[Wrf-users] Issue with restart run and adaptive time step

[Guillermo]

Dear wrf-users,

We are trying to do a restart run using WRF-ARW v3.5 and adaptive time 
step. Unfortunately the results are inconsistent, specially in near 
surface variables. This is the case for 2 meters temperature that do 
have very small diurnal oscillation compared to fixed time step run.

As an example you can see attached the pictures of 2 different restarts 
starting at 09-15-2014 00Z. Left pictures show the fixed time step run 
while right pictures are created using adaptive time step. Bottom 
pictures show results at +15h simulation time. As you can see they are 
quite different.

Simulation previous to restart runs were performed for 4 days using 
adaptive time step and the only difference in the "namelist.input" of 
the restart runs is the variable "use_adaptive_time_step".

Any clues would be appreciated.

The "namelist.input" file is as follows:

  &time_control
  run_days                            = 5,
  run_hours                           = 0,
  run_minutes                         = 0,
  run_seconds                         = 0,
  start_year                          = 2014,2014,
  start_month                         = 9,9,
  start_day                           = 15,15,
  start_hour                          = 0,0,
  start_minute                        = 00,   00,
  start_second                        = 00,   00,
  end_year                            = 2014,2014,
  end_month                           = 9,9,
  end_day                             = 20,20,
  end_hour                            = 0,0,
  end_minute                          = 00,   00,
  end_second                          = 00,   00,
  interval_seconds                    = 21600
  input_from_file                     = .true.,.true.,
  fine_input_stream                   =  2,   2,
  history_interval                    = 60,  60,
  frames_per_outfile                  = 1000, 1000,
  restart                             = .true.,
  restart_interval                    = 5760,
  io_form_history                     = 2
  io_form_restart                     = 2
  io_form_input                       = 2
  io_form_boundary                    = 2
  io_form_auxinput2                   = 2
  debug_level                         = 000
  use_netcdf_classic                  = .true.
  write_hist_at_0h_rst                = .true.
  reset_simulation_start              = .true.
  override_restart_timers             = .true.
  /

  &domains
  time_step                           = 216,
  time_step_fract_num                 = 0,
  time_step_fract_den                 = 1,
  max_dom                             = 2,
  s_we                                = 1,     1,
  e_we                                = 119,  172,
  s_sn                                = 1,     1,
  e_sn                                = 105,  139,
  s_vert                              = 1,     1,
  e_vert                              = 33,    33,
  num_metgrid_levels                  = 27
  eta_levels                          = 1.000, 0.994, 0.990, 0.986, 0.981,
                                        0.971, 0.952, 0.933, 0.904, 0.875,
                                        0.856, 0.827, 0.798, 0.760, 0.712,
                                        0.664, 0.616, 0.568, 0.520, 0.448,
                                        0.376, 0.304, 0.232, 0.184, 0.121,
                                        0.097, 0.059, 0.050, 0.040, 0.030,
                                        0.020, 0.010, 0.000,
  dx                                  = 36000, 12000,
  dy                                  = 36000, 12000,
  p_top_requested                     = 1000
  grid_id                             = 1,     2,
  parent_id                           = 0,     1,
  i_parent_start                      = 1,     47,
  j_parent_start                      = 1,     28,
  parent_grid_ratio                   = 1,     3,
  parent_time_step_ratio              = 1,     3,
  feedback                            = 0,
  smooth_option                       = 0,
  use_adaptive_time_step              = .true.,
  step_to_output_time                 = .true.,
  target_cfl                          = 1.1, 1.1,
  max_step_increase_pct               =  5,  15,
  starting_time_step                  = 216,  72
  max_time_step                       = 540, 180,
  min_time_step                       = 216,  72,
  /

  .........



Best wishes,

Guillermo
-- 
Firma @ MeteoGalicia
<http://www.xunta.es>
<http://www.meteogalicia.es>
Guillermo Fernández García
MeteoGalicia (Dpto. Predición Numérica e Investigación)
Rúa Roma, 6
15.707 Santiago de Compostela
<https://maps.google.com/maps?q=%22Rua+Roma+6,+15707+Santiago+de+Compostela%22> 
Tel: (+34) 981 957 462 <tel:0034981957462> / (+34) 881 999 646
<tel:0034881999646> Email: guillermo.fernandez at meteogalicia.es 
<mailto:guillermo.fernandez at meteogalicia.es>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20140912/c6971f26/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: T2 in wrfout_d01_2014-09-15.png
Type: image/png
Size: 282434 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/wrf-users/attachments/20140912/c6971f26/attachment-0001.png