ABSTRACT

Ocean flows at relatively large scales, typically more than tens of miles, are mostly confined to horizontal plane due to the rotation of the Earth. At the smaller submesoscale, this constraint is much reduced, allowing more vigorous vertical motion, which plays a crucial role in mediating upper ocean heat content and air-sea exchange. Improving the representation of these processes in numerical models is therefore important for accurate forecasts of a wide range of directly societally relevant phenomena, from ENSO to tropical monsoons to the rate of Arctic sea ice melt. Due to their intrinsic short lateral scales (order of km) and fast time scales (order of inertial period or faster), submesoscale flows have been difficult to observe directly in the ocean. Numerical and analytical studies, on the other hand, have made substantial progress in describing the dynamics of idealized submesoscale flows. Criteria have been developed for the onset of different submesoscale instabilities. Parametrizations have been made to account for these instabilities in global ocean models where they are not resolved. Yet little of the knowledge gained from these studies has been confirmed by or compared with observations, which makes it difficult to move forward with confidence. This project will address this gap by combining development and application of a comprehensive linear stability model with analysis of several substantial oceanographic datasets already in hand. This project largely written by and will fund an early career scientist.

The model, developed by the early career scientist, is novel in that it is forced by observed profiles of density and velocity, unlike many idealized mean states that have been used in the past. Preliminary work demonstrates its capacity to reproduce a vast range of types of submesoscale instabilities, from mixed layer baroclinic instability to symmetric instability to Langmuir circulation, and everything in between. Unlike linear stability models before, it can smoothly transition from one instability class to another; i.e. it produces a broad spectrum of instabilities from a single mean flow state input. The model will be forced by and then compared with amassed datasets from several recent oceanographic expeditions. Two of the data-sets have shown strong evidence of submesoscale activity, both with anecdotal examples and systematic statistics. Model skill will be developed and tested using profiles from several of these examples. In turn, it is anticipated that the model results will help illuminate the nature of the dynamical instabilities present in the observations; because the model is initialized with real gradient profiles and allows a fully complex superposition of a wide range of instabilities, it is uniquely suited for detangling complex observations in a broadband ocean. Comparison of model predictions with available observations that span multiple ocean basins and multiple seasons will help develop understanding of broad patterns of submesoscale variability, the sorts of patterns that can be used to validate new global parametrization development.

Please report errors in award information by writing to: awardsearch@nsf.gov.