, 2018;Tian et al, 2022;Xia et al, 2022b;Liu and Abernathey, 2023), and their trapping properties have been validated with in situ trajectories of Argo floats (Liu and Abernathey, 2023) and drifters (Encinas-Bartos et al, 2022).…”

“…To interpret biogeochemical developments in eddies, it is useful to study eddy evolution by tracking features over the entirety of their lifespans as SLA eddy atlases do. Because the LAVD method calculates coherency over fixed time intervals, an equivalent RCLV atlas does not exist, which is currently a major shortcoming of the method (Liu and Abernathey, 2023). Evaluations of RCLV evolution have only been performed for case studies and over limited domains.…”

“…Furthermore, biogeochemical states are a function of the physical environment in the recent past, rather than the future. Existing RCLV datasets detect features from satellite observations every 30 d or longer using forward-intime Lagrangian simulations (Abernathey and Haller, 2018;Tian et al, 2022;Xia et al, 2022b;Liu and Abernathey, 2023). Here we used backward-in-time Lagrangian trajectories to identify RCLVs every 8 d from 2000 to 2019 and developed an algorithm to track features over the entirety of their lifespans.…”