/[MITgcm]/manual/s_overview/text/manual_fromjm.tex
ViewVC logotype

Annotation of /manual/s_overview/text/manual_fromjm.tex

Parent Directory Parent Directory | Revision Log Revision Log | View Revision Graph Revision Graph


Revision 1.2 - (hide annotations) (download) (as text)
Thu Oct 11 19:36:57 2001 UTC (22 years, 7 months ago) by adcroft
Branch: MAIN
Changes since 1.1: +214 -169 lines
File MIME type: application/x-tex
Changes for latex2html.

1 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
2 cnh 1.1 % $Name: $
3    
4     \documentclass[12pt]{book}
5     \usepackage{amsmath}
6     \usepackage{html}
7     \usepackage{epsfig}
8     \usepackage{graphics,subfigure}
9     \usepackage{array}
10     \usepackage{multirow}
11     \usepackage{fancyhdr}
12     \usepackage{psfrag}
13    
14     %TCIDATA{OutputFilter=Latex.dll}
15     %TCIDATA{LastRevised=Thursday, October 04, 2001 14:41:22}
16     %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
17     %TCIDATA{Language=American English}
18    
19     \fancyhead{}
20     \fancyhead[LO]{\slshape \rightmark}
21     \fancyhead[RE]{\slshape \leftmark}
22     \fancyhead[RO,LE]{\thepage}
23     \fancyfoot[CO,CE]{\today}
24     \fancyfoot[RO,LE]{ }
25     \renewcommand{\headrulewidth}{0.4pt}
26     \renewcommand{\footrulewidth}{0.4pt}
27     \setcounter{secnumdepth}{3}
28     \input{tcilatex}
29    
30     \begin{document}
31    
32     \tableofcontents
33    
34    
35 adcroft 1.2 \part{MIT GCM basics}
36 cnh 1.1
37     % Section: Overview
38    
39 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
40 cnh 1.1 % $Name: $
41    
42     \section{Introduction}
43    
44     This documentation provides the reader with the information necessary to
45     carry out numerical experiments using MITgcm. It gives a comprehensive
46     description of the continuous equations on which the model is based, the
47     numerical algorithms the model employs and a description of the associated
48     program code. Along with the hydrodynamical kernel, physical and
49     biogeochemical parameterizations of key atmospheric and oceanic processes
50     are available. A number of examples illustrating the use of the model in
51     both process and general circulation studies of the atmosphere and ocean are
52     also presented.
53    
54     MITgcm has a number of novel aspects:
55    
56     \begin{itemize}
57     \item it can be used to study both atmospheric and oceanic phenomena; one
58     hydrodynamical kernel is used to drive forward both atmospheric and oceanic
59 adcroft 1.2 models - see fig
60 cnh 1.1 \marginpar{
61     Fig.1 One model}\ref{fig:onemodel}
62    
63 adcroft 1.2 %% CNHbegin
64     %notci%\input{part1/one_model_figure}
65     %% CNHend
66    
67 cnh 1.1 \item it has a non-hydrostatic capability and so can be used to study both
68 adcroft 1.2 small-scale and large scale processes - see fig
69 cnh 1.1 \marginpar{
70     Fig.2 All scales}\ref{fig:all-scales}
71    
72 adcroft 1.2 %% CNHbegin
73     %notci%\input{part1/all_scales_figure}
74     %% CNHend
75    
76 cnh 1.1 \item finite volume techniques are employed yielding an intuitive
77     discretization and support for the treatment of irregular geometries using
78 adcroft 1.2 orthogonal curvilinear grids and shaved cells - see fig
79 cnh 1.1 \marginpar{
80 adcroft 1.2 Fig.3 Finite volumes}\ref{fig:finite-volumes}
81    
82     %% CNHbegin
83     %notci%\input{part1/fvol_figure}
84     %% CNHend
85 cnh 1.1
86     \item tangent linear and adjoint counterparts are automatically maintained
87     along with the forward model, permitting sensitivity and optimization
88     studies.
89    
90     \item the model is developed to perform efficiently on a wide variety of
91     computational platforms.
92     \end{itemize}
93    
94     Key publications reporting on and charting the development of the model are
95     listed in an Appendix.
96    
97     We begin by briefly showing some of the results of the model in action to
98     give a feel for the wide range of problems that can be addressed using it.
99     \pagebreak
100    
101 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
102 cnh 1.1 % $Name: $
103    
104     \section{Illustrations of the model in action}
105    
106     The MITgcm has been designed and used to model a wide range of phenomena,
107     from convection on the scale of meters in the ocean to the global pattern of
108     atmospheric winds - see fig.2\ref{fig:all-scales}. To give a flavor of the
109     kinds of problems the model has been used to study, we briefly describe some
110     of them here. A more detailed description of the underlying formulation,
111     numerical algorithm and implementation that lie behind these calculations is
112     given later. Indeed many of the illustrative examples shown below can be
113     easily reproduced: simply download the model (the minimum you need is a PC
114     running linux, together with a FORTRAN\ 77 compiler) and follow the examples
115     described in detail in the documentation.
116    
117     \subsection{Global atmosphere: `Held-Suarez' benchmark}
118    
119     A novel feature of MITgcm is its ability to simulate both atmospheric and
120     oceanographic flows at both small and large scales.
121    
122 adcroft 1.2 Fig.E1a.\ref{fig:eddy_cs} shows an instantaneous plot of the 500$mb$
123 cnh 1.1 temperature field obtained using the atmospheric isomorph of MITgcm run at
124     2.8$^{\circ }$ resolution on the cubed sphere. We see cold air over the pole
125     (blue) and warm air along an equatorial band (red). Fully developed
126     baroclinic eddies spawned in the northern hemisphere storm track are
127     evident. There are no mountains or land-sea contrast in this calculation,
128     but you can easily put them in. The model is driven by relaxation to a
129     radiative-convective equilibrium profile, following the description set out
130     in Held and Suarez; 1994 designed to test atmospheric hydrodynamical cores -
131     there are no mountains or land-sea contrast.
132    
133 adcroft 1.2 %% CNHbegin
134     %notci%\input{part1/cubic_eddies_figure}
135     %% CNHend
136    
137 cnh 1.1 As described in Adcroft (2001), a `cubed sphere' is used to discretize the
138     globe permitting a uniform gridding and obviated the need to fourier filter.
139     The `vector-invariant' form of MITgcm supports any orthogonal curvilinear
140     grid, of which the cubed sphere is just one of many choices.
141    
142     Fig.E1b shows the 5-year mean, zonally averaged potential temperature, zonal
143     wind and meridional overturning streamfunction from a 20-level version of
144     the model. It compares favorable with more conventional spatial
145     discretization approaches.
146    
147     A regular spherical lat-lon grid can also be used.
148    
149 adcroft 1.2 %% CNHbegin
150     %notci%\input{part1/hs_zave_u_figure}
151     %% CNHend
152    
153 cnh 1.1 \subsection{Ocean gyres}
154    
155     Baroclinic instability is a ubiquitous process in the ocean, as well as the
156     atmosphere. Ocean eddies play an important role in modifying the
157     hydrographic structure and current systems of the oceans. Coarse resolution
158     models of the oceans cannot resolve the eddy field and yield rather broad,
159     diffusive patterns of ocean currents. But if the resolution of our models is
160     increased until the baroclinic instability process is resolved, numerical
161     solutions of a different and much more realistic kind, can be obtained.
162    
163     Fig. ?.? shows the surface temperature and velocity field obtained from
164     MITgcm run at $\frac{1}{6}^{\circ }$ horizontal resolution on a $lat-lon$
165     grid in which the pole has been rotated by 90$^{\circ }$ on to the equator
166     (to avoid the converging of meridian in northern latitudes). 21 vertical
167     levels are used in the vertical with a `lopped cell' representation of
168     topography. The development and propagation of anomalously warm and cold
169     eddies can be clearly been seen in the Gulf Stream region. The transport of
170     warm water northward by the mean flow of the Gulf Stream is also clearly
171     visible.
172    
173 adcroft 1.2 %% CNHbegin
174     %notci%\input{part1/ocean_gyres_figure}
175     %% CNHend
176    
177    
178 cnh 1.1 \subsection{Global ocean circulation}
179    
180     Fig.E2a shows the pattern of ocean currents at the surface of a 4$^{\circ }$
181     global ocean model run with 15 vertical levels. Lopped cells are used to
182     represent topography on a regular $lat-lon$ grid extending from 70$^{\circ
183     }N $ to 70$^{\circ }S$. The model is driven using monthly-mean winds with
184     mixed boundary conditions on temperature and salinity at the surface. The
185     transfer properties of ocean eddies, convection and mixing is parameterized
186     in this model.
187    
188     Fig.E2b shows the meridional overturning circulation of the global ocean in
189     Sverdrups.
190    
191 adcroft 1.2 %%CNHbegin
192     %notci%\input{part1/global_circ_figure}
193     %%CNHend
194    
195 cnh 1.1 \subsection{Convection and mixing over topography}
196    
197     Dense plumes generated by localized cooling on the continental shelf of the
198     ocean may be influenced by rotation when the deformation radius is smaller
199     than the width of the cooling region. Rather than gravity plumes, the
200     mechanism for moving dense fluid down the shelf is then through geostrophic
201     eddies. The simulation shown in the figure (blue is cold dense fluid, red is
202     warmer, lighter fluid) employs the non-hydrostatic capability of MITgcm to
203     trigger convection by surface cooling. The cold, dense water falls down the
204     slope but is deflected along the slope by rotation. It is found that
205     entrainment in the vertical plane is reduced when rotational control is
206     strong, and replaced by lateral entrainment due to the baroclinic
207     instability of the along-slope current.
