QTCM version 2.3 includes a simple atmospheric boundary layer.
It assumes a steady state, vertically homogeneous mixed layer
of fixed height 
. This reduces the momentum equation in the
boundary layer to a simple balance between the Coriolis force due
to the boundary layer wind 
, the boundary
layer's average gradient of the geopotential 
and the turbulent
stress 
at top and bottom
(see Stevens et al. [10]):
The stress 
at the top of the mixed layer is assumed to be
proportional to the velocity jump at the interface
with an entrainment velocity
as its proportionality constant. For the surface stress the wind speed
in the standard bulk formula is augmented by a minimum
surface wind speed 
which crudely parameterizes the additional
stress due to surface wind variance not captured in QTCM (gustiness etc.):
, 
.
The result is a nonlinear vector equation for the boundary layer wind. In components:
These equations are solved for 
and 
in subroutine abl using
a Newton-Ralphon procedure. The routine uses the surface wind of the prior time
step as the initial guess. Initially, the routine uses the 
as
the initiall guess(*).
The input of abl are the surface geopotential gradient 
and the velocity at the top of the ABL 
.
Besides and abl returns 
which is then used in the
bulk formulas for the latent and sensible heat fluxes.
The contribution
of the minimum wind speed is incorporated in . The surface wind
components ,
, however, are not augmented by the minimum wind speed and are
thus not directly used as the surface wind in the bulk formulas for scalar fluxes.
(*)The surface wind is currently not saved in the restart file which may make the model not exactly restartable, i.e., the trajectories of a restarted run may diverge from the original trajectory.