User's Guide to namelist.input file (located in "run" directory) CM1 Numerical Model, Release 16 (cm1r16) Last updated: 6 February 2012 Entries with "--- NEW ---" are new (or modified) in cm1r16 NOTE: some namelist.input files that have been configured for certain idealized simulations ... e.g., squall lines, supercells, hurricanes ... are available on the CM1 website: http://www.mmm.ucar.edu/people/bryan/cm1/ ------------------------------------------------------------------------- param0 section -- all are INTEGER, except for terrain_flag (a logical) nx - Total number of grid points in x direction ny - Total number of grid points in y direction nz - Total number of grid points in z direction nodex - Total number of processors in x direction (for MPI runs only) nodey - Total number of processors in y direction (for MPI runs only) timeformat - Format for text printout of model integration time: 1 = seconds 2 = minutes 3 = hours 4 = days timestats - 0 = Do not provide provide timing statistics 1 = Provide timing statistics at end of simulation 2 = The same as 1, but include time required to complete each time step terrain_flag - .true. = With terrain .false. = No terrain (flat lower boundary) --- NEW --- procfiles - .true. = Text printout/config files for every MPI process .false. = Only one text printout/config file (Default) ------------------------------------------------------------------------- param1 section -- enter REAL values dx - Horizontal grid spacing in x direction (m). dy - Horizontal grid spacing in y direction (m). dz - Vertical grid spacing (m). NOTE: The variables dx,dy,dz are only used when stretch=0. When stretch=1, these variables should be set to an approximately average value (to minimize roundoff errors). See README.stretch for more information. dtl - Large time step (s). For psolver = 2,3,4, or 5, this time step is limited by the fastest nonacoustic speed. For thunderstorm simulations, this is usually the maximum vertical velocity. Otherwise, this would be the propagation speed of gravity waves. The following usually works well: dtl = min(dx,dy,dz)/67 (rounded to an appropriate value, of course) For psolver=1, this time step is limited by the propagation speed of sound waves. dtl of about min(dx,dy,dz)/700 is recommended. When using adaptive time-stepping, set dtl to a reasonable "target" value (i.e., a value you think would probably be best for your simulation). This will be used as the initial timestep. timax - Maximum forecast time (s). tapfrq - Frequency of three-dimensional model output (s). Output is in cm1out files. rstfrq - Frequency to save model restart files (s). Set to a negative number if restart files are not desired. statfrq - Frequency for calculating some interesting output. (seconds) Set to negative number to output stats every timestep. Output is in cm1out_stats.dat file. See param10 section below for the information that can be requested. prclfrq - Frequency to output parcel data (s). Note ... this does not affect the parcel calculations themselves, which are always updated every timestep; it merely tells the model how frequently to output the information. Set to a negative number to output parcel data every time step. ------------------------------------------------------------------------- param2 section -- enter INTEGER values (value in CAPS is recommended, where applicable) adapt_dt - Use adaptive timestep? (0=NO, 1=yes) Model automatically adjusts timestep to maintain stability. NOTES: - nsound is ignored for adapt_dt=1 - a reasonable value must still be assigned to dtl (above), which will also be used as the initial timestep irst - Is this a restart? (0=no, 1=yes) rstnum - If this is a restart, this variable specifies the number of the restart file. For example, for file cm1out_0000_0002_rst.dat, rstnum is 2. iconly - Setup initial conditions only? 1 = creates initial conditions, but does not run model. 0 = creates initial conditions, and proceeds with integration. hadvorder - Order of horizontal advection scheme. 5=fifth order, 6=sixth order vadvorder - Order of vertical advection scheme. 