208    
209 adcroft 1.2 %%CNHbegin
210     %notci%\input{part1/convect_and_topo}
211     %%CNHend
212    
213 cnh 1.1 \subsection{Boundary forced internal waves}
214    
215     The unique ability of MITgcm to treat non-hydrostatic dynamics in the
216     presence of complex geometry makes it an ideal tool to study internal wave
217     dynamics and mixing in oceanic canyons and ridges driven by large amplitude
218     barotropic tidal currents imposed through open boundary conditions.
219    
220     Fig. ?.? shows the influence of cross-slope topographic variations on
221     internal wave breaking - the cross-slope velocity is in color, the density
222     contoured. The internal waves are excited by application of open boundary
223     conditions on the left.\ They propagate to the sloping boundary (represented
224     using MITgcm's finite volume spatial discretization) where they break under
225     nonhydrostatic dynamics.
226    
227 adcroft 1.2 %%CNHbegin
228     %notci%\input{part1/boundary_forced_waves}
229     %%CNHend
230    
231 cnh 1.1 \subsection{Parameter sensitivity using the adjoint of MITgcm}
232    
233     Forward and tangent linear counterparts of MITgcm are supported using an
234     `automatic adjoint compiler'. These can be used in parameter sensitivity and
235     data assimilation studies.
236    
237     As one example of application of the MITgcm adjoint, Fig.E4 maps the
238     gradient $\frac{\partial J}{\partial \mathcal{H}}$where $J$ is the magnitude
239 adcroft 1.2 of the overturning streamfunction shown in fig?.? at 40$^{\circ }$N and $
240 cnh 1.1 \mathcal{H}$ is the air-sea heat flux 100 years before. We see that $J$ is
241     sensitive to heat fluxes over the Labrador Sea, one of the important sources
242     of deep water for the thermohaline circulations. This calculation also
243     yields sensitivities to all other model parameters.
244    
245 adcroft 1.2 %%CNHbegin
246     %notci%\input{part1/adj_hf_ocean_figure}
247     %%CNHend
248    
249 cnh 1.1 \subsection{Global state estimation of the ocean}
250    
251     An important application of MITgcm is in state estimation of the global
252     ocean circulation. An appropriately defined `cost function', which measures
253     the departure of the model from observations (both remotely sensed and
254     insitu) over an interval of time, is minimized by adjusting `control
255     parameters' such as air-sea fluxes, the wind field, the initial conditions
256     etc. Figure ?.? shows an estimate of the time-mean surface elevation of the
257     ocean obtained by bringing the model in to consistency with altimetric and
258     in-situ observations over the period 1992-1997.
259    
260 adcroft 1.2 %% CNHbegin
261     %notci%\input{part1/globes_figure}
262     %% CNHend
263    
264 cnh 1.1 \subsection{Ocean biogeochemical cycles}
265    
266     MITgcm is being used to study global biogeochemical cycles in the ocean. For
267     example one can study the effects of interannual changes in meteorological
268     forcing and upper ocean circulation on the fluxes of carbon dioxide and
269     oxygen between the ocean and atmosphere. The figure shows the annual air-sea
270     flux of oxygen and its relation to density outcrops in the southern oceans
271     from a single year of a global, interannually varying simulation.
272    
273 adcroft 1.2 %%CNHbegin
274     %notci%\input{part1/biogeo_figure}
275     %%CNHend
276 cnh 1.1
277     \subsection{Simulations of laboratory experiments}
278    
279     Figure ?.? shows MITgcm being used to simulate a laboratory experiment
280     enquiring in to the dynamics of the Antarctic Circumpolar Current (ACC). An
281     initially homogeneous tank of water ($1m$ in diameter) is driven from its
282     free surface by a rotating heated disk. The combined action of mechanical
283     and thermal forcing creates a lens of fluid which becomes baroclinically
284     unstable. The stratification and depth of penetration of the lens is
285     arrested by its instability in a process analogous to that whic sets the
286     stratification of the ACC.
287    
288 adcroft 1.2 %%CNHbegin
289     %notci%\input{part1/lab_figure}
290     %%CNHend
291    
292     % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
293 cnh 1.1 % $Name: $
294    
295     \section{Continuous equations in `r' coordinates}
296    
297     To render atmosphere and ocean models from one dynamical core we exploit
298     `isomorphisms' between equation sets that govern the evolution of the
299 adcroft 1.2 respective fluids - see fig.4
300 cnh 1.1 \marginpar{
301     Fig.4. Isomorphisms}. One system of hydrodynamical equations is written down
302     and encoded. The model variables have different interpretations depending on
303     whether the atmosphere or ocean is being studied. Thus, for example, the
304     vertical coordinate `$r$' is interpreted as pressure, $p$, if we are
305     modeling the atmosphere and height, $z$, if we are modeling the ocean.
306    
307 adcroft 1.2 %%CNHbegin
308     %notci%\input{part1/zandpcoord_figure.tex}
309     %%CNHend
310    
311 cnh 1.1 The state of the fluid at any time is characterized by the distribution of
312     velocity $\vec{\mathbf{v}}$, active tracers $\theta $ and $S$, a
313     `geopotential' $\phi $ and density $\rho =\rho (\theta ,S,p)$ which may
314     depend on $\theta $, $S$, and $p$. The equations that govern the evolution
315     of these fields, obtained by applying the laws of classical mechanics and
316     thermodynamics to a Boussinesq, Navier-Stokes fluid are, written in terms of
317 adcroft 1.2 a generic vertical coordinate, $r$, see fig.5
318 cnh 1.1 \marginpar{
319     Fig.5 The vertical coordinate of model}:
320    
321 adcroft 1.2 %%CNHbegin
322     %notci%\input{part1/vertcoord_figure.tex}
323     %%CNHend
324    
325 cnh 1.1 \begin{equation*}
326 adcroft 1.2 \frac{D\vec{\mathbf{v}_{h}}}{Dt}+\left( 2\vec{\Omega}\times \vec{\mathbf{v}}
327     \right) _{h}+\mathbf{\nabla }_{h}\phi =\mathcal{F}_{\vec{\mathbf{v}_{h}}}
328 cnh 1.1 \text{ horizontal mtm}
329     \end{equation*}
330    
331     \begin{equation*}
332 adcroft 1.2 \frac{D\dot{r}}{Dt}+\widehat{k}\cdot \left( 2\vec{\Omega}\times \vec{\mathbf{
333 cnh 1.1 v}}\right) +\frac{\partial \phi }{\partial r}+b=\mathcal{F}_{\dot{r}}\text{
334     vertical mtm}
335     \end{equation*}
336    
337     \begin{equation}
338 adcroft 1.2 \mathbf{\nabla }_{h}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial \dot{r}}{
339 cnh 1.1 \partial r}=0\text{ continuity} \label{eq:continuous}
340     \end{equation}
341    
342     \begin{equation*}
343     b=b(\theta ,S,r)\text{ equation of state}
344     \end{equation*}
345    
346     \begin{equation*}
347     \frac{D\theta }{Dt}=\mathcal{Q}_{\theta }\text{ potential temperature}
348     \end{equation*}
349    
350     \begin{equation*}
351     \frac{DS}{Dt}=\mathcal{Q}_{S}\text{ humidity/salinity}
352     \end{equation*}
353    
354     Here:
355    
356     \begin{equation*}
357     r\text{ is the vertical coordinate}
358     \end{equation*}
359    
360     \begin{equation*}
361     \frac{D}{Dt}=\frac{\partial }{\partial t}+\vec{\mathbf{v}}\cdot \nabla \text{
362     is the total derivative}
363     \end{equation*}
364    
365     \begin{equation*}
366 adcroft 1.2 \mathbf{\nabla }=\mathbf{\nabla }_{h}+\widehat{k}\frac{\partial }{\partial r}
367 cnh 1.1 \text{ is the `grad' operator}
368     \end{equation*}
369 adcroft 1.2 with $\mathbf{\nabla }_{h}$ operating in the horizontal and $\widehat{k}
370 cnh 1.1 \frac{\partial }{\partial r}$ operating in the vertical, where $\widehat{k}$
371     is a unit vector in the vertical
372    
373     \begin{equation*}
374     t\text{ is time}
375     \end{equation*}
376    
377     \begin{equation*}
378     \vec{\mathbf{v}}=(u,v,\dot{r})=(\vec{\mathbf{v}}_{h},\dot{r})\text{ is the
379     velocity}
380     \end{equation*}
381    
382     \begin{equation*}
383     \phi \text{ is the `pressure'/`geopotential'}
384     \end{equation*}
385    
386     \begin{equation*}
387     \vec{\Omega}\text{ is the Earth's rotation}
388     \end{equation*}
389    
390     \begin{equation*}
391     b\text{ is the `buoyancy'}
392     \end{equation*}
393    
394     \begin{equation*}
395     \theta \text{ is potential temperature}
396     \end{equation*}
397    
398     \begin{equation*}
399     S\text{ is specific humidity in the atmosphere; salinity in the ocean}
400     \end{equation*}
401    
402     \begin{equation*}
403 adcroft 1.2 \mathcal{F}_{\vec{\mathbf{v}}}\text{ are forcing and dissipation of }\vec{
404 cnh 1.1 \mathbf{v}}
405     \end{equation*}
406    
407     \begin{equation*}
408     \mathcal{Q}_{\theta }\mathcal{\ }\text{are forcing and dissipation of }\theta
409     \end{equation*}
410    
411     \begin{equation*}
412     \mathcal{Q}_{S}\mathcal{\ }\text{are forcing and dissipation of }S
413     \end{equation*}
414    
415     The $\mathcal{F}^{\prime }s$ and $\mathcal{Q}^{\prime }s$ are provided by
416     extensive `physics' packages for atmosphere and ocean described in Chapter 6.
