Beyond Spectra: Macroturbulence Observations Select High-Resolution Ocean Models
Abstract
High resolution is not a panacea. Presently, the most advanced high-resolution modelsuse numerical and sub-grid schemes which are c""hosen for simplicity, stability, and tradition. Some of these schemes are inconsistent with physical principles and observations but" are used anyway in the hope that resolved phenomena will be sufficient to overcome these deficiencies. Other schemes may be inconsi"stent with observations, but no existing analysis of the data makes this obvious. In this project, we intend to provide two novel sy""nthesis datasets custom-built to provide accurate, precise statistical information~globally, seasonally, and at depth~that can be us""ed to identify when a subgrid model is failing to produce realistic turbulence at the smallest resolved scales (i.e., macroturbulenc""e). These data will also inform our understanding of ocean turbulence regimes, through variations by region and season in the scale"" dependence, magnitude, direction, and type of turbulent cascades where they are observable.Our results indicate that choosing amon""g reasonable, extant, and in-development sub-gridclosures in simulations with grids significantly finer than the deformation radius"" is not trivial.Scale-selective turbulence statistics indicate that differences arise near the gridscale, but these accumulate some""times to affect larger separation scales, even beyond the deformation radius. While satellite and in situ observations exist which m""ay help identify these problems, in practice preparation and analysis of these resources is much harder than for datasets prepared f"or coarse-resolution model evaluation. Our datasets will streamline this evaluation process just as climatological mean datasets str"eamline the evaluation of coarse-resolution models.This project will examine macroturbulence in satellite, in situ, and other remot""e data, focusingon high-resolution data such as the 350m VIIRS SST observations and GLAD, LASER, and Oleander data prepared by our" team. The analytic framework of structure functions will be used to organize the collection. Structure functions are one of the old"est tools in turbulence theory and experiment. The simplest, second-order structure function contains much the same information as a" power spectrum (although with certain advantages when used with gappy or irregular data). This trusted form will be used to form ou"r first global ocean macroturbulence analysis dataset. Higher-order and blended structure functions, under certain simplifying assum""ptions, can give deep insight into intermittency, scale-to-scale property transfers, sources of energy and enstrophy, and identifica"tion of turbulence regimes. Our second dataset will contain calculations of higher order and blended-variable structure functions where data exist and robustly exceed noise levels. It is a stringent test to match these precise constraints with high-resolution mode"ls, and these data will directly further our understanding of macroturbulence regimes and parameters throughout the global oceans.I""ntellectual Merit: This project will provide a collection of observational analyses, custom-designed to act as evaluation tools for"" high-resolution ocean models and macroturbulence theory. The dynamical regimes, familiar, unfamiliar, and novel worldwide will be i"dentified regionally and seasonally using observations. Uncertainty estimates will be provided alongside the datasets.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 29, 2017
- Source ID
- N000141712963
Entities
People
- Baylor Fox-Kemper
Organizations
- Brown University
- Office of Naval Research
- United States Navy