5=fifth order, 6=sixth order Fifth order scheme has implicit diffusion. If fifth order scheme is used, idiff can be set to 0 (i.e., no additional artifical diffusion is typically necessary). Sixth order scheme requires additional artifical diffusion for stability (i.e., idiff=1 is recommended). User should use difforder=6 and value of kdiff6 between about 0.02-0.24 Reference: Wicker and Skamarock, 2002, MWR, p 2088. pdscheme - Scheme to ensure positive definiteness of moisture: 0 = set negative values back to zero (not mass conserving) 1 = simple, nearly mass-conserving scheme. Takes water from neighboring grid cells to conserve water (similar to that in Cohen, 2000, JAS, p. 1661) advweno - Advect scalars (except for pressure) with WENO scheme? 0 = no 1 = yes, apply on every Runge-Kutta step 2 = yes, apply on final Runge-Kutta step only (default) Reference: Shu, 2001: High order finite difference and finite volume WENO schemes and discontinuous Galerkin methods for CFD. ICASE Rep. 2001-11. ---NEW--- for cm1r16: weights are now based on study by Shen and Zha, 2010, Int. J. Numer. Meth. Fluids, 64, 653-675. idiff - Include additional artificial diffusion? (0=no, 1=yes) (in addition to any diffusion from subgrid turbulence scheme) For idiff=1, diffusion of all variables. For idiff=2, diffusion only applied to winds (u,v,w). User must also set difforder and kdiff2 or kdiff6. vdiff - When idiff=1, apply artificial diffusion in vertical direction? (0=no, 1=yes) mdiff - When idiff=1 and difforder=6, apply monotonic version of artificial diffusion? (0=no, 1=yes) Reference: Xue, 2000, MWR, p 2853. difforder - Order of diffusion scheme. 2=second order, 6=sixth order. Second order diffusion is not generally recommended. It is only used for certain idealized cases. kdiff2 must be set appropriately when difforder=2. Sixth order diffusion is recommended for general use when diffusion of small scales (2-6 delta) is needed. User must also set kdiff6 when difforder=6. imoist - Include moisture? (0=no, 1=yes) iturb - Include subgrid turbulence scheme? 0 = no 1 = TKE (for LES) 2 = Smagorinsky (for LES) 3 = parameterized turbulence (user must set l_h and l_inf below) tconfig - Calculation of turbulence coefficients in turbulence scheme: 1 = horizontal and vertical turbulence coefficients are the same; use this if dx,dy are about equal to dz (default) 2 = horizontal turbulence coefficient is different from vertical turbulence coefficient; use this if dx,dy are much greater than dz, or if iturb=3 is used. bcturbu - Lower/upper boundary condition for vertical diffusion of winds (u and v). This includes diffusion from subgrid turbulence (when iturb >= 1) and artificial diffusion (when vdiff=1). 1 = zero flux (default) 2 = zero gradient 3 = no slip (note: bcturbu=3 is required if idrag=1) bcturbs - Lower/upper boundary condition for vertical diffusion of all scalars. This includes diffusion from subgrid turbulence (when iturb >= 1) and artificial diffusion (when vdiff=1). 1 = zero flux (default) 2 = zero gradient dns - Run model in Direct Numerical Simulation mode. (iturb must be = 0) For this option, the user must set the six parameters in param7 below. irdamp - Use upper-level Rayleigh damping zone? (0=no, 1=YES) (acts on u,v,w, and theta) (User must set rdalpha and zd below) hrdamp - Use Rayleigh damping near lateral boundaries (0=NO, 1=yes) (acts on u,v,w only) (User must set rdalpha and xhd below) psolver - Option for pressure solver. 1 = truly compressible: no acoustic damping, no small time steps, no implicit numerics 2 = Klemp-Wilhelmson time-splitting, fully explicit: uses K-W split time steps for acoustic modes, with explicit treatment of acoustic terms in both vertical and horizontal directions. (use if dx,dy,dz are approximately equal) 3 = Klemp-Wilhelmson time-splitting, vertically implicit: uses K-W split time steps for acoustic modes, with a vetically implicit solver, and horizontally explicit calculations (as in MM5, ARPS, WRF). (use if dz is much smaller than dx,dy) 4 = Anelastic solver: Uses the anelastic mass continuity equation. Pressure is retrieved diagnostically. 