417    
418     \subsection{Kinematic Boundary conditions}
419    
420     \subsubsection{vertical}
421    
422     at fixed and moving $r$ surfaces we set (see fig.5):
423    
424     \begin{equation}
425     \dot{r}=0atr=R_{fixed}(x,y)\text{ (ocean bottom, top of the atmosphere)}
426     \label{eq:fixedbc}
427     \end{equation}
428    
429     \begin{equation}
430     \dot{r}=\frac{Dr}{Dt}atr=R_{moving}\text{ \
431     (oceansurface,bottomoftheatmosphere)} \label{eq:movingbc}
432     \end{equation}
433    
434     Here
435    
436     \begin{equation*}
437     R_{moving}=R_{o}+\eta
438     \end{equation*}
439     where $R_{o}(x,y)$ is the `$r-$value' (height or pressure, depending on
440     whether we are in the atmosphere or ocean) of the `moving surface' in the
441     resting fluid and $\eta $ is the departure from $R_{o}(x,y)$ in the presence
442     of motion.
443    
444     \subsubsection{horizontal}
445    
446     \begin{equation}
447     \vec{\mathbf{v}}\cdot \vec{\mathbf{n}}=0 \label{eq:noflow}
448 adcroft 1.2 \end{equation}
449 cnh 1.1 where $\vec{\mathbf{n}}$ is the normal to a solid boundary.
450    
451     \subsection{Atmosphere}
452    
453     In the atmosphere, see fig.5, we interpret:
454    
455     \begin{equation}
456     r=p\text{ is the pressure} \label{eq:atmos-r}
457     \end{equation}
458    
459     \begin{equation}
460     \dot{r}=\frac{Dp}{Dt}=\omega \text{ is the vertical velocity in }p\text{
461     coordinates} \label{eq:atmos-omega}
462     \end{equation}
463    
464     \begin{equation}
465     \phi =g\,z\text{ is the geopotential height} \label{eq:atmos-phi}
466     \end{equation}
467    
468     \begin{equation}
469     b=\frac{\partial \Pi }{\partial p}\theta \text{ is the buoyancy}
470     \label{eq:atmos-b}
471     \end{equation}
472    
473     \begin{equation}
474     \theta =T(\frac{p_{c}}{p})^{\kappa }\text{ is potential temperature}
475     \label{eq:atmos-theta}
476     \end{equation}
477    
478     \begin{equation}
479     S=q,\text{ is the specific humidity} \label{eq:atmos-s}
480     \end{equation}
481     where
482    
483     \begin{equation*}
484     T\text{ is absolute temperature}
485 adcroft 1.2 \end{equation*}
486 cnh 1.1 \begin{equation*}
487     p\text{ is the pressure}
488 adcroft 1.2 \end{equation*}
489 cnh 1.1 \begin{eqnarray*}
490     &&z\text{ is the height of the pressure surface} \\
491     &&g\text{ is the acceleration due to gravity}
492     \end{eqnarray*}
493    
494     In the above the ideal gas law, $p=\rho RT$, has been expressed in terms of
495     the Exner function $\Pi (p)$ given by (see Appendix Atmosphere)
496     \begin{equation}
497     \Pi (p)=c_{p}(\frac{p}{p_{c}})^{\kappa } \label{eq:exner}
498 adcroft 1.2 \end{equation}
499 cnh 1.1 where $p_{c}$ is a reference pressure and $\kappa =R/c_{p}$ with $R$ the gas
500     constant and $c_{p}$ the specific heat of air at constant pressure.
501    
502     At the top of the atmosphere (which is `fixed' in our $r$ coordinate):
503    
504     \begin{equation*}
505     R_{fixed}=p_{top}=0
506     \end{equation*}
507     In a resting atmosphere the elevation of the mountains at the bottom is
508     given by
509     \begin{equation*}
510     R_{moving}=R_{o}(x,y)=p_{o}(x,y)
511     \end{equation*}
512     i.e. the (hydrostatic) pressure at the top of the mountains in a resting
513     atmosphere.
514    
515     The boundary conditions at top and bottom are given by:
516    
517     \begin{eqnarray}
518     &&\omega =0~\text{at }r=R_{fixed} \text{ (top of the atmosphere)}
519     \label{eq:fixed-bc-atmos} \\
520     \omega &=&\frac{Dp_{s}}{Dt}\text{; at }r=R_{moving}\text{ (bottom of the
521     atmosphere)} \label{eq:moving-bc-atmos}
522     \end{eqnarray}
523    
524     Then the (hydrostatic form of) eq(\ref{eq:continuous}) yields a consistent
525     set of atmospheric equations which, for convenience, are written out in $p$
526     coordinates in Appendix Atmosphere - see eqs(\ref{eq:atmos-prime}).
527    
528     \subsection{Ocean}
529    
530     In the ocean we interpret:
531     \begin{eqnarray}
532     r &=&z\text{ is the height} \label{eq:ocean-z} \\
533     \dot{r} &=&\frac{Dz}{Dt}=w\text{ is the vertical velocity}
534     \label{eq:ocean-w} \\
535     \phi &=&\frac{p}{\rho _{c}}\text{ is the pressure} \label{eq:ocean-p} \\
536     b(\theta ,S,r) &=&\frac{g}{\rho _{c}}\left( \rho (\theta ,S,r)-\rho
537     _{c}\right) \text{ is the buoyancy} \label{eq:ocean-b}
538     \end{eqnarray}
539     where $\rho _{c}$ is a fixed reference density of water and $g$ is the
540     acceleration due to gravity.\noindent
541    
542     In the above
543    
544     At the bottom of the ocean: $R_{fixed}(x,y)=-H(x,y)$.
545    
546     The surface of the ocean is given by: $R_{moving}=\eta $
547    
548 adcroft 1.2 The position of the resting free surface of the ocean is given by $
549 cnh 1.1 R_{o}=Z_{o}=0$.
550    
551     Boundary conditions are:
552    
553     \begin{eqnarray}
554     w &=&0~\text{at }r=R_{fixed}\text{ (ocean bottom)} \label{eq:fixed-bc-ocean}
555     \\
556 adcroft 1.2 w &=&\frac{D\eta }{Dt}\text{ at }r=R_{moving}=\eta \text{ (ocean surface)
557 cnh 1.1 \label{eq:moving-bc-ocean}}
558     \end{eqnarray}
559     where $\eta $ is the elevation of the free surface.
560    
561     Then eq(\ref{eq:continuous}) yields a consistent set of oceanic equations
562     which, for convenience, are written out in $z$ coordinates in Appendix Ocean
563     - see eqs(\ref{eq:ocean-mom}) to (\ref{eq:ocean-salt}).
564    
565     \subsection{Hydrostatic, Quasi-hydrostatic, Quasi-nonhydrostatic and
566     Non-hydrostatic forms}
567    
568     Let us separate $\phi $ in to surface, hydrostatic and non-hydrostatic terms:
569    
570     \begin{equation}
571     \phi (x,y,r)=\phi _{s}(x,y)+\phi _{hyd}(x,y,r)+\phi _{nh}(x,y,r)
572     \label{eq:phi-split}
573 adcroft 1.2 \end{equation}
574 cnh 1.1 and write eq(\ref{incompressible}a,b) in the form:
575    
576     \begin{equation}
577     \frac{\partial \vec{\mathbf{v}_{h}}}{\partial t}+\mathbf{\nabla }_{h}\phi
578     _{s}+\mathbf{\nabla }_{h}\phi _{hyd}+\epsilon _{nh}\mathbf{\nabla }_{h}\phi
579     _{nh}=\vec{\mathbf{G}}_{\vec{v}_{h}} \label{eq:mom-h}
580     \end{equation}
581    
582     \begin{equation}
583     \frac{\partial \phi _{hyd}}{\partial r}=-b \label{eq:hydrostatic}
584     \end{equation}
585    
586     \begin{equation}
587 adcroft 1.2 \epsilon _{nh}\frac{\partial \dot{r}}{\partial t}+\frac{\partial \phi _{nh}}{
588 cnh 1.1 \partial r}=G_{\dot{r}} \label{eq:mom-w}
589     \end{equation}
590     Here $\epsilon _{nh}$ is a non-hydrostatic parameter.
591    
592 adcroft 1.2 The $\left( \vec{\mathbf{G}}_{\vec{v}},G_{\dot{r}}\right) $ in eq(\ref
593 cnh 1.1 {eq:mom-h}) and (\ref{eq:mom-w}) represent advective, metric and Coriolis
594 adcroft 1.2 terms in the momentum equations. In spherical coordinates they take the form
595     \footnote{
596 cnh 1.1 In the hydrostatic primitive equations (\textbf{HPE}) all underlined terms
597 adcroft 1.2 in (\ref{eq:gu-speherical}), (\ref{eq:gv-spherical}) and (\ref
598 cnh 1.1 {eq:gw-spherical}) are omitted; the singly-underlined terms are included in
599 adcroft 1.2 the quasi-hydrostatic model (\textbf{QH}). The fully non-hydrostatic model (
600 cnh 1.1 \textbf{NH}) includes all terms.} - see Marshall et al 1997a for a full
601     discussion:
602    
603     \begin{equation}
604     \left.
605     \begin{tabular}{l}
606     $G_{u}=-\vec{\mathbf{v}}.\nabla u$ \\
607     $-\left\{ \underline{\frac{u\dot{r}}{{r}}}-\frac{uv\tan lat}{{r}}\right\} $
608     \\
609     $-\left\{ -2\Omega v\sin lat+\underline{2\Omega \dot{r}\cos lat}\right\} $
610     \\
611 adcroft 1.2 $+\mathcal{F}_{u}$
612     \end{tabular}
613 cnh 1.1 \ \right\} \left\{
614     \begin{tabular}{l}
615     \textit{advection} \\
616     \textit{metric} \\
617     \textit{Coriolis} \\
618 adcroft 1.2 \textit{\ Forcing/Dissipation}
619     \end{tabular}
620 cnh 1.1 \ \right. \qquad \label{eq:gu-speherical}
621     \end{equation}
622    
623     \begin{equation}
624     \left.
