The ROMS IAS Data Assimilation and Prediction System: Quantifying Uncertainty
Abstract
The main objectives of this project were: (i) to assess the impact of observations on ocean state estimates and the ensuing forecasts; (ii) to quantify the expected errors in 4-dimensional variational (4D-Var) ocean circulation estimates: and (iii) to develop multimodel ensemble and superensemble methods for ocean models. The primary tool used in this project was the Regional Ocean Modeling System (ROMS). To address the aforementioned goals and objectives, we used a recently developed suite of tools that utilize the tangent linear , adjoint, and finite-amplitude tangent linear versions of the ROMS code. Three 4D-Var data assimilation systems have been developed for ROMS, one based on the primal formulation and two based on the dual formulation. During the project we completed the following tasks: (I) observation impact, (2) observation sensitivity, (3) Analysis and forecast error estimates based on adjoint 4D-Var, (4) Graduate student and post-doc training and mentoring, (5) a 4D-Var training workshop, and (6) preparation and publication of five manuscripts.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 13, 2011
- Accession Number
- ADA550400
Entities
People
- Andrew M. Moore
- Brian Powell
Organizations
- University of California, Santa Cruz