Mesoscale eddies are a well-recognized feature of the global oceans, being of dynamical relevance to various ocean processes and of significance to transport of various species. In the last few decades, advancements in satellite observations of the sea surface have revealed mesoscale and sub-mesoscale eddy features in many parts of the ocean. An objective definition and detection of eddies has, however, remained a challenge, thus resulting in unreliable estimates of eddy formation, evolution and dissipation. In this study, we present an implementation of a Lagrangian averaged vorticity deviation (LAVD) based method to detect and characterize eddies in the Bay of Bengal (BoB), a freshwater-dominated semi-enclosed basin in the northern Indian Ocean. Systematic sensitivity studies are performed to identify parameter values that result in robust eddy detection in the BoB, and finite-time coherence in material transport associated with the LAVD-based eddies is demonstrated. LAVD-based eddy detection is then used to statistically characterize the Sri Lanka dome (SLD) eddy, an annual mesoscale cyclonic eddy of dynamical and physical importance in the southern BoB. Analysing satellite-based ocean surface currents from 27 years (1993–2019) of data, the inter-annual variability in the SLD formation, its spatio-temporal evolution and dissipation are elucidated. On an average, SLD has a lifetime of 15 weeks and a maximum area of 5 square degrees, with the longest lifetime of 38 weeks in 1996 and a maximum area of 12 square degrees in 1998. Long-lived SLDs are shown to spatially excurse into the BoB, potentially influencing large-scale transport of physical and biogeochemical properties. A potential correlation between large SLD lifetime and large variations in the monthly evolution of Dipole Mode Index and NINO3.4 is highlighted. An SSH-based Eulerian eddy detection method is also implemented to understand its efficacy in identifying materially coherent eddies in the BoB region. The paper concludes with a summary of our results, and a brief discussion of potential applications of LAVD-based eddy detection in the northern Indian Ocean.

中尺度涡流是全球海洋的一个公认特征，与各种海洋过程具有动力学相关性，并且对各种物种的运输具有重要意义。在过去的几十年里，卫星对海面观测的进步揭示了海洋许多部分的中尺度和亚中尺度涡流特征。然而，涡流的客观定义和检测仍然是一个挑战，从而导致对涡流形成、演变和消散的估计不可靠。在这项研究中，我们提出了一种基于拉格朗日平均涡度偏差（LAVD）的方法来检测和表征孟加拉湾（BoB）的涡流，孟加拉湾是北印度洋以淡水为主的半封闭盆地。进行了系统灵敏度研究，以确定在 BoB 中实现稳健涡流检测的参数值，并证明了与基于 LAVD 的涡流相关的物质传输的有限时间相干性。然后，使用基于 LAVD 的涡流检测对斯里兰卡圆顶 (SLD) 涡流进行统计表征，这是南部 BoB 具有动力学和物理重要性的年度中尺度气旋涡流。通过分析 27 年（1993 年至 2019 年）的卫星海洋表面流数据，阐明了 SLD 形成的年际变化、其时空演化和消散。平均而言，SLD 的寿命为 15 周，最大面积为 5 平方度，最长寿命为 38 周，1996 年，最大面积为 12 平方度，1998 年。 BoB，可能影响物理和生物地球化学特性的大规模传输。强调了较长的 SLD 寿命与偶极模指数和 NINO3.4 每月演变的较大变化之间的潜在相关性。还实施了基于 SSH 的欧拉涡流检测方法，以了解其在识别 BoB 区域中的物质相干涡流方面的功效。本文最后总结了我们的研究结果，并简要讨论了基于 LAVD 的涡流探测在北印度洋的潜在应用。