625     \begin{tabular}{l}
626     $G_{v}=-\vec{\mathbf{v}}.\nabla v$ \\
627     $-\left\{ \underline{\frac{v\dot{r}}{{r}}}-\frac{u^{2}\tan lat}{{r}}\right\}
628     $ \\
629     $-\left\{ -2\Omega u\sin lat\right\} $ \\
630 adcroft 1.2 $+\mathcal{F}_{v}$
631     \end{tabular}
632 cnh 1.1 \ \right\} \left\{
633     \begin{tabular}{l}
634     \textit{advection} \\
635     \textit{metric} \\
636     \textit{Coriolis} \\
637 adcroft 1.2 \textit{\ Forcing/Dissipation}
638     \end{tabular}
639 cnh 1.1 \ \right. \qquad \label{eq:gv-spherical}
640 adcroft 1.2 \end{equation}
641 cnh 1.1 \qquad \qquad \qquad \qquad \qquad
642    
643     \begin{equation}
644     \left.
645     \begin{tabular}{l}
646     $G_{\dot{r}}=-\underline{\underline{\vec{\mathbf{v}}.\nabla \dot{r}}}$ \\
647     $+\left\{ \underline{\frac{u^{_{^{2}}}+v^{2}}{{r}}}\right\} $ \\
648     ${+}\underline{{2\Omega u\cos lat}}$ \\
649 adcroft 1.2 $\underline{\underline{\mathcal{F}_{\dot{r}}}}$
650     \end{tabular}
651 cnh 1.1 \ \right\} \left\{
652     \begin{tabular}{l}
653     \textit{advection} \\
654     \textit{metric} \\
655     \textit{Coriolis} \\
656 adcroft 1.2 \textit{\ Forcing/Dissipation}
657     \end{tabular}
658 cnh 1.1 \ \right. \label{eq:gw-spherical}
659 adcroft 1.2 \end{equation}
660 cnh 1.1 \qquad \qquad \qquad \qquad \qquad
661    
662 adcroft 1.2 In the above `${r}$' is the distance from the center of the earth and `$lat$
663 cnh 1.1 ' is latitude.
664    
665     Grad and div operators in spherical coordinates are defined in appendix
666 adcroft 1.2 OPERATORS.
667 cnh 1.1 \marginpar{
668     Fig.6 Spherical polar coordinate system.}
669    
670 adcroft 1.2 %%CNHbegin
671     %notci%\input{part1/sphere_coord_figure.tex}
672     %%CNHend
673    
674 cnh 1.1 \subsubsection{Shallow atmosphere approximation}
675    
676     Most models are based on the `hydrostatic primitive equations' (HPE's) in
677     which the vertical momentum equation is reduced to a statement of
678     hydrostatic balance and the `traditional approximation' is made in which the
679     Coriolis force is treated approximately and the shallow atmosphere
680     approximation is made.\ The MITgcm need not make the `traditional
681     approximation'. To be able to support consistent non-hydrostatic forms the
682 adcroft 1.2 shallow atmosphere approximation can be relaxed - when dividing through by $
683 cnh 1.1 r $ in, for example, (\ref{eq:gu-speherical}), we do not replace $r$ by $a$,
684     the radius of the earth.
685    
686     \subsubsection{Hydrostatic and quasi-hydrostatic forms}
687    
688     These are discussed at length in Marshall et al (1997a).
689    
690     In the `hydrostatic primitive equations' (\textbf{HPE)} all the underlined
691     terms in Eqs. (\ref{eq:gu-speherical} $\rightarrow $\ \ref{eq:gw-spherical})
692     are neglected and `${r}$' is replaced by `$a$', the mean radius of the
693     earth. Once the pressure is found at one level - e.g. by inverting a 2-d
694     Elliptic equation for $\phi _{s}$ at $r=R_{moving}$ - the pressure can be
695 adcroft 1.2 computed at all other levels by integration of the hydrostatic relation, eq(
696 cnh 1.1 \ref{eq:hydrostatic}).
697    
698     In the `quasi-hydrostatic' equations (\textbf{QH)} strict balance between
699     gravity and vertical pressure gradients is not imposed. The $2\Omega u\cos
700     \phi $ Coriolis term are not neglected and are balanced by a non-hydrostatic
701 adcroft 1.2 contribution to the pressure field: only the terms underlined twice in Eqs. (
702 cnh 1.1 \ref{eq:gu-speherical}$\rightarrow $\ \ref{eq:gw-spherical}) are set to zero
703     and, simultaneously, the shallow atmosphere approximation is relaxed. In
704     \textbf{QH}\ \textit{all} the metric terms are retained and the full
705     variation of the radial position of a particle monitored. The \textbf{QH}\
706     vertical momentum equation (\ref{eq:mom-w}) becomes:
707    
708     \begin{equation*}
709     \frac{\partial \phi _{nh}}{\partial r}=2\Omega u\cos lat
710     \end{equation*}
711     making a small correction to the hydrostatic pressure.
712    
713     \textbf{QH} has good energetic credentials - they are the same as for
714     \textbf{HPE}. Importantly, however, it has the same angular momentum
715     principle as the full non-hydrostatic model (\textbf{NH)} - see Marshall
716     et.al., 1997a. As in \textbf{HPE }only a 2-d elliptic problem need be solved.
717    
718     \subsubsection{Non-hydrostatic and quasi-nonhydrostatic forms}
719    
720     The MIT model presently supports a full non-hydrostatic ocean isomorph, but
721     only a quasi-non-hydrostatic atmospheric isomorph.
722    
723     \paragraph{Non-hydrostatic Ocean}
724    
725 adcroft 1.2 In the non-hydrostatic ocean model all terms in equations Eqs.(\ref
726 cnh 1.1 {eq:gu-speherical} $\rightarrow $\ \ref{eq:gw-spherical}) are retained. A
727     three dimensional elliptic equation must be solved subject to Neumann
728     boundary conditions (see below). It is important to note that use of the
729     full \textbf{NH} does not admit any new `fast' waves in to the system - the
730     incompressible condition eq(\ref{eq:continuous})c has already filtered out
731     acoustic modes. It does, however, ensure that the gravity waves are treated
732     accurately with an exact dispersion relation. The \textbf{NH} set has a
733     complete angular momentum principle and consistent energetics - see White
734     and Bromley, 1995; Marshall et.al.\ 1997a.
735    
736     \paragraph{Quasi-nonhydrostatic Atmosphere}
737    
738 adcroft 1.2 In the non-hydrostatic version of our atmospheric model we approximate $\dot{
739 cnh 1.1 r}$ in the vertical momentum eqs(\ref{eq:mom-w}) and (\ref{eq:gv-spherical})
740     (but only here) by:
741    
742     \begin{equation}
743     \dot{r}=\frac{Dp}{Dt}=\frac{1}{g}\frac{D\phi }{Dt} \label{eq:quasi-nh-w}
744 adcroft 1.2 \end{equation}
745 cnh 1.1 where $p_{hy}$ is the hydrostatic pressure.
746    
747     \subsubsection{Summary of equation sets supported by model}
748    
749     \paragraph{Atmosphere}
750    
751     Hydrostatic, and quasi-hydrostatic and quasi non-hydrostatic forms of the
752     compressible non-Boussinesq equations in $p-$coordinates are supported.
753    
754     \subparagraph{Hydrostatic and quasi-hydrostatic}
755    
756     The hydrostatic set is written out in $p-$coordinates in appendix Atmosphere
757     - see eq(\ref{eq:atmos-prime}).
758    
759     \subparagraph{Quasi-nonhydrostatic}
760    
761     A quasi-nonhydrostatic form is also supported.
762    
763     \paragraph{Ocean}
764    
765     \subparagraph{Hydrostatic and quasi-hydrostatic}
766    
767     Hydrostatic, and quasi-hydrostatic forms of the incompressible Boussinesq
768     equations in $z-$coordinates are supported.
769    
770     \subparagraph{Non-hydrostatic}
771    
772 adcroft 1.2 Non-hydrostatic forms of the incompressible Boussinesq equations in $z-$
773     coordinates are supported - see eqs(\ref{eq:ocean-mom}) to (\ref
774 cnh 1.1 {eq:ocean-salt}).
775    
776     \subsection{Solution strategy}
777    
778 adcroft 1.2 The method of solution employed in the \textbf{HPE}, \textbf{QH} and \textbf{
779     NH} models is summarized in Fig.7.
780 cnh 1.1 \marginpar{
781     Fig.7 Solution strategy} Under all dynamics, a 2-d elliptic equation is
782     first solved to find the surface pressure and the hydrostatic pressure at
783     any level computed from the weight of fluid above. Under \textbf{HPE} and
784     \textbf{QH} dynamics, the horizontal momentum equations are then stepped
785     forward and $\dot{r}$ found from continuity. Under \textbf{NH} dynamics a
786     3-d elliptic equation must be solved for the non-hydrostatic pressure before
787     stepping forward the horizontal momentum equations; $\dot{r}$ is found by
788     stepping forward the vertical momentum equation.
789    
790 adcroft 1.2 %%CNHbegin
791     %notci%\input{part1/solution_strategy_figure.tex}
792     %%CNHend
793    
794 cnh 1.1 There is no penalty in implementing \textbf{QH} over \textbf{HPE} except, of
795 adcroft 1.2 course, some complication that goes with the inclusion of $\cos \phi \ $
796 cnh 1.1 Coriolis terms and the relaxation of the shallow atmosphere approximation.
797     But this leads to negligible increase in computation. In \textbf{NH}, in
798     contrast, one additional elliptic equation - a three-dimensional one - must
799     be inverted for $p_{nh}$. However the `overhead' of the \textbf{NH} model is
800     essentially negligible in the hydrostatic limit (see detailed discussion in
801     Marshall et al, 1997) resulting in a non-hydrostatic algorithm that, in the
802     hydrostatic limit, is as computationally economic as the \textbf{HPEs}.
