Backward FOKKER-PLANCK Equation for Determination of Model Predictability with Uncertain Initial Errors

Abstract

It is widely recognized that uncertainty in atmospheric and oceanic models can be traced back to two factors. First, in defining the state of atmosphere (or ocean), a number of errors are involved arising from the finite resolution of measurement or from discretization in a numerical experiment, as a result of which small-scale "subgrid" processes are either discarded or parameterized. Second, once present, small errors of the kind mentioned above trigger a complex response leading to their subsequent amplification. The model predictability versus boundary condition error was discussed by Chu (1999) using the Lorenz system. The model predictability can be measured by two parameters: instantaneous error (IE) and predictability time (PT). The IE and PT are used for models with and without given initial condition errors, respectively. In this study, the authors first develop a theoretical framework for predictability evaluation using the PT measure, and then they illustrate its usefulness using the one-dimensional probabilistic error growth model proposed by Nicolis (1992).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2001
Accession Number
ADA483220

Entities

People

  • Leonid M. Ivanov
  • Peter C. Chuu

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Atmospheres
  • Boundaries
  • Computational Science
  • Differential Equations
  • Dynamics
  • Ellipsoids
  • Equations
  • Fluid Dynamics
  • Fokker Planck Equations
  • Linear Differential Equations
  • Liouville Equation
  • Noise
  • Oceans
  • Probability
  • Probability Density Functions
  • Random Variables
  • Test And Evaluation

Readers

  • Calculus or Mathematical Analysis
  • Computational Modeling and Simulation
  • Ocean-Atmosphere Mesoscale Modeling, Data Assimilation, and Flux Boundary Layers