1 |
\section{Ocean vertical mixing -- |
\subsection{KPP: Nonlocal K-Profile Parameterization for |
2 |
the nonlocal K-profile parameterization scheme KPP |
Diapycnal Mixing} |
|
\label{sectionkpp}} |
|
3 |
|
|
4 |
|
\label{sec:pkg:kpp} |
5 |
|
\begin{rawhtml} |
6 |
|
<!-- CMIREDIR:package_kpp: --> |
7 |
|
\end{rawhtml} |
8 |
|
|
9 |
|
Authors: Dimitris Menemenlis and Patrick Heimbach |
10 |
|
|
11 |
|
\subsubsection{Introduction |
12 |
|
\label{sec:pkg:kpp:intro}} |
13 |
|
|
14 |
|
%---------------------------------------------------------------------- |
15 |
|
|
16 |
|
\subsubsection{KPP configuration and compiling} |
17 |
|
|
18 |
|
As with all MITgcm packages, KPP can be turned on or off at compile time |
19 |
|
% |
20 |
|
\begin{itemize} |
21 |
|
% |
22 |
|
\item |
23 |
|
using the \texttt{packages.conf} file by adding \texttt{kpp} to it, |
24 |
|
% |
25 |
|
\item |
26 |
|
or using \texttt{genmake2} adding |
27 |
|
\texttt{-enable=kpp} or \texttt{-disable=kpp} switches |
28 |
|
% |
29 |
|
\end{itemize} |
30 |
|
(see Section \ref{sect:buildingCode}). |
31 |
|
|
32 |
|
Parts of the KPP code can be enabled or disabled at compile time |
33 |
|
via CPP preprocessor flags. These options are set in |
34 |
|
\texttt{KPP\_OPTIONS.h}. Table \ref{tab:pkg:kpp:cpp} summarizes them. |
35 |
|
|
36 |
|
\begin{table}[h!] |
37 |
|
\label{tab:pkg:kpp:cpp} |
38 |
|
{\footnotesize |
39 |
|
\begin{tabular}{|l|l|} |
40 |
|
\hline |
41 |
|
\textbf{CPP option} & \textbf{Description} \\ |
42 |
|
\hline \hline |
43 |
|
\texttt{\_KPP\_RL} & |
44 |
|
~ \\ |
45 |
|
\texttt{FRUGAL\_KPP} & |
46 |
|
~ \\ |
47 |
|
\texttt{KPP\_SMOOTH\_SHSQ} & |
48 |
|
~ \\ |
49 |
|
\texttt{KPP\_SMOOTH\_DVSQ} & |
50 |
|
~ \\ |
51 |
|
\texttt{KPP\_SMOOTH\_DENS} & |
52 |
|
~ \\ |
53 |
|
\texttt{KPP\_SMOOTH\_VISC} & |
54 |
|
~ \\ |
55 |
|
\texttt{KPP\_SMOOTH\_DIFF} & |
56 |
|
~ \\ |
57 |
|
\texttt{KPP\_ESTIMATE\_UREF} & |
58 |
|
~ \\ |
59 |
|
\texttt{INCLUDE\_DIAGNOSTICS\_INTERFACE\_CODE} & |
60 |
|
~ \\ |
61 |
|
\texttt{KPP\_GHAT} & |
62 |
|
~ \\ |
63 |
|
\texttt{EXCLUDE\_KPP\_SHEAR\_MIX} & |
64 |
|
~ \\ |
65 |
|
\hline |
66 |
|
\end{tabular} |
67 |
|
} |
68 |
|
\caption{~} |
69 |
|
\end{table} |
70 |
|
|
71 |
|
|
72 |
|
|
73 |
|
%---------------------------------------------------------------------- |
74 |
|
|
75 |
|
\subsubsection{Run-time parameters |
76 |
|
\label{sec:pkg:kpp:runtime}} |
77 |
|
|
78 |
|
Run-time parameters are set in files |
79 |
|
\texttt{data.pkg} and \texttt{data.kpp} |
80 |
|
which are read in \texttt{kpp\_readparms.F}. |
81 |
|
Run-time parameters may be broken into 3 categories: |
82 |
|
(i) switching on/off the package at runtime, |
83 |
|
(ii) required MITgcm flags, |
84 |
|
(iii) package flags and parameters. |
85 |
|
|
86 |
|
\paragraph{Enabling the package} |
87 |
|
~ \\ |
88 |
|
% |
89 |
|
The KPP package is switched on at runtime by setting |
90 |
|
\texttt{useKPP = .TRUE.} in \texttt{data.pkg}. |
91 |
|
|
92 |
|
\paragraph{Required MITgcm flags} |
93 |
|
~ \\ |
94 |
|
% |
95 |
|
The following flags/parameters of the MITgcm dynamical |
96 |
|
kernel need to be set in conjunction with KPP: |
97 |
|
|
98 |
|
\begin{tabular}{ll} |
99 |
|
\texttt{implicitViscosity = .TRUE.} & enable implicit vertical viscosity \\ |
100 |
|
\texttt{implicitDiffusion = .TRUE.} & enable implicit vertical diffusion \\ |
101 |
|
\end{tabular} |
102 |
|
|
103 |
|
|
104 |
|