803    
804     \subsection{Finding the pressure field}
805    
806     Unlike the prognostic variables $u$, $v$, $w$, $\theta $ and $S$, the
807     pressure field must be obtained diagnostically. We proceed, as before, by
808     dividing the total (pressure/geo) potential in to three parts, a surface
809     part, $\phi _{s}(x,y)$, a hydrostatic part $\phi _{hyd}(x,y,r)$ and a
810     non-hydrostatic part $\phi _{nh}(x,y,r)$, as in (\ref{eq:phi-split}), and
811     writing the momentum equation as in (\ref{eq:mom-h}).
812    
813     \subsubsection{Hydrostatic pressure}
814    
815     Hydrostatic pressure is obtained by integrating (\ref{eq:hydrostatic})
816     vertically from $r=R_{o}$ where $\phi _{hyd}(r=R_{o})=0$, to yield:
817    
818     \begin{equation*}
819 adcroft 1.2 \int_{r}^{R_{o}}\frac{\partial \phi _{hyd}}{\partial r}dr=\left[ \phi _{hyd}
820 cnh 1.1 \right] _{r}^{R_{o}}=\int_{r}^{R_{o}}-bdr
821     \end{equation*}
822     and so
823    
824     \begin{equation}
825     \phi _{hyd}(x,y,r)=\int_{r}^{R_{o}}bdr \label{eq:hydro-phi}
826     \end{equation}
827    
828     The model can be easily modified to accommodate a loading term (e.g
829     atmospheric pressure pushing down on the ocean's surface) by setting:
830    
831     \begin{equation}
832     \phi _{hyd}(r=R_{o})=loading \label{eq:loading}
833     \end{equation}
834    
835     \subsubsection{Surface pressure}
836    
837 adcroft 1.2 The surface pressure equation can be obtained by integrating continuity, (
838 cnh 1.1 \ref{eq:continuous})c, vertically from $r=R_{fixed}$ to $r=R_{moving}$
839    
840     \begin{equation*}
841 adcroft 1.2 \int_{R_{fixed}}^{R_{moving}}\left( \mathbf{\nabla }_{h}\cdot \vec{\mathbf{v}
842 cnh 1.1 }_{h}+\partial _{r}\dot{r}\right) dr=0
843     \end{equation*}
844    
845     Thus:
846    
847     \begin{equation*}
848     \frac{\partial \eta }{\partial t}+\vec{\mathbf{v}}.\nabla \eta
849 adcroft 1.2 +\int_{R_{fixed}}^{R_{moving}}\mathbf{\nabla }_{h}\cdot \vec{\mathbf{v}}
850 cnh 1.1 _{h}dr=0
851     \end{equation*}
852 adcroft 1.2 where $\eta =R_{moving}-R_{o}$ is the free-surface $r$-anomaly in units of $
853 cnh 1.1 r $. The above can be rearranged to yield, using Leibnitz's theorem:
854    
855     \begin{equation}
856     \frac{\partial \eta }{\partial t}+\mathbf{\nabla }_{h}\cdot
857     \int_{R_{fixed}}^{R_{moving}}\vec{\mathbf{v}}_{h}dr=\text{source}
858     \label{eq:free-surface}
859 adcroft 1.2 \end{equation}
860 cnh 1.1 where we have incorporated a source term.
861    
862     Whether $\phi $ is pressure (ocean model, $p/\rho _{c}$) or geopotential
863     (atmospheric model), in (\ref{mtm-split}), the horizontal gradient term can
864     be written
865     \begin{equation}
866     \mathbf{\nabla }_{h}\phi _{s}=\mathbf{\nabla }_{h}\left( b_{s}\eta \right)
867     \label{eq:phi-surf}
868 adcroft 1.2 \end{equation}
869 cnh 1.1 where $b_{s}$ is the buoyancy at the surface.
870    
871 adcroft 1.2 In the hydrostatic limit ($\epsilon _{nh}=0$), Eqs(\ref{eq:mom-h}), (\ref
872 cnh 1.1 {eq:free-surface}) and (\ref{eq:phi-surf}) can be solved by inverting a 2-d
873     elliptic equation for $\phi _{s}$ as described in Chapter 2. Both `free
874     surface' and `rigid lid' approaches are available.
875    
876     \subsubsection{Non-hydrostatic pressure}
877    
878 adcroft 1.2 Taking the horizontal divergence of (\ref{hor-mtm}) and adding $\frac{
879 cnh 1.1 \partial }{\partial r}$ of (\ref{vertmtm}), invoking the continuity equation
880     (\ref{incompressible}), we deduce that:
881    
882     \begin{equation}
883 adcroft 1.2 \nabla _{3}^{2}\phi _{nh}=\nabla .\vec{\mathbf{G}}_{\vec{v}}-\left( \mathbf{
884     \nabla }_{h}^{2}\phi _{s}+\mathbf{\nabla }^{2}\phi _{hyd}\right) =\nabla .
885 cnh 1.1 \vec{\mathbf{F}} \label{eq:3d-invert}
886     \end{equation}
887    
888     For a given rhs this 3-d elliptic equation must be inverted for $\phi _{nh}$
889     subject to appropriate choice of boundary conditions. This method is usually
890     called \textit{The Pressure Method} [Harlow and Welch, 1965; Williams, 1969;
891     Potter, 1976]. In the hydrostatic primitive equations case (\textbf{HPE}),
892     the 3-d problem does not need to be solved.
893    
894     \paragraph{Boundary Conditions}
895    
896     We apply the condition of no normal flow through all solid boundaries - the
897     coasts (in the ocean) and the bottom:
898    
899     \begin{equation}
900     \vec{\mathbf{v}}.\widehat{n}=0 \label{nonormalflow}
901     \end{equation}
902     where $\widehat{n}$ is a vector of unit length normal to the boundary. The
903     kinematic condition (\ref{nonormalflow}) is also applied to the vertical
904 adcroft 1.2 velocity at $r=R_{moving}$. No-slip $\left( v_{T}=0\right) \ $or slip $
905 cnh 1.1 \left( \partial v_{T}/\partial n=0\right) \ $conditions are employed on the
906     tangential component of velocity, $v_{T}$, at all solid boundaries,
907     depending on the form chosen for the dissipative terms in the momentum
908     equations - see below.
909    
910     Eq.(\ref{nonormalflow}) implies, making use of (\ref{mtm-split}), that:
911    
912     \begin{equation}
913     \widehat{n}.\nabla \phi _{nh}=\widehat{n}.\vec{\mathbf{F}}
914     \label{eq:inhom-neumann-nh}
915     \end{equation}
916     where
917    
918     \begin{equation*}
919     \vec{\mathbf{F}}=\vec{\mathbf{G}}_{\vec{v}}-\left( \mathbf{\nabla }_{h}\phi
920     _{s}+\mathbf{\nabla }\phi _{hyd}\right)
921 adcroft 1.2 \end{equation*}
922 cnh 1.1 presenting inhomogeneous Neumann boundary conditions to the Elliptic problem
923     (\ref{eq:3d-invert}). As shown, for example, by Williams (1969), one can
924     exploit classical 3D potential theory and, by introducing an appropriately
925     chosen $\delta $-function sheet of `source-charge', replace the
926     inhomogeneous boundary condition on pressure by a homogeneous one. The
927 adcroft 1.2 source term $rhs$ in (\ref{eq:3d-invert}) is the divergence of the vector $
928     \vec{\mathbf{F}}.$ By simultaneously setting $
929 cnh 1.1 \begin{array}{l}
930 adcroft 1.2 \widehat{n}.\vec{\mathbf{F}}
931     \end{array}
932 cnh 1.1 =0$\ and $\widehat{n}.\nabla \phi _{nh}=0\ $on the boundary the following
933     self-consistent but simpler homogenized Elliptic problem is obtained:
934    
935     \begin{equation*}
936     \nabla ^{2}\phi _{nh}=\nabla .\widetilde{\vec{\mathbf{F}}}\qquad
937 adcroft 1.2 \end{equation*}
938 cnh 1.1 where $\widetilde{\vec{\mathbf{F}}}$ is a modified $\vec{\mathbf{F}}$ such
939 adcroft 1.2 that $\widetilde{\vec{\mathbf{F}}}.\widehat{n}=0$. As is implied by (\ref
940 cnh 1.1 {eq:inhom-neumann-nh}) the modified boundary condition becomes:
941    
942     \begin{equation}
943     \widehat{n}.\nabla \phi _{nh}=0 \label{eq:hom-neumann-nh}
944     \end{equation}
945    
946     If the flow is `close' to hydrostatic balance then the 3-d inversion
947     converges rapidly because $\phi _{nh}\ $is then only a small correction to
948     the hydrostatic pressure field (see the discussion in Marshall et al, a,b).
949    
950     The solution $\phi _{nh}\ $to (\ref{eq:3d-invert}) and (\ref{homneuman})
951     does not vanish at $r=R_{moving}$, and so refines the pressure there.
952    
953     \subsection{Forcing/dissipation}
954    
955     \subsubsection{Forcing}
956    
957     The forcing terms $\mathcal{F}$ on the rhs of the equations are provided by
958     `physics packages' described in detail in chapter ??.
959    
960     \subsubsection{Dissipation}
961    
962     \paragraph{Momentum}
963    
964     Many forms of momentum dissipation are available in the model. Laplacian and
965     biharmonic frictions are commonly used:
966    
967     \begin{equation}
968 adcroft 1.2 D_{V}=A_{h}\nabla _{h}^{2}v+A_{v}\frac{\partial ^{2}v}{\partial z^{2}}
969 cnh 1.1 +A_{4}\nabla _{h}^{4}v \label{eq:dissipation}
970     \end{equation}
971     where $A_{h}$ and $A_{v}\ $are (constant) horizontal and vertical viscosity
972     coefficients and $A_{4}\ $is the horizontal coefficient for biharmonic
973     friction. These coefficients are the same for all velocity components.
