next up previous
Next: Conclusions Up: No Title Previous: The optimization method

Preliminary results

The figures below illustrate first results obtained after 12 minimization iterations. The total cost function (see below) is decreasing continuously with the model adjusting first to the climatologies of both T and S. The misfits to the altimetric observations,which initially increase, begin decreasing with the 11th iteration.
 
Figure 6:
\begin{figure} \setlength{\unitlength}{1cm} \begin{picture} (15,15)(1,+0) \e... ...file=wmo_fc1.eps, height=350pt, width=200pt, angle=-0}\end{picture} \end{figure}


 
Figure 7: Adjustment of the initial T/S fields.
\begin{figure} \setlength{\unitlength}{1cm} \begin{picture} (16,10)(1,-1) \e... ...le=wmo_xxinit.ps, height=400pt, width=300pt, angle=90}\end{picture} \end{figure}


 
Figure 8: Adjustment of the heat and fresh water fluxes.
\begin{figure} \setlength{\unitlength}{1cm} \begin{picture} (15,10)(1,-0) \e... ...ile=wmo_flux.eps, height=400pt, width=300pt, angle=90}\end{picture} \end{figure}


 
Figure 9: Velocity and temperature fields at 27.5 m - Mean over August. The main differences between the initial run (model unconstrained) and the optimized one are located at the western boundaries, where the currents are intensified. The eastward path of the Gulf Stream off the North American coast is moved southward, while off-shore waters are significantly cooled. Note also the enhancement of the equatorial westward currents.
\begin{figure} \setlength{\unitlength}{1cm} \begin{picture} (16,10)(2,-0) \e... ...le=agu_diff2.eps, height=200pt, width=300pt, angle=90}\end{picture} \end{figure}


Detlef Stammer
1999-06-22