1 | \documentclass[../main/NEMO_manual]{subfiles} |
---|
2 | |
---|
3 | \begin{document} |
---|
4 | % ================================================================ |
---|
5 | % Chapter --- Miscellaneous Topics |
---|
6 | % ================================================================ |
---|
7 | \chapter{Miscellaneous Topics} |
---|
8 | \label{chap:MISC} |
---|
9 | |
---|
10 | \minitoc |
---|
11 | |
---|
12 | \newpage |
---|
13 | |
---|
14 | % ================================================================ |
---|
15 | % Representation of Unresolved Straits |
---|
16 | % ================================================================ |
---|
17 | \section{Representation of unresolved straits} |
---|
18 | \label{sec:MISC_strait} |
---|
19 | |
---|
20 | In climate modeling, it often occurs that a crucial connections between water masses is broken as |
---|
21 | the grid mesh is too coarse to resolve narrow straits. |
---|
22 | For example, coarse grid spacing typically closes off the Mediterranean from the Atlantic at |
---|
23 | the Strait of Gibraltar. |
---|
24 | In this case, it is important for climate models to include the effects of salty water entering the Atlantic from |
---|
25 | the Mediterranean. |
---|
26 | Likewise, it is important for the Mediterranean to replenish its supply of water from the Atlantic to |
---|
27 | balance the net evaporation occurring over the Mediterranean region. |
---|
28 | This problem occurs even in eddy permitting simulations. |
---|
29 | For example, in ORCA 1/4\deg several straits of the Indonesian archipelago (Ombai, Lombok...) |
---|
30 | are much narrow than even a single ocean grid-point. |
---|
31 | |
---|
32 | We describe briefly here the three methods that can be used in \NEMO to handle such improperly resolved straits. |
---|
33 | The first two consist of opening the strait by hand while ensuring that the mass exchanges through |
---|
34 | the strait are not too large by either artificially reducing the surface of the strait grid-cells or, |
---|
35 | locally increasing the lateral friction. |
---|
36 | In the third one, the strait is closed but exchanges of mass, heat and salt across the land are allowed. |
---|
37 | Note that such modifications are so specific to a given configuration that no attempt has been made to |
---|
38 | set them in a generic way. |
---|
39 | However, examples of how they can be set up is given in the ORCA 2\deg and 0.5\deg configurations. |
---|
40 | For example, for details of implementation in ORCA2, search: \texttt{IF( cp\_cfg == "orca" .AND. jp\_cfg == 2 )} |
---|
41 | |
---|
42 | % ------------------------------------------------------------------------------------------------------------- |
---|
43 | % Hand made geometry changes |
---|
44 | % ------------------------------------------------------------------------------------------------------------- |
---|
45 | \subsection{Hand made geometry changes} |
---|
46 | \label{subsec:MISC_strait_hand} |
---|
47 | |
---|
48 | $\bullet$ reduced scale factor in the cross-strait direction to a value in better agreement with |
---|
49 | the true mean width of the strait (\autoref{fig:MISC_strait_hand}). |
---|
50 | This technique is sometime called "partially open face" or "partially closed cells". |
---|
51 | The key issue here is only to reduce the faces of $T$-cell |
---|
52 | ($i.e.$ change the value of the horizontal scale factors at $u$- or $v$-point) but not the volume of the $T$-cell. |
---|
53 | Indeed, reducing the volume of strait $T$-cell can easily produce a numerical instability at |
---|
54 | that grid point that would require a reduction of the model time step. |
---|
55 | The changes associated with strait management are done in \mdl{domhgr}, |
---|
56 | just after the definition or reading of the horizontal scale factors. |
---|
57 | |
---|