974    
975     \paragraph{Tracers}
976    
977     The mixing terms for the temperature and salinity equations have a similar
978     form to that of momentum except that the diffusion tensor can be
979 adcroft 1.2 non-diagonal and have varying coefficients. $\qquad $
980 cnh 1.1 \begin{equation}
981     D_{T,S}=\nabla .[\underline{\underline{K}}\nabla (T,S)]+K_{4}\nabla
982     _{h}^{4}(T,S) \label{eq:diffusion}
983     \end{equation}
984 adcroft 1.2 where $\underline{\underline{K}}\ $is the diffusion tensor and the $K_{4}\ $
985 cnh 1.1 horizontal coefficient for biharmonic diffusion. In the simplest case where
986     the subgrid-scale fluxes of heat and salt are parameterized with constant
987     horizontal and vertical diffusion coefficients, $\underline{\underline{K}}$,
988     reduces to a diagonal matrix with constant coefficients:
989    
990     \begin{equation}
991     \qquad \qquad \qquad \qquad K=\left(
992     \begin{array}{ccc}
993     K_{h} & 0 & 0 \\
994     0 & K_{h} & 0 \\
995 adcroft 1.2 0 & 0 & K_{v}
996 cnh 1.1 \end{array}
997     \right) \qquad \qquad \qquad \label{eq:diagonal-diffusion-tensor}
998     \end{equation}
999     where $K_{h}\ $and $K_{v}\ $are the horizontal and vertical diffusion
1000     coefficients. These coefficients are the same for all tracers (temperature,
1001     salinity ... ).
1002    
1003     \subsection{Vector invariant form}
1004    
1005 adcroft 1.2 For some purposes it is advantageous to write momentum advection in eq(\ref
1006 cnh 1.1 {hor-mtm}) and (\ref{vertmtm}) in the (so-called) `vector invariant' form:
1007    
1008     \begin{equation}
1009 adcroft 1.2 \frac{D\vec{\mathbf{v}}}{Dt}=\frac{\partial \vec{\mathbf{v}}}{\partial t}
1010     +\left( \nabla \times \vec{\mathbf{v}}\right) \times \vec{\mathbf{v}}+\nabla
1011 cnh 1.1 \left[ \frac{1}{2}(\vec{\mathbf{v}}\cdot \vec{\mathbf{v}})\right]
1012     \label{eq:vi-identity}
1013 adcroft 1.2 \end{equation}
1014 cnh 1.1 This permits alternative numerical treatments of the non-linear terms based
1015     on their representation as a vorticity flux. Because gradients of coordinate
1016     vectors no longer appear on the rhs of (\ref{eq:vi-identity}), explicit
1017 adcroft 1.2 representation of the metric terms in (\ref{eq:gu-speherical}), (\ref
1018 cnh 1.1 {eq:gv-spherical}) and (\ref{eq:gw-spherical}), can be avoided: information
1019     about the geometry is contained in the areas and lengths of the volumes used
1020     to discretize the model.
1021    
1022     \subsection{Adjoint}
1023    
1024     Tangent linear and adjoint counterparts of the forward model and described
1025     in Chapter 5.
1026    
1027 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
1028 cnh 1.1 % $Name: $
1029    
1030     \section{Appendix ATMOSPHERE}
1031    
1032     \subsection{Hydrostatic Primitive Equations for the Atmosphere in pressure
1033     coordinates}
1034    
1035     \label{sect-hpe-p}
1036    
1037     The hydrostatic primitive equations (HPEs) in p-coordinates are:
1038     \begin{eqnarray}
1039 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1040 cnh 1.1 _{h}+\mathbf{\nabla }_{p}\phi &=&\vec{\mathbf{\mathcal{F}}}
1041     \label{eq:atmos-mom} \\
1042     \frac{\partial \phi }{\partial p}+\alpha &=&0 \label{eq-p-hydro-start} \\
1043 adcroft 1.2 \mathbf{\nabla }_{p}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial \omega }{
1044 cnh 1.1 \partial p} &=&0 \label{eq:atmos-cont} \\
1045     p\alpha &=&RT \label{eq:atmos-eos} \\
1046     c_{v}\frac{DT}{Dt}+p\frac{D\alpha }{Dt} &=&\mathcal{Q} \label{eq:atmos-heat}
1047 adcroft 1.2 \end{eqnarray}
1048 cnh 1.1 where $\vec{\mathbf{v}}_{h}=(u,v,0)$ is the `horizontal' (on pressure
1049     surfaces) component of velocity,$\frac{D}{Dt}=\vec{\mathbf{v}}_{h}\cdot
1050     \mathbf{\nabla }_{p}+\omega \frac{\partial }{\partial p}$ is the total
1051     derivative, $f=2\Omega \sin lat$ is the Coriolis parameter, $\phi =gz$ is
1052 adcroft 1.2 the geopotential, $\alpha =1/\rho $ is the specific volume, $\omega =\frac{Dp
1053     }{Dt}$ is the vertical velocity in the $p-$coordinate. Equation(\ref
1054     {eq:atmos-heat}) is the first law of thermodynamics where internal energy $
1055     e=c_{v}T$, $T$ is temperature, $Q$ is the rate of heating per unit mass and $
1056 cnh 1.1 p\frac{D\alpha }{Dt}$ is the work done by the fluid in compressing.
1057    
1058     It is convenient to cast the heat equation in terms of potential temperature
1059     $\theta $ so that it looks more like a generic conservation law.
1060     Differentiating (\ref{eq:atmos-eos}) we get:
1061     \begin{equation*}
1062     p\frac{D\alpha }{Dt}+\alpha \frac{Dp}{Dt}=R\frac{DT}{Dt}
1063 adcroft 1.2 \end{equation*}
1064     which, when added to the heat equation (\ref{eq:atmos-heat}) and using $
1065 cnh 1.1 c_{p}=c_{v}+R$, gives:
1066     \begin{equation}
1067     c_{p}\frac{DT}{Dt}-\alpha \frac{Dp}{Dt}=\mathcal{Q}
1068     \label{eq-p-heat-interim}
1069 adcroft 1.2 \end{equation}
1070 cnh 1.1 Potential temperature is defined:
1071     \begin{equation}
1072     \theta =T(\frac{p_{c}}{p})^{\kappa } \label{eq:potential-temp}
1073 adcroft 1.2 \end{equation}
1074 cnh 1.1 where $p_{c}$ is a reference pressure and $\kappa =R/c_{p}$. For convenience
1075     we will make use of the Exner function $\Pi (p)$ which defined by:
1076     \begin{equation}
1077     \Pi (p)=c_{p}(\frac{p}{p_{c}})^{\kappa } \label{Exner}
1078 adcroft 1.2 \end{equation}
1079 cnh 1.1 The following relations will be useful and are easily expressed in terms of
1080     the Exner function:
1081     \begin{equation*}
1082     c_{p}T=\Pi \theta \;\;;\;\;\frac{\partial \Pi }{\partial p}=\frac{\kappa \Pi
1083 adcroft 1.2 }{p}\;\;;\;\;\alpha =\frac{\kappa \Pi \theta }{p}=\frac{\partial \ \Pi }{
1084     \partial p}\theta \;\;;\;\;\frac{D\Pi }{Dt}=\frac{\partial \Pi }{\partial p}
1085 cnh 1.1 \frac{Dp}{Dt}
1086 adcroft 1.2 \end{equation*}
1087 cnh 1.1 where $b=\frac{\partial \ \Pi }{\partial p}\theta $ is the buoyancy.
1088    
1089     The heat equation is obtained by noting that
1090     \begin{equation*}
1091     c_{p}\frac{DT}{Dt}=\frac{D(\Pi \theta )}{Dt}=\Pi \frac{D\theta }{Dt}+\theta
1092     \frac{D\Pi }{Dt}=\Pi \frac{D\theta }{Dt}+\alpha \frac{Dp}{Dt}
1093     \end{equation*}
1094     and on substituting into (\ref{eq-p-heat-interim}) gives:
1095     \begin{equation}
1096     \Pi \frac{D\theta }{Dt}=\mathcal{Q}
1097     \label{eq:potential-temperature-equation}
1098     \end{equation}
1099     which is in conservative form.
1100    
1101 adcroft 1.2 For convenience in the model we prefer to step forward (\ref
1102 cnh 1.1 {eq:potential-temperature-equation}) rather than (\ref{eq:atmos-heat}).
1103    
1104     \subsubsection{Boundary conditions}
1105    
1106     The upper and lower boundary conditions are :
1107     \begin{eqnarray}
1108     \mbox{at the top:}\;\;p=0 &&\text{, }\omega =\frac{Dp}{Dt}=0 \\
1109     \mbox{at the surface:}\;\;p=p_{s} &&\text{, }\phi =\phi _{topo}=g~Z_{topo}
1110     \label{eq:boundary-condition-atmosphere}
1111     \end{eqnarray}
1112     In $p$-coordinates, the upper boundary acts like a solid boundary ($\omega
1113     =0 $); in $z$-coordinates and the lower boundary is analogous to a free
1114     surface ($\phi $ is imposed and $\omega \neq 0$).
