Combining Ensemble and Variational Data Assimilation

Abstract

The long-term goal of this project is to develop and apply practical methods for data assimilation to improve the short-range prediction of mesoscale ocean variability. The primary objective of this work is to develop an ocean data assimilation system that exploits the strengths of both the ensemble-based (e.g., Evensen 2003; Houtekamer and Mitchell 1998; Tippett et al. 2003) and variational (e.g., Bennett 2002) approaches to data assimilation. The first step in this project is to perform a comprehensive inter-comparison of an ensemble-based data assimilation system with a 4d-var system for a suite of coastal model configurations. The second step is to identify the strengths and weaknesses of each system and to improve both systems by borrowing components from the other system. Ultimately, a single ensemble-var system will be developed. We will investigate the extent to which the ensemble-var system can outperform both the ensemble-based and variational approaches, both in terms of forecast skill (accuracy) and computational efficiency (throughput).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 30, 2010
Accession Number
ADA542745

Entities

People

  • Alexandra Kurapov
  • Peter R. Oke

Organizations

  • Oregon State University

Tags

Communities of Interest

  • Space

DTIC Thesaurus Topics

  • Agreements
  • Assimilation
  • Atmospheric Sciences
  • Australia
  • Four Dimensional
  • Heat Flux
  • Observation
  • Ocean Currents
  • Oceans
  • Sea Level
  • Sea Surface Temperature
  • Stresses
  • Surface Temperature
  • Test Beds
  • Universities
  • Upwelling
  • Wind Stress

Fields of Study

  • Environmental science

Readers

  • Atmospheric Science/Meteorology
  • Systems Analysis and Design