Stochastic Parameterization of Ocean Turbulence for Observational Networks

Abstract

We propose to apply a newly developed methodology combining mathematics, simulations and data assimilation to create a stochastic re,gional ocean model (SROM) which is systematically in- formed by observed and simulated submesoscale ocean data from both ONR measure,ments and high-resolution MITgcm and ROMS simulations. To accomplish this objective, the new mathematics and data assimilation proto,col will be used to coordinate deterministic fine-scale and stochastic coarse-scale simulations as bench-marked by observed data set,s from the LatMix ONR mission and (potentially) also the FLEAT ONR mission.In its first step, our new methodology (called SALT) cali,brates stochastic transport on the coarser grid based on correlation information derived from the simulation on the finer grid. The,next step quantifies the uncertainty due to coarser grid resolution by running an ensemble of calibrated stochastic simulations on t,he coarser grid. The final step reduces the uncertainty of runs on the coarser grid by using the particle-filtering method for data,assimilation based on a few sparse observations of the finer scale information evaluated on the coarser grid points. The data assi,milation step with particle filtering has been found in earlier investigations to drastically reduce the uncertainty of the simulati,ons of the coarser grid [CCH+ 19a, CCH+ 19b, CCH+ 20].[CCH+ 19a] Colin Cotter, Dan Crisan, Darryl D Holm, Wei Pan, and Igor Shevchen,ko. Numerically model- ing stochastic Lie transport in fluid dynamics. Multiscale Modeling & Simulation, 17(1):192-232, 2019.[CCH+ 1,9b] Colin Cotter, Dan Crisan, Darryl D Holm, Wei Pan, and Igor Shevchenko. A particle filter for stochastic advection by Lie transpo,rt (SALT): A case study for the damped and forced incompressible 2D Euler equation. arXiv preprint arXiv:1907.11884, 2019, , Accepte,d for publication in SIAM/ASA Journal on Uncertainty Quantification. 16 Sept 2020. [CCH+ 20] Colin Cotter, Dan Crisan, Darryl Holm,,Wei Pan, and Igor Shevchenko. Data assimilation for a quasi- geostrophic model with circulation-preserving stochastic transport nois,e. Journal of Statistical Physics, pages 136, 2020.

Document Details

Document Type
DoD Grant Award
Publication Date
Mar 05, 2022
Source ID
N000142212082

Entities

People

  • Darryl D. Holm

Organizations

  • Imperial College London
  • Office of Naval Research
  • United States Navy

Tags

Readers

  • Atmospheric Science/Meteorology
  • Computational Fluid Dynamics (CFD)
  • Ocean-Atmosphere Mesoscale Modeling, Data Assimilation, and Flux Boundary Layers