Data Assimilation Efforts for Oregon Shelf Flows
Alexander L. Kurapov, Peter R. Oke, Lana Erofeeva,
J. S. Allen, Gary D. Egbert, Robert N. Miller
College of Oceanic and Atmospheric Sciences
Oregon State University
Our objective is to build a coastal data assimilation (DA) system to
constrain circulation models with observations such as HF radar surface
currents, moored ADCP currents, and ship-borne data collected in
Oregon coastal waters. Two main components of the system are a POM-based
DA model of wind-driven circulation and a spectral in time tidal DA
model, that also uses sigma-coordinates. In the wind driven circulation
DA model, surface HF radar currents for the summer of 1998 were
assimilated using an optimal interpolation (OI) sequential algorithm.
A time-distributed averaging procedure was incorporated in an OI DA
scheme to overcome problems of data compatibility and initialization.
The baroclinic tidal model is based on the linearized primitive
equations. HF radar surface currents harmonicly analyzed in short time
windows were assimilated into this model using a variational generalized
inverse method and an efficient representer algorithm. Validation of
the DA results against independent ADCP measurements suggests that
surface current measurements contain valuable information about
subsurface flows, both subinertial and superinertial. Connections
between the two components of the DA are being established. Theoretical
analysis indicates that the spectral model, applied at subinertial
frequencies, can be used to build a forecast error covariance for the
time-dependent wind circulation DA model. The tidal model may be used
to detide ship-borne observations before they are assimilated in a
wind-forced circulation model. The latter, in turn, can supply the
tidal model with time-variable hydrographic fields and subinertial
currents that are believed to cause internal tide.