5 = Incomprssible solver: Uses the incompressible mass continuity equation. Pressure is retrieved diagnostically. NOTE: psolver = 4 and 5 are only recommended for simulations without moisture (for now). nsound - Number of small (acoustic) time steps per large time step. Only used for psolver = 2 and 3. note: nsound is ignored for adaptive time-stepping (adapt_dt=1) NOTE! nsound must be an EVEN integer (exactly) for the split time steps to work properly. There is no check for this ... so please remember. Typically, nsound of 4,6,8,10,12 is appropriate. For nsound less than 4, you should probably use psolver=1 instead. A value of nsound greater than 12 is not generally recommended. Also keep in mind the small time step (dts=dtl/nsound) that is required for stability: For psolver=2, dts should be about min(dx,dy,dz)/700. For psolver=3, dts should be about min(dx,dy)/700. ptype - Explicit moisture scheme. 1 = Kessler scheme (water only) 2 = NASA-Goddard version of LFO scheme 3 = Thompson scheme 4 = Gilmore/Straka/Rasmussen version of LFO scheme (default) 5 = Morrison double-moment scheme 6 = Rotunno-Emanuel (1987) simple water-only scheme ihail - Use hail or graupel for large ice category when ptype=2,5. (Goddard-LFO and Morrison schemes only) 1 = hail 0 = graupel iautoc - Include autoconversion of qc to qr when ptype = 2? (0=no, 1=yes) (Goddard-LFO scheme only) icor - Include Coriolis acceleration? (0=no, 1=yes) (If user chooses 1, then fcor must be set below) f-plane is assumed. pertcor - Apply Coriolis acceleration to perturbation winds only? (0=NO, 1=yes) neweqts - Use new equations (from Bryan and Fritsch 2002) ? 0 = no, use traditional (approximate) equation set 1 = yes (original formulation) 2 - yes, but apply condensation adjustment on RK steps (RECOMMENDED) (note: switch is irrelevent if imoist = 0, i.e. the equations are equivalent in a dry environment) (only available with ptype=1,2,3,5,6) idiss - Include dissipative heating? (0=no, 1=yes) efall - Include energy fallout term? (0=no, 1=yes) rterm - Include simple relaxation term that mimics atmospheric radiation? (0=no, 1=yes) (Note: this is a very simple approach, and is only recommended for highly idealized model simulations. See Rotunno and Emanuel 1987, JAS, p. 546 for a description of this term.) wbc - West lateral boundary condition. ebc - East lateral boundary condition. sbc - South lateral boundary condition. nbc - North lateral boundary condition. where: 1 = periodic 2 = open-radiative 3 = rigid walls irbc - For bc=2, this is the type of radiative scheme to use: 1 = Klemp-Wilhelmson (1978) on large steps 2 = Klemp-Wilhelmson (1978) on small steps 4 = Durran-Klemp (1983) formulation roflux - Restrict outward flux? (0=no, 1=yes) When this option is activated, the total outward mass flux at open boundary conditions is not allowed to exceed total inward mass flux. This is a requirement for the anelastic solver. For the compressible solvers, this scheme helps prevent runaway outward mass flux that can cause domain-total mass loss and pressure falls. isnd - Base-state sounding: 1 = Dry adiabatic 2 = Dry isothermal 3 = Dry, constant dT/dz [some variables can be 4 = Saturated neutrally stable (BF02 sounding) set in base.F file] 5 = Weisman-Klemp analytic sounding 7 = External file (named 'input_sounding') (see isnd=7 section of base.F for info) (some soundings are available at http://www.mmm.ucar.edu/people/bryan/cm1) 8 = Dry, constant d(theta)/dz 9 = Dry, constant Brunt-Vaisala frequency 10 = Saturated, constant Brunt-Vaisala frequency 11 = Saturated, constant equiv. pot. temp. iwnd - Base-state wind profile: (ignored if isnd=7) 0 = zero winds [additional variables 1 = RKW-type profile need to be set in 2 = Weisman-Klemp supercell base.F file] 3 = multicell 4 = Weisman-Klemp multicell 5 = Dornbrack etal analytic profile 6 = constant wind itern - Initial topography specifications. User must also set zs array in init_terrain.F: 1 = bell-shaped hill 2 = Schaer test case 3 = (case from T. Lane and J. Doyle) 4 = specified in external GrADS file iinit - 3D initialization option: 1 = warm bubble 2 = cold pool [additional variables 3 = line of warm bubbles need to be set in 4 = initialization for moist benchmark init3d.F file] 5 = cold blob 7 = Rotunno-Emanuel