source: palm/trunk/SOURCE/CURRENT_MODIFICATIONS @ 191

Last change on this file since 191 was 191, checked in by raasch, 16 years ago

further adjustments and bugfixes for SGI system

  • Property svn:keywords set to Id
File size: 7.1 KB
Line 
1New:
2---
3The plant canopy model has been extended to the prognostic equation for   
4the potential temperature and the prognostic equation for the concentration
5of a passive scalar, as proposed by Watanabe (2004) and Shaw and Schumann
6(1992). (prognostic_equations, plant_canopy_model) 
7
8Within a job chain, the number of processors or the virtual processor grid
9is now allowed to be changed. This modifies the domain decomposition and
10requires that the PEs within a restart job have to know from which file(s) they
11have to read their restart data (up to now there is a one to one relation
12between the restart files and the PEs, i.e. processor 4 only reads data
13from file "_0004"). Informations about the subdomain index limits of all
14restart files are now written by PE0 (to file "_0000"). Also, all variables
15containing global values which do not depend on the subdomain are written
16to file "_0000" only. In a restart run, this file is opened by all PEs to
17read these global values and the informations about the subdomain index
18limits. Depending on these index limit information it is decided which
19restart file(s) has/have to be opened in order to read the complete data
20necessary for the current subdomain.
21
22The turbulence recycling method of Kataoka and Mizuno (2002) has been
23implemented. A pre-run with smaller model domain can be used to initialize
24the real run, in order to spare the CPU time needed to get the turbulence
25into a quasi-stationary state (see initializing_actions =
26'read_data_for_recycling').
27
28Particle advection adjusted for ocean runs.
29
30User-defined spectra.
31
32Argument -c introduced to most of the subjob calls, which allows the user to
33choose his own job catalog by setting job_catalog in the configuration file
34(default is ~/job_queue). Workaround for mpixec with -env option.
35Adjustments for lcxt4 (Bergen Center for Computational Science) (mrun, subjob)
36
37advec_particles, calc_spectra, check_open, check_parameters, data_output_spectra, header, init_particles, init_pegrid, init_3d_model, modules, netcdf, parin, particle_boundary_conds, plant_canopy_model, prognostic_equations, read_var_list, read_3d_binary, time_integration, user_interface, write_var_list, write_3d_binary
38
39New: inflow_turbulence
40
41
42Changed:
43-------
44Allow 100 spectra levels instead of 10 for consistency with
45define_netcdf_header (calc_spectra, data_output_spectra, header, modules,
46netcdf)
47
48Modification of the integrated version of the profile function for momentum
49for unstable stratification; more consistent flux definitions. (wall_fluxes,
50production_e)
51
52Strict grid matching along z is not needed for mg-solver. (check_parameters)
53
54Sharp lateral edges of the plant canopy introduced. Before smaller
55values of the leaf area density than those applied in the prognostic
56equation for the vertical velocity were used in the prognostic equations
57for u and v at the boundaries of the plant_canopy. (init_3d_model) 
58
59-cross_ts_* (read_var_list, write_var_list)
60-dopr_time_count, old profil parameters (cross_..., dopr_crossindex,
61 profile_***) removed (read_3d_binary, write_3d_binary)
62
63hom_sum moved, volume_flow_area, volume_flow_initial moved from ..._3d_binary
64to ..._var_list, hom_sum is now allocated in parin instead of init_3d_model
65
66npex, npey moved from inipar- to d3par-namelist (parin)
67
68myid_char14 removed. This was used since on IBM machines the process rank may
69be changed when the final communicator is defined, in order to save the
70preliminary processor-id for opening the binary output file for restarts
71(unit 14), because otherwise a mismatch occured when reading this file in a
72restart job. This mechanism is not needed any more since the subdomain
73informations are now contained in file _0000. (parin, check_open)
74
75Transpositions for the 2D domain decomposition have been optimized by using
76f_inv as an automatic array instead of providing the memory by a dummy argument.
77This spares one copy loop per transposition. Order of indices in the 3D loops
78in some of the transpose routines have been rearranged for better cache utilization.
79Both have been suggested by Roland Richter (SGI) as part of the
80HLRN-II benchmark process. (transpose)
81
82Workaround for getting information about the coupling mode. (palm)
83
84Format adjustments in order to allow CPU# > 999. (cpu_statistics)
85
86Provisional solution for run_control_1d output: add 'CALL check_open( 15 )'
87(init_1d_model)
88
89Subdomain data are read into temporary arrays pf_tmp/pf3d_tmp in order to
90avoid INTEL compiler warnings about (automatic) creation of temporary arrays
91(combine_plot_fields)
92
93advec_s_ups, advec_u_ups, advec_v_ups, advec_w_ups, calc_spectra, check_open, check_parameters, cpu_statistics, data_output_spectra, header, init_1d_model, init_3d_model, modules, netcdf, palm, parin, poisfft, production_e, read_var_list, read_3d_binary, transpose, wall_fluxes, write_var_list, write_3d_binary
94
95
96Errors:
97------
98Bugfix: change definition of us_wall from 1D to 2D. Tests showed that this decreases
99u* by some 10% and increases TKE and momentum fluxes by some 10% because friction
100was underestimated and momentum fluxes were wrongly calculated due to the bug.
101(prandtl_fluxes, wall_fluxes)
102
103Bugfix: calculation of horizontal fluxes at vertical walls (diffusion_s)
104
105Bugfix: zero assignments to tendency arrays in case of restarts (init_3d_model)
106
107Bugfix: wrong grid level used in multigrid solver, routine restrict (poismg)
108
109Bugfix: Construction of vertical profiles when 10 gradients have been
110specified in the parameter list (ug, vg, pt, q, sa, lad). (check_parameters)
111
112Bugfix: Determination of the number of vertical levels for which spectra
113shall be output - 100 replaced by 10. (netcdf)
114
115Bugfix: Lengths of those characters that are needed for the output of the
116characteristic levels of potential temperature, the geostrophic wind,
117scalar concentration, humidity and leaf area density to the header file
118has been increased. (header)
119
120Bugfix: add ')' in user_read_restart_data.
121
122The leaf area density is explicitly set to its surface value at k=0. (check_parameters)
123
124Introduce prefix_chr to ensure unique dvrp_file path.
125
126small bugfixes for user_interface sample code (comments):
127- initialize ustvst with 0.0 as it is now computed only until nxr and nyn
128- two ALLOCATE statements moved from user_read_restart_data back to user_init
129- remove 'READ (13) u2_av' statement in user_read_restart_data
130
131Bugfix: remove IF statement in plant_canopy_model_ij (plant_canopy_model)
132
133Bugfix: divide sums(k,8) (e) and sums(k,34) (e*) by ngp_2dh_s_inner(k,sr)
134(like other scalars) (flow_statistics)
135
136Bugfix: dopr_time_count was written on the binary file, which caused that
137NetCDF files newly created by restart files (no append of existing files!)
138contained uneccessary time levels. (read_3d_binary, write_3d_binary)
139
140Bugfix: extra '*' removed in user_statistics sample code (user_interface)
141
142Bugfix: a stop command was missing in some cases of the parallel branch (local_stop)
143
144Bugfix in volume flow control for non-cyclic boundary conditions (pres)
145
146Bugfix: misplaced #endif directives (combine_plot_fields)
147
148check_parameters, diffusion_s, flow_statistics, init_dvrp, init_3d_model, local_stop, plant_canopy_model, poismg, prandtl_fluxes, pres, read_3d_binary, user_interface, wall_fluxes, write_3d_binary
149
Note: See TracBrowser for help on using the repository browser.