\paragraph{Package flags and parameters} |
105 |
|
~ \\ |
106 |
|
% |
107 |
|
\begin{table}[h!] |
108 |
|
\label{tab:pkg:kpp:runtime_flags} |
109 |
|
{\footnotesize |
110 |
|
\begin{tabular}{|l|c|l|} |
111 |
|
\hline |
112 |
|
\textbf{Flag/parameter} & \textbf{default} & \textbf{Description} \\ |
113 |
|
\hline \hline |
114 |
|
\multicolumn{3}{|c|}{\textit{I/O related parameters} } \\ |
115 |
|
\hline |
116 |
|
kpp\_freq & \texttt{deltaTClock} & |
117 |
|
Recomputation frequency for KPP fields \\ |
118 |
|
kpp\_dumpFreq & \texttt{dumpFreq} & |
119 |
|
Dump frequency of KPP field snapshots \\ |
120 |
|
kpp\_taveFreq & \texttt{taveFreq} & |
121 |
|
Averaging and dump frequency of KPP fields \\ |
122 |
|
KPPmixingMaps & \texttt{.FALSE.} & |
123 |
|
include KPP diagnostic maps in STDOUT \\ |
124 |
|
KPPwriteState & \texttt{.FALSE.} & |
125 |
|
write KPP state to file \\ |
126 |
|
KPP\_ghatUseTotalDiffus & \texttt{.FALSE.} & |
127 |
|
if \texttt{.T.} compute non-local term using total vertical diffusivity \\ |
128 |
|
~ & ~ & |
129 |
|
if \texttt{.F.} use KPP vertical diffusivity \\ |
130 |
|
\hline |
131 |
|
\multicolumn{3}{|c|}{\textit{Genral KPP parameters} } \\ |
132 |
|
\hline |
133 |
|
minKPPhbl & \texttt{delRc(1)} & |
134 |
|
Minimum boundary layer depth \\ |
135 |
|
epsilon & 0.1 & |
136 |
|
nondimensional extent of the surface layer \\ |
137 |
|
vonk & 0.4 & |
138 |
|
von Karman constant \\ |
139 |
|
dB\_dz & 5.2E-5 1/s$^2$ & |
140 |
|
maximum dB/dz in mixed layer hMix \\ |
141 |
|
concs & 98.96 & |
142 |
|
~ \\ |
143 |
|
concv & 1.8 & |
144 |
|
~ \\ |
145 |
|
\hline |
146 |
|
\multicolumn{3}{|c|}{\textit{Boundary layer parameters (S/R \texttt{bldepth})} } \\ |
147 |
|
\hline |
148 |
|
Ricr & 0.3 & |
149 |
|
critical bulk Richardson number \\ |
150 |
|
cekman & 0.7 & |
151 |
|
coefficient for Ekman depth \\ |
152 |
|
cmonob & 1.0 & |
153 |
|
coefficient for Monin-Obukhov depth \\ |
154 |
|
concv & 1.8 & |
155 |
|
ratio of interior to entrainment depth buoyancy frequency \\ |
156 |
|
hbf & 1.0 & |
157 |
|
fraction of depth to which absorbed solar radiation contributes \\ |
158 |
|
~ & ~ & |
159 |
|
to surface buoyancy forcing \\ |
160 |
|
Vtc & \texttt{~} & |
161 |
|
non-dim. coeff. for velocity scale of turbulant velocity shear \\ |
162 |
|
~ & ~ & |
163 |
|
( = function of concv,concs,epsilon,vonk,Ricr) \\ |
164 |
|
\hline |
165 |
|
\multicolumn{3}{|c|}{\textit{Boundary layer mixing parameters (S/R \texttt{blmix})} } \\ |
166 |
|
\hline |
167 |
|
cstar & 10. & |
168 |
|
proportionality coefficient for nonlocal transport \\ |
169 |
|
cg & ~ & |
170 |
|
non-dimensional coefficient for counter-gradient term \\ |
171 |
|
~ & ~ & |
172 |
|
( = function of cstar,vonk,concs,epsilon) \\ |
173 |
|
\hline |
174 |
|
\multicolumn{3}{|c|}{\textit{Interior mixing parameters (S/R \texttt{Ri\_iwmix})} } \\ |
175 |
|
\hline |
176 |
|
Riinfty & 0.7 & |
177 |
|
gradient Richardson number limit for shear instability \\ |
178 |
|
BVDQcon & -0.2E-4 1/s$^2$ & |
179 |
|
Brunt-V\"ai\"sal\"a squared \\ |
180 |
|
difm0 & 0.005 m$^2$/s & |
181 |
|
viscosity max. due to shear instability \\ |
182 |
|
difs0 & 0.005 m$^2$/s & |
183 |
|
tracer diffusivity max. due to shear instability \\ |
184 |
|
dift0 & 0.005 m$^2$/s & |
185 |
|
heat diffusivity max. due to shear instability \\ |
186 |
|
difmcon & 0.1 & |
187 |
|
viscosity due to convective instability \\ |
188 |
|
difscon & 0.1 & |