tropical cyclone 8 = line thermal with random perturbations 9 = forced convergence (Loftus et al 2008) irandp - Include random potential temperature perturbations in the initial conditions? (0=no, 1=yes) (set magnitude of perturbations in init3d.F) ibalance - Specified balance assumption for initial 3D pressure field (ignored if iinit=7) 0 = no balance (initial pressure perturbation is zero everywhere, except for iinit=7) 1 = hydrostatic balance (appropriate for small aspect ratios) 2 = anelastic balance (initial pressure perturbation is the buoyancy pressure perturbation field for an anelastic atmosphere). (Does not currently work with MPI setup.) iorigin - Specifies location of the origin in horizontal space 1 = At the bottom-left corner of the domain (x goes from 0 km to nx*dx km) (y goes from 0 km to ny*dy km) 2 = At the center of the domain (x goes from -nx*dx/2 km to +nx*dx/2 km) (y goes from -ny*dy/2 km to +ny*dy/2 km) axisymm - Run axisymmetric version of model (0=no, 1=yes) (for axisymm=1, ny must be 1, wbc must be 3, and sbc,nbc must be 1) (see README.axisymm for more information) imove - Move domain at constant speed (0=no, 1=yes) For imove=1, user must set umove and vmove. iptra - Integrate passive fluid tracer? (0=no, 1=yes) User must initialize "pta" array in init3d.F. npt - Total number of passive fluid tracers. iprcl - Integrate passive parcels? (0=no, 1=yes) User must initialize "pdata" array in init3d.F. nparcels - Total number of parcels. ------------------------------------------------------------------------- param3 section -- enter REAL values kdiff2 - Diffusion coefficient for difforder=2. Specified in m^2/s. kdiff6 - Diffusion coefficient for difforder=6. Specified as a fraction of one-dimensional stability. A value between 0.02-0.24 is recommended. fcor - Coriolis parameter (1/s). kdiv - Coefficient for divergence damper. Value of ~0.1 is recommended. This is only used when psolver=2,3. (The divergence damper is an artificial term designed to damp acoustic waves.) alph - Off-centering coefficient for vertically implicit acoustic solver. A value of 0.5 is centered-in-time. Slight forward-in-time bias is recommended. Default value is 0.60. (only used for psolver=3) rdalpha - Inverse e-folding time for upper-level Rayleigh damping layer (1/s). Value of about 1/300 is recommended. zd - Height above which Rayleigh damping is applied (m). (when irdamp = 1) xhd - Distance from lateral boundaries where Rayleigh damping is applied (m). (when hrdamp = 1) umove - Constant speed for domain translation in x-direction (m/s) (for imove = 1) vmove - Constant speed for domain translation in y-direction (m/s) (for imove = 1) v_t - Constant terminal fall velocity of liquid water (m/s) when ptype=6 l_h - Horizontal turbulence length scale (m) used when iturb=3 (parameterized turbulence) --- NEW --- (modified for cm1r16) l_inf - Asymptotic vertical turbulence length scale (m) (i.e., vertical length scale at z = infinity) used for iturb=3 (parameterized turbulence) --- NEW --- ndcnst - specified cloud droplet concentration for default version of Morrison microphysics scheme (units of cm-3) Note: typical value for maritime environments: 100 cm-3 typical value for continental environments: 300 cm-3 ------------------------------------------------------------------------- param11 section - atmospheric radiation NOTE: The parameters in this section ONLY apply to atmospheric radation. You do not need to set anything here unless radopt = 1. radopt - Use atmospheric radiation code? 0 = no 1 = yes The only option, at present, is the NASA-Goddard longwave and shortwave radiation codes, which are kindly provided by the ARPS/CAPS group at the University of Oklahoma. (Note: TIPA option is not implemented in this version of CM1) (Another Note: technically, the interaction of radiation with clouds is configured specifically for the NASA-Goddard LFO scheme, i.e., ptype=2. A future version of CM1 will pass the proper variables from all microphysics schemes into the radiation code so that consistent calculations are performed. That said, the radiative tendencies should still be reasonable for all ice microphysics