58 | $\bullet$ increase of the viscous boundary layer thickness by local increase of the fmask value at the coast |
---|
59 | (\autoref{fig:MISC_strait_hand}). |
---|
60 | This is done in \mdl{dommsk} together with the setting of the coastal value of fmask (see \autoref{sec:LBC_coast}). |
---|
61 | |
---|
62 | %>>>>>>>>>>>>>>>>>>>>>>>>>>>> |
---|
63 | \begin{figure}[!tbp] |
---|
64 | \begin{center} |
---|
65 | \includegraphics[width=0.80\textwidth]{Fig_Gibraltar} |
---|
66 | \includegraphics[width=0.80\textwidth]{Fig_Gibraltar2} |
---|
67 | \caption{ |
---|
68 | \protect\label{fig:MISC_strait_hand} |
---|
69 | Example of the Gibraltar strait defined in a $1^{\circ} \times 1^{\circ}$ mesh. |
---|
70 | \textit{Top}: using partially open cells. |
---|
71 | The meridional scale factor at $v$-point is reduced on both sides of the strait to account for |
---|
72 | the real width of the strait (about 20 km). |
---|
73 | Note that the scale factors of the strait $T$-point remains unchanged. |
---|
74 | \textit{Bottom}: using viscous boundary layers. |
---|
75 | The four fmask parameters along the strait coastlines are set to a value larger than 4, |
---|
76 | $i.e.$ "strong" no-slip case (see \autoref{fig:LBC_shlat}) creating a large viscous boundary layer that |
---|
77 | allows a reduced transport through the strait. |
---|
78 | } |
---|
79 | \end{center} |
---|
80 | \end{figure} |
---|
81 | %>>>>>>>>>>>>>>>>>>>>>>>>>>>> |
---|
82 | |
---|
83 | |
---|
84 | % ================================================================ |
---|
85 | % Closed seas |
---|
86 | % ================================================================ |
---|
87 | \section{Closed seas (\protect\mdl{closea})} |
---|
88 | \label{sec:MISC_closea} |
---|
89 | |
---|
90 | \colorbox{yellow}{Add here a short description of the way closed seas are managed} |
---|
91 | |
---|
92 | |
---|
93 | % ================================================================ |
---|
94 | % Sub-Domain Functionality |
---|
95 | % ================================================================ |
---|
96 | \section{Sub-domain functionality} |
---|
97 | \label{sec:MISC_zoom} |
---|
98 | |
---|
99 | \subsection{Simple subsetting of input files via NetCDF attributes} |
---|
100 | |
---|
101 | The extended grids for use with the under-shelf ice cavities will result in redundant rows around Antarctica if |
---|
102 | the ice cavities are not active. |
---|
103 | A simple mechanism for subsetting input files associated with the extended domains has been implemented to |
---|
104 | avoid the need to maintain different sets of input fields for use with or without active ice cavities. |
---|
105 | The existing 'zoom' options are overly complex for this task and marked for deletion anyway. |
---|
106 | This alternative subsetting operates for the j-direction only and works by optionally looking for and |
---|
107 | using a global file attribute (named: \np{open\_ocean\_jstart}) to determine the starting j-row for input. |
---|
108 | The use of this option is best explained with an example: |
---|
109 | consider an ORCA1 configuration using the extended grid bathymetry and coordinate files: |
---|
110 | \vspace{-10pt} |
---|
111 | \ifile{eORCA1\_bathymetry\_v2} |
---|
112 | \ifile{eORCA1\_coordinates} |
---|
113 | \noindent These files define a horizontal domain of 362x332. |
---|
114 | Assuming the first row with open ocean wet points in the non-isf bathymetry for this set is row 42 |
---|
115 | (Fortran indexing) then the formally correct setting for \np{open\_ocean\_jstart} is 41. |
---|
116 | Using this value as the first row to be read will result in a 362x292 domain which is the same size as |
---|
117 | the original ORCA1 domain. |
---|
118 | Thus the extended coordinates and bathymetry files can be used with all the original input files for ORCA1 if |
---|