189 |
|
tracer diffusivity due to convective instability \\ |
190 |
|
diftcon & 0.1 & |
191 |
|
heat diffusivity due to convective instability \\ |
192 |
|
\hline |
193 |
|
\multicolumn{3}{|c|}{\textit{Double-diffusive mixing parameters (S/R \texttt{ddmix})} } \\ |
194 |
|
\hline |
195 |
|
Rrho0 & not used & |
196 |
|
limit for double diffusive density ratio \\ |
197 |
|
dsfmax & not used & |
198 |
|
maximum diffusivity in case of salt fingering \\ |
199 |
|
\hline |
200 |
|
\hline |
201 |
|
\end{tabular} |
202 |
|
} |
203 |
|
\caption{~} |
204 |
|
\end{table} |
205 |
|
|
206 |
|
|
207 |
|
|
208 |
|
%---------------------------------------------------------------------- |
209 |
|
|
210 |
|
\subsubsection{Equations |
211 |
|
\label{sec:pkg:kpp:equations}} |
212 |
|
|
213 |
|
%---------------------------------------------------------------------- |
214 |
|
|
215 |
|
\subsubsection{Key subroutines |
216 |
|
\label{sec:pkg:kpp:subroutines}} |
217 |
|
|
218 |
|
\paragraph{kpp\_calc:} Top-level routine. \\ |
219 |
|
~ |
220 |
|
|
221 |
|
\paragraph{kpp\_mix:} Intermediate-level routine \\ |
222 |
|
~ |
223 |
|
|
224 |
|
\paragraph{ri\_iwmix:} ~ \\ |
225 |
|
% |
226 |
|
Compute interior viscosity and diffusivity coefficients due to |
227 |
|
% |
228 |
|
\begin{itemize} |
229 |
|
% |
230 |
|
\item |
231 |
|
shear instability (dependent on a local gradient Richardson number), |
232 |
|
% |
233 |
|
\item |
234 |
|
to background internal wave activity, and |
235 |
|
% |
236 |
|
\item |
237 |
|
to static instability (local Richardson number < 0). |
238 |
|
% |
239 |
|
\end{itemize} |
240 |
|
|
241 |
|
|
242 |
|
\paragraph{bldepth:} ~ \\ |
243 |
|
% |
244 |
|
The oceanic planetary boundary layer depth, \texttt{hbl}, is determined as |
245 |
|
the shallowest depth where the bulk Richardson number is |
246 |
|
equal to the critical value, \texttt{Ricr}. |
247 |
|
|
248 |
|
Bulk Richardson numbers are evaluated by computing velocity and |
249 |
|
buoyancy differences between values at zgrid(kl) < 0 and surface |
250 |
|
reference values. |
251 |
|
In this configuration, the reference values are equal to the |
252 |
|
values in the surface layer. |
253 |
|
When using a very fine vertical grid, these values should be |
254 |
|
computed as the vertical average of velocity and buoyancy from |
255 |
|
the surface down to epsilon*zgrid(kl). |
256 |
|
|
257 |
|
When the bulk Richardson number at k exceeds Ricr, hbl is |
258 |
|
linearly interpolated between grid levels zgrid(k) and zgrid(k-1). |
259 |
|
|
260 |
|
The water column and the surface forcing are diagnosed for |
261 |
|
stable/ustable forcing conditions, and where hbl is relative |
262 |
|
to grid points (caseA), so that conditional branches can be |
263 |
|
avoided in later subroutines. |
264 |
|
|
265 |
|
\paragraph{blmix:} ~ \\ |
266 |
|
% |
267 |
|
Compute boundary layer mixing coefficients. |
268 |
|
Mixing coefficients within boundary layer depend on surface |
269 |
|
forcing and the magnitude and gradient of interior mixing below |
270 |
|
the boundary layer ("matching"). |
271 |
|
% |
272 |
|
\begin{enumerate} |
273 |
|
% |
274 |
|
\item |
275 |
|
compute velocity scales at hbl |
276 |
|
% |
277 |
|
\item |
278 |
|
find the interior viscosities and derivatives at hbl |
279 |
|
% |
280 |
|
\item |
281 |
|
compute turbulent velocity scales on the interfaces |
282 |
|
% |
283 |
|
\item |
284 |
|