schemes, and there is no issue for clear-sky conditions.) If radopt=1, set the following parameters: dtrad - Time increment (seconds) between calculation of radiation tendency. (Radiative tendencies are held fixed in-between calls to the atmospheric radiation subroutine.) ctrlat - Latitude (applies to entire domain, for now) ctrlon - Longitude (applies to entire domain, for now) NOTE: because ctrlat and ctrlon are fixed (for now) the radiation scheme is only appropriate for domains having horizontal extent of order 100--1000 km or less (FAQ: Why are lat and lon fixed across the entire domain? It's because George doesn't have time, at the moment, to deal with map projections in CM1.) year - Year (integer) at start of simulation month - Month (integer) at start of simulation day - Day (integer) at start of simulation hour - Hour (integer) at start of simulation minute - Minute (integer) at start of simulation second - Second (integer) at start of simulation Yet Another Note: the radiation scheme uses three important pieces of information from the surface section (param12) below: surface temperature, land/water flag, and land-use type. Make sure you have the desired settings for your simulation below (even if you are not using surface fluxes!). ------------------------------------------------------------------------- param12 section: surface model, ocean model, boundary layer: NOTE: By default, surface conditions are the same everywhere at the initial time. But, users can define spatially varying initial surface conditions in init_surface.F idrag - Include surface drag? (0=no, 1=yes) isfcflx - Include surface fluxes of heat and moisture (0=no, 1=yes) sfcmodel - Surface model: (Specifically, method to calculate surface fluxes over land and water) 1 = original CM1 formulation 2 = version from WRF model (see further details below) - sfcmodel=1 uses simple formulations wherein surface exchange coefficients are specified: see "Options for sfcmodel = 1" section below. For diagnostic surface layer calculations (such as 10-m winds), a neutrally stratified surface layer is assumed. Notes: - surface temperature remains fixed over time - surface moisture availability remains fixed over time - sfcmodel=1 requires oceanmodel=1 - sfcmodel=2 uses the MM5/WRF similarity theory code for the surface layer: based on Monin-Obukhov with Carslon-Boland viscous sub-layer and standard similarity functions from look-up tables. ("sf_sfclay_physics = 1" in WRF) See also "Options for sfcmodel = 2" section below. The soil model is the "Thermal diffusion" model from MM5/WRF: it updates soil temperature only ... soil moisture availability is held fixed over time. (Same as "sf_surface_physics = 1" in WRF) Notes: - sfcmodel=2 can be used with either oceanmodel=1,2 oceanmodel - Model for ocean conditions: 1 = fixed sea-surface temperature 2 = ocean mixed layer model (Same as "omlcall = 1" in WRF) (Note: oceanmodel=2 requires sfcmodel=2) Ref: Pollard et al, 1973, Geophys. Fluid Dyn., 3, 381-404. ipbl - Use Yonsei University (YSU) PBL parameterization? 0 = no 1 = yes NOTE: ipbl=1 cannot be used with iturb=1,2 (that is, you cannot run an LES configuration and parameterize the PBL at the same time) NOTE: if ipbl=1 and iturb=3, then l_inf must be 0 (in this case, iturb=3 becomes a 2D, horizontal, scheme) Reference: Hong et al, 2006, MWR, p 2318) ("bl_pbl_physics = 1" in WRF) -------- Options for initialization of surface conditions: initsfc - initial surface conditions: 1 = constant values (set tsk0,tmn0,xland0,lu0 below) 2 = sea breeze test case from WRF for any other value: you must initialize the surface conditions yourself in the "init_surface.F" file. tsk0 - default initial value for "skin temperature" (K) of soil/water (~1 cm deep) NOTE: this replaces sea surface temperature (tsurf) in cm1r15 tmn0 - default initial value for deep-layer temperature (K) of soil (Note: remains fixed throughout simulation) (only used if sfcmodel=2) (only used over land ... ignored over water) xland0 - default initial value for land/water flag: 1 for land, 2 for water lu0 - default initial value for land-use index (see LANDUSE.TBL file) (NOTE: for