New URL for NEMO forge!   http://forge.nemo-ocean.eu

Since March 2022 along with NEMO 4.2 release, the code development moved to a self-hosted GitLab.
This present forge is now archived and remained online for history.
chap_misc.tex in branches/2017/dev_merge_2017/DOC/texfiles/chapters – NEMO

source: branches/2017/dev_merge_2017/DOC/texfiles/chapters/chap_misc.tex @ 9376

Last change on this file since 9376 was 9376, checked in by nicolasmartin, 6 years ago

Global reorganisation of DOC directory: refactoring & cleaning of TeX source code for readability

File size: 15.4 KB
Line 
1\documentclass[NEMO_book]{subfiles}
2\begin{document}
3% ================================================================
4% Chapter ---€” Miscellaneous Topics
5% ================================================================
6\chapter{Miscellaneous Topics}
7\label{MISC}
8\minitoc
9
10\newpage
11$\ $\newline    % force a new ligne
12
13% ================================================================
14% Representation of Unresolved Straits
15% ================================================================
16\section{Representation of Unresolved Straits}
17\label{MISC_strait}
18
19In climate modeling, it often occurs that a crucial connections between water masses
20is broken as the grid mesh is too coarse to resolve narrow straits. For example, coarse
21grid spacing typically closes off the Mediterranean from the Atlantic at the Strait of
22Gibraltar. In this case, it is important for climate models to include the effects of salty
23water entering the Atlantic from the Mediterranean. Likewise, it is important for the
24Mediterranean to replenish its supply of water from the Atlantic to balance the net
25evaporation occurring over the Mediterranean region. This problem occurs even in
26eddy permitting simulations. For example, in ORCA 1/4\deg several straits of the Indonesian
27archipelago (Ombai, Lombok...) are much narrow than even a single ocean grid-point.
28
29We describe briefly here the three methods that can be used in \NEMO to handle
30such improperly resolved straits. The first two consist of opening the strait by hand
31while ensuring that the mass exchanges through the strait are not too large by
32either artificially reducing the surface of the strait grid-cells or, locally increasing
33the lateral friction. In the third one, the strait is closed but exchanges of mass,
34heat and salt across the land are allowed.
35Note that such modifications are so specific to a given configuration that no attempt
36has been made to set them in a generic way. However, examples of how
37they can be set up is given in the ORCA 2\deg and 0.5\deg configurations. For example,
38for details of implementation in ORCA2, search:
39\texttt{ IF( cp\_cfg == "orca" .AND. jp\_cfg == 2 ) }
40
41% -------------------------------------------------------------------------------------------------------------
42%       Hand made geometry changes
43% -------------------------------------------------------------------------------------------------------------
44\subsection{Hand made geometry changes}
45\label{MISC_strait_hand}
46
47$\bullet$ reduced scale factor in the cross-strait direction to a value in better agreement
48with the true mean width of the strait. (Fig.~\ref{Fig_MISC_strait_hand}).
49This technique is sometime called "partially open face" or "partially closed cells".
50The key issue here is only to reduce the faces of $T$-cell ($i.e.$ change the value
51of the horizontal scale factors at $u$- or $v$-point) but not the volume of the $T$-cell.
52Indeed, reducing the volume of strait $T$-cell can easily produce a numerical
53instability at that grid point that would require a reduction of the model time step.
54The changes associated with strait management are done in \mdl{domhgr},
55just after the definition or reading of the horizontal scale factors.
56
57$\bullet$ increase of the viscous boundary layer thickness by local increase of the
58fmask value at the coast (Fig.~\ref{Fig_MISC_strait_hand}). This is done in
59\mdl{dommsk} together with the setting of the coastal value of fmask
60(see Section \ref{LBC_coast})
61
62%>>>>>>>>>>>>>>>>>>>>>>>>>>>>
63\begin{figure}[!tbp]     \begin{center}
64\includegraphics[width=0.80\textwidth]{Fig_Gibraltar}
65\includegraphics[width=0.80\textwidth]{Fig_Gibraltar2}
66\caption{   \protect\label{Fig_MISC_strait_hand} 
67Example of the Gibraltar strait defined in a $1^{\circ} \times 1^{\circ}$ mesh.
