_BML.dox

/*! \page BML Bulk Surface Mixed Layer

This bulk surface mixed layer scheme was designed to be used with a

purely isopycnal model. Following \cite niiler1977, \cite oberhuber1993,

and Hallberg (\cite muller2003) the TKE budget is used to construct a

time-evolving homogeneous mixed layer. A buffer layer sits between

the mixed layer and the interior ocean to mediate between the two.

 The following processes are executed, in the order listed.

\li 1. Undergo convective adjustment into mixed layer.

\li 2. Apply surface heating and cooling.

\li 3. Starting from the top, entrain whatever fluid the TKE budget

      permits.  Penetrating shortwave radiation is also applied at

     this point.

\li 4. If there is any unentrained fluid that was formerly in the

     mixed layer, detrain this fluid into the buffer layer.  This

    is equivalent to the mixed layer detraining to the Monin-

      Obukhov depth.

\li 5. Divide the fluid in the mixed layer evenly into CS\%nkml pieces.

\li 6. Split the buffer layer if appropriate.

Layers 1 to nkml are the mixed layer, nkml+1 to nkml+nkbl are the

buffer layers. The results of this subroutine are mathematically

identical if there are multiple pieces of the mixed layer with

the same density or if there is just a single layer. There is no

stability limit on the time step.

The key parameters for the mixed layer are found in the control structure.

These include mstar, nstar, nstar2, pen\_SW\_frac, pen\_SW\_scale, and TKE\_decay.

For the \cite oberhuber1993 and \cite kraus1967 mixed layers, the values of these are:

<table>

<caption id="table_symbols_bml">Model variables used in the bulk mixed layer</caption>

<tr><th>Symbol <th>Value in Oberhuber (1993) <th>Value in Kraus-Turner (1967)

<tr><td>pen\_SW\_frac    <td> 0.42   <td> 0.0

<tr><td>pen\_SW\_scale   <td> 15.0 m <td> 0.0 m

<tr><td>mstar            <td> 1.25   <td> 1.25

<tr><td>nstar            <td> 1      <td> 0.4

<tr><td>TKE\_decay       <td> 2.5    <td> 0.0

<tr><td>conv\_decay      <td> 0.5    <td> 0.0

</table>

TKE\_decay is \f$1/\kappa\f$ in eq. 28 of \cite oberhuber1993, while

conv\_decay is \f$1/\mu\f$. Conv\_decay has been eliminated in favor of

the well-calibrated form for the efficiency of penetrating convection

from \cite wang2003.

*/