119 | the ice cavities are not active (\np{ln\_isfcav = .false.}). |
---|
120 | Full instructions for achieving this are: |
---|
121 | |
---|
122 | \noindent Add the new attribute to any input files requiring a j-row offset, i.e: |
---|
123 | \vspace{-10pt} |
---|
124 | \begin{cmds} |
---|
125 | ncatted -a open_ocean_jstart,global,a,d,41 eORCA1_coordinates.nc |
---|
126 | ncatted -a open_ocean_jstart,global,a,d,41 eORCA1_bathymetry_v2.nc |
---|
127 | \end{cmds} |
---|
128 | |
---|
129 | \noindent Add the logical switch to \ngn{namcfg} in the configuration namelist and set true: |
---|
130 | %--------------------------------------------namcfg-------------------------------------------------------- |
---|
131 | |
---|
132 | \nlst{namcfg} |
---|
133 | %-------------------------------------------------------------------------------------------------------------- |
---|
134 | |
---|
135 | \noindent Note the j-size of the global domain is the (extended j-size minus \np{open\_ocean\_jstart} + 1 ) and |
---|
136 | this must match the size of all datasets other than bathymetry and coordinates currently. |
---|
137 | However the option can be extended to any global, 2D and 3D, netcdf, input field by adding the: |
---|
138 | \vspace{-10pt} |
---|
139 | \begin{forlines} |
---|
140 | lrowattr=ln_use_jattr |
---|
141 | \end{forlines} |
---|
142 | optional argument to the appropriate \np{iom\_get} call and the \np{open\_ocean\_jstart} attribute to |
---|
143 | the corresponding input files. |
---|
144 | It remains the users responsibility to set \np{jpjdta} and \np{jpjglo} values in |
---|
145 | the \np{namelist\_cfg} file according to their needs. |
---|
146 | |
---|
147 | %>>>>>>>>>>>>>>>>>>>>>>>>>>>> |
---|
148 | \begin{figure}[!ht] |
---|
149 | \begin{center} |
---|
150 | \includegraphics[width=0.90\textwidth]{Fig_LBC_zoom} |
---|
151 | \caption{ |
---|
152 | \protect\label{fig:LBC_zoom} |
---|
153 | Position of a model domain compared to the data input domain when the zoom functionality is used. |
---|
154 | } |
---|
155 | \end{center} |
---|
156 | \end{figure} |
---|
157 | %>>>>>>>>>>>>>>>>>>>>>>>>>>>> |
---|
158 | |
---|
159 | |
---|
160 | % ================================================================ |
---|
161 | % Accuracy and Reproducibility |
---|
162 | % ================================================================ |
---|
163 | \section{Accuracy and reproducibility (\protect\mdl{lib\_fortran})} |
---|
164 | \label{sec:MISC_fortran} |
---|
165 | |
---|
166 | \subsection{Issues with intrinsinc SIGN function (\protect\key{nosignedzero})} |
---|
167 | \label{subsec:MISC_sign} |
---|
168 | |
---|
169 | The SIGN(A, B) is the \textsc {Fortran} intrinsic function delivers the magnitude of A with the sign of B. |
---|
170 | For example, SIGN(-3.0,2.0) has the value 3.0. |
---|
171 | The problematic case is when the second argument is zero, because, on platforms that support IEEE arithmetic, |
---|
172 | zero is actually a signed number. |
---|
173 | There is a positive zero and a negative zero. |
---|
174 | |
---|
175 | In \textsc{Fortran}~90, the processor was required always to deliver a positive result for SIGN(A, B) if B was zero. |
---|
176 | Nevertheless, in \textsc{Fortran}~95, the processor is allowed to do the correct thing and deliver ABS(A) when |
---|
177 | B is a positive zero and -ABS(A) when B is a negative zero. |
---|
178 | This change in the specification becomes apparent only when B is of type real, and is zero, |
---|
179 | and the processor is capable of distinguishing between positive and negative zero, |
---|
180 | and B is negative real zero. |
---|
181 | Then SIGN delivers a negative result where, under \textsc{Fortran}~90 rules, it used to return a positive result. |
---|
182 | This change may be especially sensitive for the ice model, |
---|