water/ocean, use lu0 = 16) season - which set of land-use conditions to use from LANDUSE.TBL file: 1 = summer values 2 = winter values c-------------------------------------------------------------c To reiterate: if you want to use spatially varying values of tsk,tmn,xland,lu then you must code it up yourself in the "init_surface.F" file. c-------------------------------------------------------------c -------- Options for sfcmodel = 1: cecd - When idrag=1 or isfcflx=1, this allows the user to choose the formulation for the surface exchange coefficients for enthalphy (Ce) and momentum (Cd). Options are: 1 = constant value: user must set cnstce and/or cnstcd below 2 = Deacon's formula [eg, Rotunno and Emanuel (1987, JAS)] (over water only) default: 3 = Cd based roughly on Fairall et al (2003) at low wind speeds and Donelan (2004, GRL) at high wind speeds Ce constant, based on Drennan et al. (2007, JAS) (over water only) pertflx - Use only perturbation winds for calculation of surface fluxes? (0=NO, 1=yes) (Only available with sfcmodel=1) cnstce - Constant value of Ce (surface exchange coefficient for enthalpy) if isfcflx=1 and cecd=1 cnstcd - Constant value of Cd (surface exchange coefficient for momentum) if idrag=1 and cecd=1 -------- Options for sfcmodel = 2: isftcflx - Use alternative Ck and Cd for tropical storm applications: (0=off) (For Cd: 1,2 = Donelan) (For Ce: 1=constant Z0q, 2=Garratt) (Cannot be used with sfcmodel=1) iz0tlnd - When using sfcmodel=2, option for thermal roughness length: 0 = Carlson-Boland (original mm5/wrf version) 1 = Czil_new (new: depends on vegetation height) -------- Options for oceanmodel = 2: oml_hml0 - default ocean mixed layer depth (m) at initial time oml_gamma - default deep water lapse rate (K m-1) ------------------------------------------------------------------------- param4 section -- Horizontally stretched (x) grid options. See README.stretch for more information. stretch_x - Use horizontally stretched grid in x? 0 = no 1 = yes, stretching on both west and east sides of domain 2 = yes, stretching on west side of domain only dx_inner - Smallest grid spacing (m). dx_outer - Largest grid spacing, at edge of domain (m). nos_x_len - Length of the no-stretching part of domain (m). tot_x_len - Total length of the domain (m). ------------------------------------------------------------------------- param5 section -- Horizontally stretched (y) grid options. See README.stretch for more information. stretch_y - Use horizontally stretched grid in y? 0 = no 1 = yes, stretching on both south and north sides of domain 2 = yes, stretching on south side of domain only dy_inner - Smallest grid spacing (m). dy_outer - Largest grid spacing, at edge of domain (m). nos_y_len - Length of the no-stretching part of domain (m). tot_y_len - Total length of the domain (m). ------------------------------------------------------------------------- param6 section -- Vertically stretched grid options. See README.stretch for more information. stretch_z - Use vertically stretched grid spacing? (0=no, 1=yes) ztop - Total depth of the domain (i.e., the height of the top of the domain) (m). str_bot - Level where stretching begins (m). str_top - Level where stretching ends (m). dz_bot - Grid spacing at (and below) str_bot (m). dz_top - Grid spacing at (and above) str_top (m). ------------------------------------------------------------------------- param7 section -- Options relating to dns (used ONLY when dns = 1) bc_wind - top/bottom boundary condition for horizontal winds 1 = free slip 2 = no slip bc_temp - top/bottom boundary condition for potential temperature 1 = constant theta is specified at boundaries 2 = constant flux is specified at boundaries ptc_top - potential temperature closure for top of model if bc_temp = 1, this is theta at the top boundary (K) if bc_temp = 2, this is the flux at the top boundary (K m/s) ptc_bot - potential temperature closure for bottom of model if bc_temp = 1, this is theta at the bottom boundary (K) if bc_temp = 2, this is the flux at the bottom boundary (K m/s) viscosity - value for kinematic viscosity (m^2/s) pr_num - value for Prandtl number (unitless) ------------------------------------------------------------------------- param8 section - flex variables var1,var2,var3 ... var10 - Use these variables to easily change parameters in the model without re-compiling the code. Example: A user wants to run a series of simulations in which the low-level shear is changed. In file "base.F", the user can set "uconst2 = var1", then compile the code. Then, var1 can be changed in the namelist.input file and the initial wind profile will change without needing to re-compile. Example: A user wants to change the location, size, and amplitude of the initial thermal bubble. Using the flex vars, the code could be modified in this manner: ric = var1 rjc = var2 zc = var3 bhrad = var4 bvrad = var5 bptpert = var6 and then the values in namelist.input might be set as follows: var1 = 50000.0, var2 = 25000.0, var3 = 2000.0, var4 = 10000.0, var5 = 1000.0, var6 = 1.5, Notes: - the flex variables are all real (float) variables. - the flex variables are automatically passed into most subroutines in the model ------------------------------------------------------------------------- param9 section - output options output_path - Specifies the path where the output will be placed. By default, the output is placed in the same directory as the executable (cm1.exe), by specifying './' IMPORTANT NOTE: Always end the path with a "/". output_basename - Specifies the base file name for output. By default, all files will begin with 'cm1out' output_format - Specifies the format of the output files: 1 = GrADS format 2 = netcdf (note: GrADS users might want to use the COARDS convention for netcdf output ... set coards = .true. in writeout_nc.F) 3 = HDF5: scale-offset lossy compression followed by gzip compression (smallest files) 4 = HDF5: gzip compression (level 1) on floating point data and is lossless 5 = HDF5: no compression whatsoever, just floating point (largest files) (see top of writeout_hdf.F file for explanation of the difference between format 3,4,5) (see also this site at the CM1 Users Group webpages http://groups.google.com/group/cm1-users-group/web/a-brief-hdftools-primer for more information about using the HDF output) output_filetype - Type of output file: 1 = all output goes into one file (note: the file size can become very big!) 2 = one output file per output time (note: good to use if you want to keep individual file sizes relatively small ... but this produces many files) 3 = for MPI runs only: one output file per output time ...AND... one output file per MPI process (note: creates many output files that need to be combined together using special code ... see, eg, http://www.mmm.ucar.edu/people/bryan/cm1/programs/) but this is the most efficient way to write output for MPI runs with very large number of processes (say, > 200). NOTE: for HDF output, option 3 is required output_interp - For simulations with terrain, this option (0=no, 1=yes) will generate a second set of output files, with "_i" as part of the filename, wherein the output has been interpolated to the nominal model height levels. This differs from all other model output files, for which the output is on the native terrain-following coordinate surfaces. (only available for output_format=1) ------ For the remaining variables, 0=no and 1=yes ------ output_rain - surface rainfall. If imove = 1, two output fields are generated. The first (rn) is the accumulated rainfall at model grid points. The second (rn2) is the translated rainfall pattern, assuming a lower surface is moving at umove and vmove. output_sws - maximum surface wind speed (aka, surface wind swath, sws) If imove = 1, two output fields are generated. The first (sws) is the max sws at model grid points. The second (sws2) is the translated sws pattern, assuming a lower surface is moving at umove and vmove. When output_sws = 1, a whole bunch of additional "swaths" are calculated: svs = maximum vertical vorticity at lowest model level sps = minumum pressure at lowest model level srs = maximum rainwater mixing ratio at lowest model level sgs = maximum graupel/hail mixing ratio at lowest model level sus = maximum w at 5 km AGL (i.e., maximum updraft swath) shs = maximum