compute the dimensionless shape functions at the interfaces |
285 |
|
% |
286 |
|
\item |
287 |
|
compute boundary layer diffusivities at the interfaces |
288 |
|
% |
289 |
|
\item |
290 |
|
compute nonlocal transport term |
291 |
|
% |
292 |
|
\item |
293 |
|
find diffusivities at kbl-1 grid level |
294 |
|
% |
295 |
|
\end{enumerate} |
296 |
|
|
297 |
|
\paragraph{kpp\_calc\_diff\_t/s, kpp\_calc\_visc:} ~ \\ |
298 |
|
% |
299 |
|
Add contribution to net diffusivity/viscosity from |
300 |
|
KPP diffusivity/viscosity. |
301 |
|
|
302 |
|
\paragraph{kpp\_transport\_t/s/ptr:} ~ \\ |
303 |
|
% |
304 |
|
Add non local KPP transport term (ghat) to diffusive |
305 |
|
temperature/salinity/passive tracer flux. |
306 |
|
The nonlocal transport term is nonzero only for scalars |
307 |
|
in unstable (convective) forcing conditions. |
308 |
|
|
309 |
|
{\footnotesize |
310 |
|
\begin{verbatim} |
311 |
|
|
312 |
|
C !CALLING SEQUENCE: |
313 |
|
c ... |
314 |
|
c kpp_calc (TOP LEVEL ROUTINE) |
315 |
|
c | |
316 |
|
c |-- statekpp: o compute all EOS/density-related arrays |
317 |
|
c | o uses S/R FIND_ALPHA, FIND_BETA, FIND_RHO |
318 |
|
c | |
319 |
|
c |-- kppmix |
320 |
|
c | |--- ri_iwmix (compute interior mixing coefficients due to constant |
321 |
|
c | | internal wave activity, static instability, |
322 |
|
c | | and local shear instability). |
323 |
|
c | | |
324 |
|
c | |--- bldepth (diagnose boundary layer depth) |
325 |
|
c | | |
326 |
|
c | |--- blmix (compute boundary layer diffusivities) |
327 |
|
c | | |
328 |
|
c | |--- enhance (enhance diffusivity at interface kbl - 1) |
329 |
|
c | o |
330 |
|
c | |
331 |
|
c |-- swfrac |
332 |
|
c o |
333 |
|
|
334 |
|
\end{verbatim} |
335 |
|
} |
336 |
|
|
337 |
|
%---------------------------------------------------------------------- |
338 |
|
|
339 |
|
\subsubsection{KPP diagnostics |
340 |
|
\label{sec:pkg:kpp:diagnostics}} |
341 |
|
|
342 |
|
Diagnostics output is available via the diagnostics package |
343 |
|
(see Section \ref{sec:pkg:diagnostics}). |
344 |
|
Available output fields are summarized in |
345 |
|
Table \ref{tab:pkg:kpp:diagnostics}. |
346 |
|
|
347 |
|
\begin{table}[h!] |
348 |
|
\label{tab:pkg:kpp:diagnostics} |
349 |
|
{\footnotesize |
350 |
|
\begin{verbatim} |
351 |
|
------------------------------------------------------ |
352 |
|
<-Name->|Levs|grid|<-- Units -->|<- Tile (max=80c) |
353 |
|
------------------------------------------------------ |
354 |
|
KPPviscA| 23 |SM |m^2/s |KPP vertical eddy viscosity coefficient |
355 |
|
KPPdiffS| 23 |SM |m^2/s |Vertical diffusion coefficient for salt & tracers |
356 |
|
KPPdiffT| 23 |SM |m^2/s |Vertical diffusion coefficient for heat |
357 |
|
KPPghat | 23 |SM |s/m^2 |Nonlocal transport coefficient |
358 |
|
KPPhbl | 1 |SM |m |KPP boundary layer depth, bulk Ri criterion |
359 |
|
KPPmld | 1 |SM |m |Mixed layer depth, dT=.8degC density criterion |
360 |
|
KPPfrac | 1 |SM | |Short-wave flux fraction penetrating mixing layer |
361 |
|
\end{verbatim} |
362 |
|
} |
363 |
|
\caption{~} |
364 |
|
\end{table} |
365 |
|
|
366 |
|
%---------------------------------------------------------------------- |
367 |
|
|
368 |
|
\subsubsection{Reference experiments} |
369 |
|
|
370 |
|
lab\_sea: |
371 |
|
|
372 |
|
natl\_box: |
373 |
|
|
374 |
|
%---------------------------------------------------------------------- |
375 |
|
|
376 |
|
\subsubsection{References} |