Data assimilation via a hydrodynamic model
In this paper an approach to assimilate remote sensed data is presented. The chosen model ISPRAMIX is a prognostic baroclinic model, for (combined) shallow and deep sea applications, based on the primitive equations, the hydrostatic approximation and the Boussinesq assumption with respect to the effect of variable density. The computation of vertical transport by turbulence is based on a KL-turbulence model. A staggered grid is used for the spatial discretisation. The computation of the baroclinic effects is based on separate transport equations for heat and salinity. The adjoint model computes the gradient of a cost function with respect to control variables. It was validated progressively by rebuilding the ISPRAMIX subroutine sequence in successive stages and for different options. In its definitive form the data assimilation procedure allows a choice of the control variable and of the cost function. The results of the numerical experiments obtained so far show that the data assimilation approach works well in quite different, significant cases.
Bibliographic Reference: Paper presented: European Geophysical Society, XXth General Assembly, Hamburg (DE), April 3-7, 1995
Availability: Available from (1) as Paper EN 38940 ORA
Record Number: 199510674 / Last updated on: 1995-07-07
Original language: en
Available languages: en