integrated updraft helicity swath output_coldpool - properties of surface-based cold pools: cpc = cold pool intensity, C cph = cold pool depth, h NOTE: the reference profile (for calculation of buoyancy) is simply the initial sounding. output_sfcflx - surface fluxes of potential temperature and water vapor, and Ce/Cd (if isfcflx=1) output_sfcparams - parameters used in the surface/soil/ocean models output_sfcdiags - diagnostics from surface-layer parameterization (eg, 10-m winds, 2-m temp/moisture, roughness length, etc) output_zs - terrain height? output_zh - height on model levels? output_basestate - output the base-state arrays? output_th - potential temperature? output_thpert - potential temperature perturbation? output_prs - pressure? output_prspert - pressure perturbation? output_pi - nondimensional pressure (ie, Exner function)? output_pipert - nondimensional pressure perturbation? output_rho - dry air density? output_rhopert - dry air density perturbation? output_tke - subgrid turbulence kinetic energy? output_km - subgrid eddy viscosity? output_kh - subgrid eddy diffusivity? output_qv - water vapor mixing ratio? output_qvpert - perturbation water vapor mixing ratio? output_q - liquid and solid water mixing ratios? (and number concentrations for double-moment schemes) output_dbz - reflectivity (dBZ)? Only available for ptype=2,3,5 (Goddard-LFO, Thompson, Morrison schemes, respectively) also outputs composite reflectivity (cref) (i.e., max reflectivity in the column) ---- all variables above here are in the scalar file (_s) --- output_u - u-velocity? (_u file) output_upert - u-velocity perturbation? (_u file) output_uinterp - u-velocity interpolated to scalar points? (_s file) output_v - v-velocity? (_v file) output_vpert - v-velocity perturbation? (_v file) output_vinterp - v-velocity interpolated to scalar points? (_s file) output_w - w-velocity? (_w file) output_winterp - w-velocity interpolated to scalar points? (_s file) output_vort - vorticity? (all three components, interpolated to scalar points) (_s file) (not available when using terrain ... at least for now) output_uh - vertically integrated (1-6 km AGL) updraft helicity? (Kain et al, 2008, WAF, p 931) output_pblten - tendencies from PBL scheme? (only if ipbl=1) output_dissten - dissipative heating tendency? (only if idiss=1) output_radten - radiative tendencies? (only if radopt=1) ------------------------------------------------------------------------- param10 section - statistical output options (0=no, 1=yes) stat_w - max/min vertical velocity stat_u - max/min horizontal (x-direction) velocity (radial velocity if axisymm=1) stat_v - max/min horizontal (y-direction) velocity (azimuthal velocity if axisymm=1) stat_rmw - radius of maximum azimuthal velocity (for axisymmetric simulations only) stat_pipert - max/min perturbation nondimensional pressure stat_prspert - max/min perturbation pressure stat_thpert - max/min perturbation potential temperature stat_q - max/min moisture variables stat_tke - max/min subgrid turbulence kinetic energy stat_km - max/min turbulence coefficient for momentum stat_kh - max/min turbulence coefficient for scalars stat_div - max/min divergence stat_rh - max/min relative humidity (wrt liquid) stat_rhi - max/min relative humidity (wrt ice) stat_the - max/min equivalent potential temperature stat_cloud - max/min cloud top/bottom stat_sfcprs - max/min pressure at lowest model level stat_wsp - max/min wind speed at lowest model level (includes 10-m windspeed if idrag=1) stat_cfl - max Courant number - also prints KSHMAX and KSVMAX (analysis of numerical stability for diffusion terms) (NOTE: KSHMAX should ideally be less than 0.125 for 3d runs) stat_vort - max vertical vorticity at several levels stat_tmass - total dry-air mass stat_tmois - total moisture stat_qmass - total mass of moisture variables stat_tenerg - total energy stat_mo - total momentum stat_tmf - total downward/upward mass flux stat_pcn - precipitation/moisture statistics stat_qsrc - sources of moisture mass ------------------------------------------------------------------------- Questions, comments, suggestions: send email to gbryan@ucar.edu