68\textit{Top}: using partially open cells. The meridional scale factor at $v$-point
69is reduced on both sides of the strait to account for the real width of the strait
70(about 20 km). Note that the scale factors of the strait $T$-point remains unchanged.
71\textit{Bottom}: using viscous boundary layers. The four fmask parameters
72along the strait coastlines are set to a value larger than 4, $i.e.$ "strong" no-slip
73case (see Fig.\ref{Fig_LBC_shlat}) creating a large viscous boundary layer
74that allows a reduced transport through the strait.}
75\end{center}   \end{figure}
76%>>>>>>>>>>>>>>>>>>>>>>>>>>>>
77
78
79% ================================================================
80% Closed seas
81% ================================================================
82\section{Closed seas (\protect\mdl{closea})}
83\label{MISC_closea}
84
85\colorbox{yellow}{Add here a short description of the way closed seas are managed}
86
87
88% ================================================================
89% Sub-Domain Functionality
90% ================================================================
91\section{Sub-Domain Functionality}
92\label{MISC_zoom}
93
94\subsection{Simple subsetting of input files via netCDF attributes}
95
96The extended grids for use with the under-shelf ice cavities will result in redundant rows
97around Antarctica if the ice cavities are not active. A simple mechanism for subsetting
98input files associated with the extended domains has been implemented to avoid the need to
99maintain different sets of input fields for use with or without active ice cavities. The
100existing 'zoom' options are overly complex for this task and marked for deletion anyway.
101This alternative subsetting operates for the j-direction only and works by optionally
102looking for and using a global file attribute (named: \np{open\_ocean\_jstart}) to
103determine the starting j-row for input. The use of this option is best explained with an
104example: Consider an ORCA1 configuration using the extended grid bathymetry and coordinate
105files:
106\vspace{-10pt}
107\begin{verbatim}
108eORCA1_bathymetry_v2.nc
109eORCA1_coordinates.nc
110\end{verbatim}
111\noindent These files define a horizontal domain of 362x332. Assuming the first row with
112open ocean wet points in the non-isf bathymetry for this set is row 42 (Fortran indexing)
113then the formally correct setting for \np{open\_ocean\_jstart} is 41. Using this value as the
114first row to be read will result in a 362x292 domain which is the same size as the original
115ORCA1 domain. Thus the extended coordinates and bathymetry files can be used with all the
116original input files for ORCA1 if the ice cavities are not active (\np{ln\_isfcav =
117.false.}). Full instructions for achieving this are:
118
119\noindent Add the new attribute to any input files requiring a j-row offset, i.e:
120\vspace{-10pt}
121\begin{shcode}
122ncatted  -a open_ocean_jstart,global,a,d,41 eORCA1_coordinates.nc
123ncatted  -a open_ocean_jstart,global,a,d,41 eORCA1_bathymetry_v2.nc
124\end{shcode}
125 
126\noindent Add the logical switch to \ngn{namcfg} in the configuration namelist and set true:
127%--------------------------------------------namcfg--------------------------------------------------------
128\fortranfile{namelists/namcfg}
129%--------------------------------------------------------------------------------------------------------------
130
131\noindent Note the j-size of the global domain is the (extended j-size minus
132\np{open\_ocean\_jstart} + 1 ) and this must match the size of all datasets other than
133bathymetry and coordinates currently. However the option can be extended to any global, 2D
134and 3D, netcdf, input field by adding the:
135\vspace{-10pt}
136\begin{fortrancode}
137lrowattr=ln_use_jattr
138\end{fortrancode}
139optional argument to the appropriate \np{iom\_get} call and the \np{open\_ocean\_jstart} attribute to the corresponding input files. It remains the users responsibility to set \np{jpjdta} and \np{jpjglo} values in the \np{namelist\_cfg} file according to their needs.
