To activate time series output in WRF, a file named "tslist" must be present in the WRF run directory. The tslist file contains a list of locations, given by their latitude and longitude, along with a short description and an abbreviation for each location. The first three lines in the file are regarded as header information, and are ignored. The contents of an example tslist file are shown below. #-----------------------------------------------# # 24 characters for name | pfx | LAT | LON | #-----------------------------------------------# Cape Hallett hallt -72.330 170.250 McMurdo Station mcm -77.851 166.713 Bogus point A pt_a 29.718 -75.772 Bogus point B pt_b 37.614 -74.650 Given a tslist file, for each location inside a model domain (either coarse or nested), below files are created: 1. pfx*.dNN.TS containing the regular time series output of surface variables. 2. pfx.dNN.UU containing a vertical profile of u wind component for each time step 3. pfx.dNN.VV containing a vertical profile of v wind component for each time step 4. pfx.dNN.TH containing a vertical profile of potential temperature for time step 5. pfx.dNN.PH containing a vertical profile of geopotential height in time step 6. pfx.dNN.QV containing a vertical profile of water vapor mixing ratio time step Where pfx is the specified prefix for the dd location in the tslist file, and NN is the domain ID, as given in namelist.input. If locations not in any model domain are specified in the tslist file, they will be simply ignored by the time series capability. The maximum number of time series locations is controlled by the namelist variable max_ts_locs, which may be specified in the &domains namelist. Also, the number of entries in the buffers for time series output can be set with the ts_buf_size variable, also found in &domains. Decreasing the size of the buffers will reduce the amount of memory that is allocated for time series; however, smaller buffers will need to be flushed to disk more often than larger buffers. Thus, it is recommended that the size of the buffer be set to the maximum number of time steps for any domain in a model run. Namelist.input variables related to to this new capability: * max_ts_locs = maximum number of locations in 'tslist' ( default is 5) * ts_buf_size = buffer size for time series output (default is 200) * max_ts_level = number of model levels for time series vertical profiles (default is 15) The first line in a time-series output of surface variables (pfx*.dNN.TS) looks like this: NZCM McMurdo 2 7 mcm (-77.850, 166.710) ( 153, 207) (-77.768, 166.500) 81.8 meters Those are name of the station, grid ID, time-series ID, station lat/lon, grid indices (nearest grid point to the station location), grid lat/lon, elevation. The variables from the time series output are: id, ts_hour, id_tsloc, ix, iy, t, q, u, v, psfc, glw, gsw, hfx, lh, tsk, tslb(1), rainc, rainnc, clw id: grid ID ts_hour: forecast time in hours id_tsloc: time series ID ix,iy: grid location (nearest grid to the station) t: 2 m Temperature (K) q: 2 m vapor mixing ratio (kg/kg) u: 10 m U wind (earth-relative) v: 10 m V wind (earth-relative) psfc: surface pressure (Pa) glw: downward longwave radiation flux at the ground (W/m^2, downward is positive) gsw: net shortwave radiation flux at the ground (W/m^2, downward is positive) hfx: surface sensible heat flux (W/m^2, upward is positive) lh: surface latent heat flux (W/m^2, upward is positive) tsk: skin temperature (K) tslb(1): top soil layer temperature (K) rainc: rainfall from a cumulus scheme (mm) rainnc: rainfall from an explicit scheme (mm) clw: total column-integrated water vapor and cloud variables Example: 2 0.033333 7 153 207 -39.49153 0.00104 -4.57959 7.21139 99064.35938 168.17384 12.06811 -41.77642 0.00001 259.18286 271.35104 0.00000 0.00000 0.00000 Format of the files of vertical profile: each line starting with the model time in hours, followed by the variable at model level 1,2,3, ... up to the highest model level of interest