183 | so we overwrite the intrinsinc function with our own function simply performing : \\ |
---|
184 | \verb? IF( B >= 0.e0 ) THEN ; SIGN(A,B) = ABS(A) ? \\ |
---|
185 | \verb? ELSE ; SIGN(A,B) =-ABS(A) ? \\ |
---|
186 | \verb? ENDIF ? \\ |
---|
187 | This feature can be found in \mdl{lib\_fortran} module and is effective when \key{nosignedzero} is defined. |
---|
188 | We use a CPP key as the overwritting of a intrinsic function can present performance issues with |
---|
189 | some computers/compilers. |
---|
190 | |
---|
191 | |
---|
192 | \subsection{MPP reproducibility} |
---|
193 | \label{subsec:MISC_glosum} |
---|
194 | |
---|
195 | The numerical reproducibility of simulations on distributed memory parallel computers is a critical issue. |
---|
196 | In particular, within NEMO global summation of distributed arrays is most susceptible to rounding errors, |
---|
197 | and their propagation and accumulation cause uncertainty in final simulation reproducibility on |
---|
198 | different numbers of processors. |
---|
199 | To avoid so, based on \citet{He_Ding_JSC01} review of different technics, |
---|
200 | we use a so called self-compensated summation method. |
---|
201 | The idea is to estimate the roundoff error, store it in a buffer, and then add it back in the next addition. |
---|
202 | |
---|
203 | Suppose we need to calculate $b = a_1 + a_2 + a_3$. |
---|
204 | The following algorithm will allow to split the sum in two |
---|
205 | ($sum_1 = a_{1} + a_{2}$ and $b = sum_2 = sum_1 + a_3$) with exactly the same rounding errors as |
---|
206 | the sum performed all at once. |
---|
207 | \begin{align*} |
---|
208 | sum_1 \ \ &= a_1 + a_2 \\ |
---|
209 | error_1 &= a_2 + ( a_1 - sum_1 ) \\ |
---|
210 | sum_2 \ \ &= sum_1 + a_3 + error_1 \\ |
---|
211 | error_2 &= a_3 + error_1 + ( sum_1 - sum_2 ) \\ |
---|
212 | b \qquad \ &= sum_2 \\ |
---|
213 | \end{align*} |
---|
214 | An example of this feature can be found in \mdl{lib\_fortran} module. |
---|
215 | It is systematicallt used in glob\_sum function (summation over the entire basin excluding duplicated rows and |
---|
216 | columns due to cyclic or north fold boundary condition as well as overlap MPP areas). |
---|
217 | The self-compensated summation method should be used in all summation in i- and/or j-direction. |
---|
218 | See \mdl{closea} module for an example. |
---|
219 | Note also that this implementation may be sensitive to the optimization level. |
---|
220 | |
---|
221 | \subsection{MPP scalability} |
---|
222 | \label{subsec:MISC_mppsca} |
---|
223 | |
---|
224 | The default method of communicating values across the north-fold in distributed memory applications (\key{mpp\_mpi}) |
---|
225 | uses a \textsc{MPI\_ALLGATHER} function to exchange values from each processing region in |
---|
226 | the northern row with every other processing region in the northern row. |
---|
227 | This enables a global width array containing the top 4 rows to be collated on every northern row processor and then |
---|
228 | folded with a simple algorithm. |
---|
229 | Although conceptually simple, this "All to All" communication will hamper performance scalability for |
---|
230 | large numbers of northern row processors. |
---|
231 | From version 3.4 onwards an alternative method is available which only performs direct "Peer to Peer" communications |
---|
232 | between each processor and its immediate "neighbours" across the fold line. |
---|
233 | This is achieved by using the default \textsc{MPI\_ALLGATHER} method during initialisation to |
---|
234 | help identify the "active" neighbours. |
---|
235 | Stored lists of these neighbours are then used in all subsequent north-fold exchanges to |
---|
236 | restrict exchanges to those between associated regions. |
---|
237 | The collated global width array for each region is thus only partially filled but is guaranteed to |
---|