1115    
1116     \subsubsection{Splitting the geo-potential}
1117    
1118     For the purposes of initialization and reducing round-off errors, the model
1119     deals with perturbations from reference (or ``standard'') profiles. For
1120     example, the hydrostatic geopotential associated with the resting atmosphere
1121     is not dynamically relevant and can therefore be subtracted from the
1122     equations. The equations written in terms of perturbations are obtained by
1123     substituting the following definitions into the previous model equations:
1124     \begin{eqnarray}
1125     \theta &=&\theta _{o}+\theta ^{\prime } \label{eq:atmos-ref-prof-theta} \\
1126     \alpha &=&\alpha _{o}+\alpha ^{\prime } \label{eq:atmos-ref-prof-alpha} \\
1127     \phi &=&\phi _{o}+\phi ^{\prime } \label{eq:atmos-ref-prof-phi}
1128     \end{eqnarray}
1129     The reference state (indicated by subscript ``0'') corresponds to
1130     horizontally homogeneous atmosphere at rest ($\theta _{o},\alpha _{o},\phi
1131     _{o}$) with surface pressure $p_{o}(x,y)$ that satisfies $\phi
1132     _{o}(p_{o})=g~Z_{topo}$, defined:
1133     \begin{eqnarray*}
1134     \theta _{o}(p) &=&f^{n}(p) \\
1135     \alpha _{o}(p) &=&\Pi _{p}\theta _{o} \\
1136     \phi _{o}(p) &=&\phi _{topo}-\int_{p_{0}}^{p}\alpha _{o}dp
1137     \end{eqnarray*}
1138     %\begin{eqnarray*}
1139     %\phi'_\alpha & = & \int^p_{p_o} (\alpha_o -\alpha) dp \\
1140     %\phi'_s(x,y,t) & = & \int_{p_o}^{p_s} \alpha dp
1141     %\end{eqnarray*}
1142    
1143     The final form of the HPE's in p coordinates is then:
1144     \begin{eqnarray}
1145 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1146 cnh 1.1 _{h}+\mathbf{\nabla }_{p}\phi ^{\prime } &=&\vec{\mathbf{\mathcal{F}}} \\
1147     \frac{\partial \phi ^{\prime }}{\partial p}+\alpha ^{\prime } &=&0 \\
1148 adcroft 1.2 \mathbf{\nabla }_{p}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial \omega }{
1149 cnh 1.1 \partial p} &=&0 \\
1150     \frac{\partial \Pi }{\partial p}\theta ^{\prime } &=&\alpha ^{\prime } \\
1151     \frac{D\theta }{Dt} &=&\frac{\mathcal{Q}}{\Pi } \label{eq:atmos-prime}
1152     \end{eqnarray}
1153    
1154 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
1155 cnh 1.1 % $Name: $
1156    
1157     \section{Appendix OCEAN}
1158    
1159     \subsection{Equations of motion for the ocean}
1160    
1161     We review here the method by which the standard (Boussinesq, incompressible)
1162     HPE's for the ocean written in z-coordinates are obtained. The
1163     non-Boussinesq equations for oceanic motion are:
1164     \begin{eqnarray}
1165 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1166 cnh 1.1 _{h}+\frac{1}{\rho }\mathbf{\nabla }_{z}p &=&\vec{\mathbf{\mathcal{F}}} \\
1167     \epsilon _{nh}\frac{Dw}{Dt}+g+\frac{1}{\rho }\frac{\partial p}{\partial z}
1168     &=&\epsilon _{nh}\mathcal{F}_{w} \\
1169 adcroft 1.2 \frac{1}{\rho }\frac{D\rho }{Dt}+\mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}
1170 cnh 1.1 _{h}+\frac{\partial w}{\partial z} &=&0 \\
1171     \rho &=&\rho (\theta ,S,p) \\
1172     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \\
1173     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq:non-boussinesq}
1174 adcroft 1.2 \end{eqnarray}
1175 cnh 1.1 These equations permit acoustics modes, inertia-gravity waves,
1176     non-hydrostatic motions, a geostrophic (Rossby) mode and a thermo-haline
1177     mode. As written, they cannot be integrated forward consistently - if we
1178     step $\rho $ forward in (\ref{eq-zns-cont}), the answer will not be
1179 adcroft 1.2 consistent with that obtained by stepping (\ref{eq-zns-heat}) and (\ref
1180 cnh 1.1 {eq-zns-salt}) and then using (\ref{eq-zns-eos}) to yield $\rho $. It is
1181     therefore necessary to manipulate the system as follows. Differentiating the
1182     EOS (equation of state) gives:
1183    
1184     \begin{equation}
1185     \frac{D\rho }{Dt}=\left. \frac{\partial \rho }{\partial \theta }\right|
1186     _{S,p}\frac{D\theta }{Dt}+\left. \frac{\partial \rho }{\partial S}\right|
1187     _{\theta ,p}\frac{DS}{Dt}+\left. \frac{\partial \rho }{\partial p}\right|
1188     _{\theta ,S}\frac{Dp}{Dt} \label{EOSexpansion}
1189     \end{equation}
1190    
1191     Note that $\frac{\partial \rho }{\partial p}=\frac{1}{c_{s}^{2}}$ is the
1192 adcroft 1.2 reciprocal of the sound speed ($c_{s}$) squared. Substituting into \ref
1193 cnh 1.1 {eq-zns-cont} gives:
1194     \begin{equation}
1195 adcroft 1.2 \frac{1}{\rho c_{s}^{2}}\frac{Dp}{Dt}+\mathbf{\nabla }_{z}\cdot \vec{\mathbf{
1196 cnh 1.1 v}}+\partial _{z}w\approx 0 \label{eq-zns-pressure}
1197     \end{equation}
1198     where we have used an approximation sign to indicate that we have assumed
1199     adiabatic motion, dropping the $\frac{D\theta }{Dt}$ and $\frac{DS}{Dt}$.
1200     Replacing \ref{eq-zns-cont} with \ref{eq-zns-pressure} yields a system that
1201     can be explicitly integrated forward:
1202     \begin{eqnarray}
1203 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1204 cnh 1.1 _{h}+\frac{1}{\rho }\mathbf{\nabla }_{z}p &=&\vec{\mathbf{\mathcal{F}}}
1205     \label{eq-cns-hmom} \\
1206     \epsilon _{nh}\frac{Dw}{Dt}+g+\frac{1}{\rho }\frac{\partial p}{\partial z}
1207     &=&\epsilon _{nh}\mathcal{F}_{w} \label{eq-cns-hydro} \\
1208 adcroft 1.2 \frac{1}{\rho c_{s}^{2}}\frac{Dp}{Dt}+\mathbf{\nabla }_{z}\cdot \vec{\mathbf{
1209 cnh 1.1 v}}_{h}+\frac{\partial w}{\partial z} &=&0 \label{eq-cns-cont} \\
1210     \rho &=&\rho (\theta ,S,p) \label{eq-cns-eos} \\
1211     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \label{eq-cns-heat} \\
1212     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq-cns-salt}
1213     \end{eqnarray}
1214    
1215     \subsubsection{Compressible z-coordinate equations}
1216    
1217     Here we linearize the acoustic modes by replacing $\rho $ with $\rho _{o}(z)$
1218     wherever it appears in a product (ie. non-linear term) - this is the
1219     `Boussinesq assumption'. The only term that then retains the full variation
1220     in $\rho $ is the gravitational acceleration:
1221     \begin{eqnarray}
1222 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1223 cnh 1.1 _{h}+\frac{1}{\rho _{o}}\mathbf{\nabla }_{z}p &=&\vec{\mathbf{\mathcal{F}}}
1224     \label{eq-zcb-hmom} \\
1225 adcroft 1.2 \epsilon _{nh}\frac{Dw}{Dt}+\frac{g\rho }{\rho _{o}}+\frac{1}{\rho _{o}}
1226 cnh 1.1 \frac{\partial p}{\partial z} &=&\epsilon _{nh}\mathcal{F}_{w}
1227     \label{eq-zcb-hydro} \\
1228 adcroft 1.2 \frac{1}{\rho _{o}c_{s}^{2}}\frac{Dp}{Dt}+\mathbf{\nabla }_{z}\cdot \vec{
1229 cnh 1.1 \mathbf{v}}_{h}+\frac{\partial w}{\partial z} &=&0 \label{eq-zcb-cont} \\
1230     \rho &=&\rho (\theta ,S,p) \label{eq-zcb-eos} \\
1231     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \label{eq-zcb-heat} \\
1232     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq-zcb-salt}
1233     \end{eqnarray}
1234     These equations still retain acoustic modes. But, because the
1235 adcroft 1.2 ``compressible'' terms are linearized, the pressure equation \ref
1236 cnh 1.1 {eq-zcb-cont} can be integrated implicitly with ease (the time-dependent
1237     term appears as a Helmholtz term in the non-hydrostatic pressure equation).
1238     These are the \emph{truly} compressible Boussinesq equations. Note that the
1239     EOS must have the same pressure dependency as the linearized pressure term,
1240 adcroft 1.2 ie. $\left. \frac{\partial \rho }{\partial p}\right| _{\theta ,S}=\frac{1}{
1241 cnh 1.1 c_{s}^{2}}$, for consistency.