140
141%>>>>>>>>>>>>>>>>>>>>>>>>>>>>
142\begin{figure}[!ht]    \begin{center}
143\includegraphics[width=0.90\textwidth]{Fig_LBC_zoom}
144\caption{   \protect\label{Fig_LBC_zoom}
145Position of a model domain compared to the data input domain when the zoom functionality is used.}
146\end{center}   \end{figure}
147%>>>>>>>>>>>>>>>>>>>>>>>>>>>>
148
149
150% ================================================================
151% Accuracy and Reproducibility
152% ================================================================
153\section{Accuracy and Reproducibility (\protect\mdl{lib\_fortran})}
154\label{MISC_fortran}
155
156\subsection{Issues with intrinsinc SIGN function (\protect\key{nosignedzero})}
157\label{MISC_sign}
158
159The SIGN(A, B) is the \textsc {Fortran} intrinsic function delivers the magnitude
160of A with the sign of B. For example, SIGN(-3.0,2.0) has the value 3.0.
161The problematic case is when the second argument is zero, because, on platforms
162that support IEEE arithmetic, zero is actually a signed number.
163There is a positive zero and a negative zero.
164
165In \textsc{Fortran}~90, the processor was required always to deliver a positive result for SIGN(A, B)
166if B was zero. Nevertheless, in \textsc{Fortran}~95, the processor is allowed to do the correct thing
167and deliver ABS(A) when B is a positive zero and -ABS(A) when B is a negative zero.
168This change in the specification becomes apparent only when B is of type real, and is zero,
169and the processor is capable of distinguishing between positive and negative zero,
170and B is negative real zero. Then SIGN delivers a negative result where, under \textsc{Fortran}~90
171rules,  it used to return a positive result.
172This change may be especially sensitive for the ice model, so we overwrite the intrinsinc
173function with our own function simply performing :   \\
174\verb?   IF( B >= 0.e0 ) THEN   ;   SIGN(A,B) = ABS(A)  ?    \\
175\verb?   ELSE                   ;   SIGN(A,B) =-ABS(A)     ?  \\
176\verb?   ENDIF    ? \\
177This feature can be found in \mdl{lib\_fortran} module and is effective when \key{nosignedzero}
178is defined. We use a CPP key as the overwritting of a intrinsic function can present
179performance issues with some computers/compilers.
180
181
182\subsection{MPP reproducibility}
183\label{MISC_glosum}
184
185The numerical reproducibility of simulations on distributed memory parallel computers
186is a critical issue. In particular, within NEMO global summation of distributed arrays
187is most susceptible to rounding errors, and their propagation and accumulation cause
188uncertainty in final simulation reproducibility on different numbers of processors.
189To avoid so, based on \citet{He_Ding_JSC01} review of different technics,
190we use a so called self-compensated summation method. The idea is to estimate
191the roundoff error, store it in a buffer, and then add it back in the next addition.
192
193Suppose we need to calculate $b = a_1 + a_2 + a_3$. The following algorithm
194will allow to split the sum in two ($sum_1 = a_{1} + a_{2}$ and $b = sum_2 = sum_1 + a_3$)
195with exactly the same rounding errors as the sum performed all at once.
196\begin{align*}
197   sum_1 \ \  &= a_1 + a_2 \\
198   error_1     &= a_2 + ( a_1 - sum_1 ) \\
199   sum_2 \ \  &= sum_1 + a_3 + error_1 \\
200   error_2     &= a_3 + error_1 + ( sum_1 - sum_2 ) \\
201   b \qquad \ &= sum_2 \\
202\end{align*}
203An example of this feature can be found in \mdl{lib\_fortran} module.
204It is systematicallt used in glob\_sum function (summation over the entire basin excluding
205duplicated rows and columns due to cyclic or north fold boundary condition as well as
206overlap MPP areas). The self-compensated summation method should be used in all summation
207in i- and/or j-direction. See closea.F90 module for an example.
