22nd Conference on Climate Variability and Change

406

4DVAR data assimilation and Adjoint-based sensitivity analysis in an eddy resolving ocean model of the Peruvian Current System

Aneesh Subramanian, SIO/Univ. Of California, La Jolla, CA; and A. J. Miller, B. Cornuelle, and E. Di Lorenzo

Peruvian Current system plays a key role in the climate variability of the Eastern Tropical Pacific and also has teleconnections to other regions of high climate impact like the Tropical Pacific. Strong coastal jets and winds influence this region strongly. The oceanic circulation is characterized by strong eddy activity with short zonal scales. The highly active primary production in this region with a very strong concomitant upwelling is of great importance to the surrounding economies. Data assimilation of the sea surface temperature, height and subsurface temperature and salinity data measured from CTD cast from a ship cruise for a period of two months in 2008 is achieved by using the inverse Regional Ocean Modeling System (iROMS), a 4D-variational data assimilation system for high-resolution basin-wide and coastal oceanic flows. Associated with the data assimilation platform of ROMS is a suite of Generalized Stability Analysis tools, which allow the quantitative assessment of sensitivities of model solutions to various parameters. These include computation of the eigenmodes of the tangent linear model (standard linear stability analysis normal modes), eigenmodes of the adjoint model (optimal eigenmodes for exciting these normal modes), singular vectors (fasting growing modes over finite time intervals), stochastic optimals (patterns of stochastic forcing that account for the largest amount of stochastic response in the model solution) and forcing singular vectors (forcing patterns that maximize model variability over a specified region for a given time period). We present the methods and initial results for the assimilation and also sensitivity tests of the ocean state in this region to various forcing conditions.

Poster Session , Advances in Modeling
Wednesday, 20 January 2010, 2:30 PM-4:00 PM

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page