[Mitgcm-support] Impact of prescribed mixing

[mitgcm-support at dev.mitgcm.org]

HyperNews notification of new message.  See: 
  http://escher.JPL.NASA.GOV:2000/HyperNews/get/forums/assim/110/5.html

The adjoint assimilation uses prescribed mixing coefficients
pre-computed from the time-mean of KPP mixing coefficients for 1997. The
impact of which (relative to the online time-dependent KPP coefficients)
is assessed to understand the limitation of the prescribed mixing.
Idealy, the assimilation should exceeds the skill of the best control
run (i.e., the one that uses KPP). This figure[1] compares the
simulation for 1997 with prescribed mixing (used by adjoint assimilation
as prior) and our control run (c20000630) which use full kpp with TAO
data. Red solid - prescribed mixing, black dotted - control run with
KPP, blue dashed - assimilation. KPP mixing outperforms the prescribed
mixing in the mixed layer (smaller rms deviation from and larger
correlation with TAO data, lower 2 panels). I've looked at some time
series for both and found that the prescribed mixing temperature in the
mixed layer lacks the episodic time-dependent mixing events. During the
1997 El Nino, the mixed layer in the eastern tropical Pacific is deep.
The SST in the prescribed mixing runs are not mixed deep enough
comparing to the online KPP mixing because the former was computed from
1997 annual mean. Consequently, SST with prescribed mixing has poorer
skill than time-dependent KPP mixing. What's surprising is the that,
below the mixed layer, prescribed mixing does bettter than
time-dependent KPP mixing. An example[2] temperature time series at 2n,
155w, 100 m shows that, over much of 1997, the simulation using
prescribed mixing (no assimilation, green) is closer to the TAO data
(red) than the simulation with KPP does (blue). Therefore, employment of
KPP in assimilation can improve SST, but degraded subsurface
temperature. The cause of which is unknown (may have to do with the
shear instability term in KPP? - according to Dimitris.)

-----------------------------------------------------------------------
[1] http://escher.jpl.nasa.gov:2000/hosts/escher/escher4/medea/tlee/FIG_FORUM/plot_statt_ave_kpp.jpg
[2] http://escher.jpl.nasa.gov:2000/hosts/escher/escher4/medea/tlee/FIG_FORUM/t_2n155w_100m_kpp.jpg