208Note also that this implementation may be sensitive to the optimization level.
209
210\subsection{MPP scalability}
211\label{MISC_mppsca}
212
213The default method of communicating values across the north-fold in distributed memory applications
214(\key{mpp\_mpi}) uses a \textsc{MPI\_ALLGATHER} function to exchange values from each processing
215region in the northern row with every other processing region in the northern row. This enables a
216global width array containing the top 4 rows to be collated on every northern row processor and then
217folded with a simple algorithm. Although conceptually simple, this "All to All" communication will
218hamper performance scalability for large numbers of northern row processors. From version 3.4
219onwards an alternative method is available which only performs direct "Peer to Peer" communications
220between each processor and its immediate "neighbours" across the fold line. This is achieved by
221using the default \textsc{MPI\_ALLGATHER} method during initialisation to help identify the "active"
222neighbours. Stored lists of these neighbours are then used in all subsequent north-fold exchanges to
223restrict exchanges to those between associated regions. The collated global width array for each
224region is thus only partially filled but is guaranteed to be set at all the locations actually
225required by each individual for the fold operation. This alternative method should give identical
226results to the default \textsc{ALLGATHER} method and is recommended for large values of \np{jpni}.
227The new method is activated by setting \np{ln\_nnogather} to be true ({\bf nammpp}). The
228reproducibility of results using the two methods should be confirmed for each new, non-reference
229configuration.
230
231% ================================================================
232% Model optimisation, Control Print and Benchmark
233% ================================================================
234\section{Model Optimisation, Control Print and Benchmark}
235\label{MISC_opt}
236%--------------------------------------------namctl-------------------------------------------------------
237\fortranfile{namelists/namctl} 
238%--------------------------------------------------------------------------------------------------------------
239
240 \gmcomment{why not make these bullets into subsections?}
241Options are defined through the  \ngn{namctl} namelist variables.
242
243$\bullet$ Vector optimisation:
244
245\key{vectopt\_loop} enables the internal loops to collapse. This is very
246a very efficient way to increase the length of vector calculations and thus
247to speed up the model on vector computers.
248 
249% Add here also one word on NPROMA technique that has been found useless, since compiler have made significant progress during the last decade.
250 
251% Add also one word on NEC specific optimisation (Novercheck option for example)
252 
253$\bullet$ Control print %: describe here 4 things:
254
2551- \np{ln\_ctl} : compute and print the trends averaged over the interior domain
256in all TRA, DYN, LDF and ZDF modules. This option is very helpful when
257diagnosing the origin of an undesired change in model results.
258
2592- also \np{ln\_ctl} but using the nictl and njctl namelist parameters to check
260the source of differences between mono and multi processor runs.
261
262%%gm   to be removed both here and in the code
2633- last digit comparison (\np{nn\_bit\_cmp}). In an MPP simulation, the computation of
264a sum over the whole domain is performed as the summation over all processors of
265each of their sums over their interior domains. This double sum never gives exactly
266the same result as a single sum over the whole domain, due to truncation differences.
267The "bit comparison" option has been introduced in order to be able to check that
268mono-processor and multi-processor runs give exactly the same results.
269%THIS is to be updated with the mpp_sum_glo  introduced in v3.3
270% nn_bit_cmp  today only check that the nn_cla = 0 (no cross land advection)
271%%gm end
272
273$\bullet$  Benchmark (\np{nn\_bench}). This option defines a benchmark run based on
274a GYRE configuration (see \S\ref{CFG_gyre}) in which the resolution remains the same
275whatever the domain size. This allows a very large model domain to be used, just by
276changing the domain size (\jp{jpiglo}, \jp{jpjglo}) and without adjusting either the time-step
277or the physical parameterisations.
278
279% ================================================================
280\end{document}
281
282
283
284
Note: See TracBrowser for help on using the repository browser.