1242    
1243     \subsubsection{`Anelastic' z-coordinate equations}
1244    
1245     The anelastic approximation filters the acoustic mode by removing the
1246 adcroft 1.2 time-dependency in the continuity (now pressure-) equation (\ref{eq-zcb-cont}
1247     ). This could be done simply by noting that $\frac{Dp}{Dt}\approx -g\rho _{o}
1248 cnh 1.1 \frac{Dz}{Dt}=-g\rho _{o}w$, but this leads to an inconsistency between
1249     continuity and EOS. A better solution is to change the dependency on
1250     pressure in the EOS by splitting the pressure into a reference function of
1251     height and a perturbation:
1252     \begin{equation*}
1253     \rho =\rho (\theta ,S,p_{o}(z)+\epsilon _{s}p^{\prime })
1254     \end{equation*}
1255     Remembering that the term $\frac{Dp}{Dt}$ in continuity comes from
1256     differentiating the EOS, the continuity equation then becomes:
1257     \begin{equation*}
1258 adcroft 1.2 \frac{1}{\rho _{o}c_{s}^{2}}\left( \frac{Dp_{o}}{Dt}+\epsilon _{s}\frac{
1259     Dp^{\prime }}{Dt}\right) +\mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+
1260 cnh 1.1 \frac{\partial w}{\partial z}=0
1261     \end{equation*}
1262     If the time- and space-scales of the motions of interest are longer than
1263 adcroft 1.2 those of acoustic modes, then $\frac{Dp^{\prime }}{Dt}<<(\frac{Dp_{o}}{Dt},
1264 cnh 1.1 \mathbf{\nabla }\cdot \vec{\mathbf{v}}_{h})$ in the continuity equations and
1265 adcroft 1.2 $\left. \frac{\partial \rho }{\partial p}\right| _{\theta ,S}\frac{
1266 cnh 1.1 Dp^{\prime }}{Dt}<<\left. \frac{\partial \rho }{\partial p}\right| _{\theta
1267     ,S}\frac{Dp_{o}}{Dt}$ in the EOS (\ref{EOSexpansion}). Thus we set $\epsilon
1268     _{s}=0$, removing the dependency on $p^{\prime }$ in the continuity equation
1269     and EOS. Expanding $\frac{Dp_{o}(z)}{Dt}=-g\rho _{o}w$ then leads to the
1270     anelastic continuity equation:
1271     \begin{equation}
1272 adcroft 1.2 \mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial w}{\partial z}-
1273 cnh 1.1 \frac{g}{c_{s}^{2}}w=0 \label{eq-za-cont1}
1274     \end{equation}
1275     A slightly different route leads to the quasi-Boussinesq continuity equation
1276 adcroft 1.2 where we use the scaling $\frac{\partial \rho ^{\prime }}{\partial t}+
1277     \mathbf{\nabla }_{3}\cdot \rho ^{\prime }\vec{\mathbf{v}}<<\mathbf{\nabla }
1278 cnh 1.1 _{3}\cdot \rho _{o}\vec{\mathbf{v}}$ yielding:
1279     \begin{equation}
1280 adcroft 1.2 \mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+\frac{1}{\rho _{o}}\frac{
1281 cnh 1.1 \partial \left( \rho _{o}w\right) }{\partial z}=0 \label{eq-za-cont2}
1282     \end{equation}
1283     Equations \ref{eq-za-cont1} and \ref{eq-za-cont2} are in fact the same
1284     equation if:
1285     \begin{equation}
1286     \frac{1}{\rho _{o}}\frac{\partial \rho _{o}}{\partial z}=\frac{-g}{c_{s}^{2}}
1287     \end{equation}
1288     Again, note that if $\rho _{o}$ is evaluated from prescribed $\theta _{o}$
1289 adcroft 1.2 and $S_{o}$ profiles, then the EOS dependency on $p_{o}$ and the term $\frac{
1290 cnh 1.1 g}{c_{s}^{2}}$ in continuity should be referred to those same profiles. The
1291     full set of `quasi-Boussinesq' or `anelastic' equations for the ocean are
1292     then:
1293     \begin{eqnarray}
1294 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1295 cnh 1.1 _{h}+\frac{1}{\rho _{o}}\mathbf{\nabla }_{z}p &=&\vec{\mathbf{\mathcal{F}}}
1296     \label{eq-zab-hmom} \\
1297 adcroft 1.2 \epsilon _{nh}\frac{Dw}{Dt}+\frac{g\rho }{\rho _{o}}+\frac{1}{\rho _{o}}
1298 cnh 1.1 \frac{\partial p}{\partial z} &=&\epsilon _{nh}\mathcal{F}_{w}
1299     \label{eq-zab-hydro} \\
1300 adcroft 1.2 \mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+\frac{1}{\rho _{o}}\frac{
1301 cnh 1.1 \partial \left( \rho _{o}w\right) }{\partial z} &=&0 \label{eq-zab-cont} \\
1302     \rho &=&\rho (\theta ,S,p_{o}(z)) \label{eq-zab-eos} \\
1303     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \label{eq-zab-heat} \\
1304     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq-zab-salt}
1305     \end{eqnarray}
1306    
1307     \subsubsection{Incompressible z-coordinate equations}
1308    
1309     Here, the objective is to drop the depth dependence of $\rho _{o}$ and so,
1310     technically, to also remove the dependence of $\rho $ on $p_{o}$. This would
1311     yield the ``truly'' incompressible Boussinesq equations:
1312     \begin{eqnarray}
1313 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1314 cnh 1.1 _{h}+\frac{1}{\rho _{c}}\mathbf{\nabla }_{z}p &=&\vec{\mathbf{\mathcal{F}}}
1315     \label{eq-ztb-hmom} \\
1316 adcroft 1.2 \epsilon _{nh}\frac{Dw}{Dt}+\frac{g\rho }{\rho _{c}}+\frac{1}{\rho _{c}}
1317 cnh 1.1 \frac{\partial p}{\partial z} &=&\epsilon _{nh}\mathcal{F}_{w}
1318     \label{eq-ztb-hydro} \\
1319     \mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial w}{\partial z}
1320     &=&0 \label{eq-ztb-cont} \\
1321     \rho &=&\rho (\theta ,S) \label{eq-ztb-eos} \\
1322     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \label{eq-ztb-heat} \\
1323     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq-ztb-salt}
1324     \end{eqnarray}
1325     where $\rho _{c}$ is a constant reference density of water.
1326    
1327     \subsubsection{Compressible non-divergent equations}
1328    
1329     The above ``incompressible'' equations are incompressible in both the flow
1330     and the density. In many oceanic applications, however, it is important to
1331     retain compressibility effects in the density. To do this we must split the
1332     density thus:
1333     \begin{equation*}
1334     \rho =\rho _{o}+\rho ^{\prime }
1335 adcroft 1.2 \end{equation*}
1336 cnh 1.1 We then assert that variations with depth of $\rho _{o}$ are unimportant
1337     while the compressible effects in $\rho ^{\prime }$ are:
1338     \begin{equation*}
1339     \rho _{o}=\rho _{c}
1340 adcroft 1.2 \end{equation*}
1341 cnh 1.1 \begin{equation*}
1342     \rho ^{\prime }=\rho (\theta ,S,p_{o}(z))-\rho _{o}
1343 adcroft 1.2 \end{equation*}
1344 cnh 1.1 This then yields what we can call the semi-compressible Boussinesq
1345     equations:
1346     \begin{eqnarray}
1347 adcroft 1.2 \frac{D\vec{\mathbf{v}}_{h}}{Dt}+f\hat{\mathbf{k}}\times \vec{\mathbf{v}}
1348     _{h}+\frac{1}{\rho _{c}}\mathbf{\nabla }_{z}p^{\prime } &=&\vec{\mathbf{
1349 cnh 1.1 \mathcal{F}}} \label{eq:ocean-mom} \\
1350     \epsilon _{nh}\frac{Dw}{Dt}+\frac{g\rho ^{\prime }}{\rho _{c}}+\frac{1}{\rho
1351     _{c}}\frac{\partial p^{\prime }}{\partial z} &=&\epsilon _{nh}\mathcal{F}_{w}
1352     \label{eq:ocean-wmom} \\
1353     \mathbf{\nabla }_{z}\cdot \vec{\mathbf{v}}_{h}+\frac{\partial w}{\partial z}
1354     &=&0 \label{eq:ocean-cont} \\
1355     \rho ^{\prime } &=&\rho (\theta ,S,p_{o}(z))-\rho _{c} \label{eq:ocean-eos}
1356     \\
1357     \frac{D\theta }{Dt} &=&\mathcal{Q}_{\theta } \label{eq:ocean-theta} \\
1358     \frac{DS}{Dt} &=&\mathcal{Q}_{s} \label{eq:ocean-salt}
1359 adcroft 1.2 \end{eqnarray}
1360 cnh 1.1 Note that the hydrostatic pressure of the resting fluid, including that
1361     associated with $\rho _{c}$, is subtracted out since it has no effect on the
1362     dynamics.
1363    
1364     Though necessary, the assumptions that go into these equations are messy
1365     since we essentially assume a different EOS for the reference density and
1366     the perturbation density. Nevertheless, it is the hydrostatic ($\epsilon
1367     _{nh}=0$ form of these equations that are used throughout the ocean modeling
1368     community and referred to as the primitive equations (HPE).
1369    
1370 adcroft 1.2 % $Header: /u/gcmpack/mitgcmdoc/part1/manual.tex,v 1.3 2001/10/10 16:48:45 cnh Exp $
1371 cnh 1.1 % $Name: $
1372    
1373     \section{Appendix:OPERATORS}
1374    
1375     \subsection{Coordinate systems}
1376    
1377     \subsubsection{Spherical coordinates}
1378    
1379     In spherical coordinates, the velocity components in the zonal, meridional
1380     and vertical direction respectively, are given by (see Fig.2) :
1381    
1382     \begin{equation*}
1383     u=r\cos \phi \frac{D\lambda }{Dt}
1384     \end{equation*}
1385    
1386     \begin{equation*}
1387     v=r\frac{D\phi }{Dt}\qquad
1388     \end{equation*}
1389     $\qquad \qquad \qquad \qquad $
1390    
1391     \begin{equation*}
1392     \dot{r}=\frac{Dr}{Dt}
1393     \end{equation*}
1394    
1395     Here $\phi $ is the latitude, $\lambda $ the longitude, $r$ the radial
1396     distance of the particle from the center of the earth, $\Omega $ is the
1397     angular speed of rotation of the Earth and $D/Dt$ is the total derivative.
1398    
1399     The `grad' ($\nabla $) and `div' ($\nabla $.) operators are defined by, in
1400     spherical coordinates:
1401    
1402     \begin{equation*}
1403 adcroft 1.2 \nabla \equiv \left( \frac{1}{r\cos \phi }\frac{\partial }{\partial \lambda }
1404     ,\frac{1}{r}\frac{\partial }{\partial \phi },\frac{\partial }{\partial r}
1405 cnh 1.1 \right)
1406     \end{equation*}
1407    
1408     \begin{equation*}
1409     \nabla .v\equiv \frac{1}{r\cos \phi }\left\{ \frac{\partial u}{\partial
1410     \lambda }+\frac{\partial }{\partial \phi }\left( v\cos \phi \right) \right\}
1411     +\frac{1}{r^{2}}\frac{\partial \left( r^{2}\dot{r}\right) }{\partial r}
1412     \end{equation*}
1413    
1414     \end{document}

  ViewVC Help
Powered by ViewVC 1.1.22