238 | be set at all the locations actually required by each individual for the fold operation. |
---|
239 | This alternative method should give identical results to the default \textsc{ALLGATHER} method and |
---|
240 | is recommended for large values of \np{jpni}. |
---|
241 | The new method is activated by setting \np{ln\_nnogather} to be true ({\bf nammpp}). |
---|
242 | The reproducibility of results using the two methods should be confirmed for each new, |
---|
243 | non-reference configuration. |
---|
244 | |
---|
245 | % ================================================================ |
---|
246 | % Model optimisation, Control Print and Benchmark |
---|
247 | % ================================================================ |
---|
248 | \section{Model optimisation, control print and benchmark} |
---|
249 | \label{sec:MISC_opt} |
---|
250 | %--------------------------------------------namctl------------------------------------------------------- |
---|
251 | |
---|
252 | \nlst{namctl} |
---|
253 | %-------------------------------------------------------------------------------------------------------------- |
---|
254 | |
---|
255 | \gmcomment{why not make these bullets into subsections?} |
---|
256 | Options are defined through the \ngn{namctl} namelist variables. |
---|
257 | |
---|
258 | $\bullet$ Vector optimisation: |
---|
259 | |
---|
260 | \key{vectopt\_loop} enables the internal loops to collapse. |
---|
261 | This is very a very efficient way to increase the length of vector calculations and thus |
---|
262 | to speed up the model on vector computers. |
---|
263 | |
---|
264 | % Add here also one word on NPROMA technique that has been found useless, since compiler have made significant progress during the last decade. |
---|
265 | |
---|
266 | % Add also one word on NEC specific optimisation (Novercheck option for example) |
---|
267 | |
---|
268 | $\bullet$ Control print %: describe here 4 things: |
---|
269 | |
---|
270 | 1- \np{ln\_ctl}: compute and print the trends averaged over the interior domain in all TRA, DYN, LDF and |
---|
271 | ZDF modules. |
---|
272 | This option is very helpful when diagnosing the origin of an undesired change in model results. |
---|
273 | |
---|
274 | 2- also \np{ln\_ctl} but using the nictl and njctl namelist parameters to check the source of differences between |
---|
275 | mono and multi processor runs. |
---|
276 | |
---|
277 | %%gm to be removed both here and in the code |
---|
278 | 3- last digit comparison (\np{nn\_bit\_cmp}). |
---|
279 | In an MPP simulation, the computation of a sum over the whole domain is performed as the summation over |
---|
280 | all processors of each of their sums over their interior domains. |
---|
281 | This double sum never gives exactly the same result as a single sum over the whole domain, |
---|
282 | due to truncation differences. |
---|
283 | The "bit comparison" option has been introduced in order to be able to check that |
---|
284 | mono-processor and multi-processor runs give exactly the same results. |
---|
285 | % THIS is to be updated with the mpp_sum_glo introduced in v3.3 |
---|
286 | % nn_bit_cmp today only check that the nn_cla = 0 (no cross land advection) |
---|
287 | %%gm end |
---|
288 | |
---|
289 | $\bullet$ Benchmark (\np{nn\_bench}). |
---|
290 | This option defines a benchmark run based on a GYRE configuration (see \autoref{sec:CFG_gyre}) in which |
---|
291 | the resolution remains the same whatever the domain size. |
---|
292 | This allows a very large model domain to be used, just by changing the domain size (\jp{jpiglo}, \jp{jpjglo}) and |
---|
293 | without adjusting either the time-step or the physical parameterisations. |
---|
294 | |
---|
295 | % ================================================================ |
---|
296 | \biblio |
---|
297 | |
---|
298 | \end{document} |
---|