OVERVIEW

     RAS - Relaxed Arakawa/Schubert Cumulus Parameterization Scheme

     A mass based parameterization of cumulus precipitation, 
     heating and moistening based upon the Arakawa-Schubert 
     Quasi-Equilibrium idea.
   

     Given temperature, moisture and wind profiles at the model's 
     pressure levels, this module returns tendencies of temperature, 
     moisture and wind due to cumulus convection. It also returns the
     convective precipitation rate.

     If run with the prognostic cloud scheme (strat_cloud), the
     routine converts a fraction of the condensed water to
     stratiform liquid and ice as well as increases the stratiform
     cloud fraction.

     This parameterization was developed by Moorthi and Suarez 
     (see reference), with local modifications by Joe  Sirutis and
     Steve Klein. The local modifications include: 1) re-evaporation
     of convective scale precip was added, 2) the entrainment parameter
     was made a function of cloud type, and was used in tuning the 
     moisture distribution produced by the scheme, 3) an option of 
     detraining condensed liquid and ice into stratiform cloud was
     added.
    

OTHER MODULES USED

          constants_mod
     sat_vapor_pres_mod
          utilities_mod
       diag_manager_mod
       time_manager_mod

Note: Version v2.0 was used by all modules for testing this version
      of ras_mod.

PUBLIC INTERFACE

use ras_mod [,only: ras_init, ras, ras_bdgt]

ras_init  - Called once to initialize ras. Also reads namelist.
            Must be called before ras.

ras       - Driver for ras module

ras_bdgt  - Optional debugging tool. Can be called to do budget checks
            for tendencies computed in ras.

Notes:
 * A namelist interface ( &ras_nml ) controls runtime options.

PUBLIC ROUTINES

       call ras_init (axes, Time )

input
    axes       Axes for diagnostics
        	[integer, dimension(4) ]

    Time       Time used for diagnostics 
               [time_type]

---------------------------------------------------------------------

        call ras( is,     js,      Time,    temp0,      qvap0,     &
                  uwnd0,  vwnd0,   pres0,   pres0_int,  coldT0,    &
                  dtime,  dtemp0,  dqvap0,  duwnd0,     dvwnd0,    &
                  rain0,  snow0,   kbot,    ql0,        qi0,       &
                  qa0,    mc0,     Dl0,     Di0,        Da0  )

input
    is, js     Starting indices for window
               [integer]

    Time       Time used for diagnostics 
               [time_type]

    temp0      Temperature [deg k] at full model levels,
               [real, dimension(:,:,nlev)]

    qvap0      Specific humidity of water vapor [kg/kg] at full model levels,
               [real, dimension(:,:,nlev)]

    uwnd0      Zonal wind component [m/s] at full model levels,
               [real, dimension(:,:,nlev)]

    vwnd0      Meridional wind component [m/s] at full model levels,
               [real, dimension(:,:,nlev)]

    pres0      Pressure at full levels in pascals
               [real, dimension(:,:,nlev)]

    pres0_int  Pressure at half levels in pascals
               [real, dimension(:,:,nlev+1)]

    coldT0     Should the precipitation assume to be frozen?
               [logical, dimension(nlon,nlat)]

    dtime      Time step in seconds.
               [real]

output

    dtemp0    Temperature tendency [deg k/time step] 
              [real, dimension(:,:,nlev)]

    dqvap0    Specific humidity of water vapor tendency [1/time step]
              [real, dimension(:,:,nlev)]

    duwnd0    Zonal wind component tendency [m/s/time step]
              [real, dimension(:,:,nlev)]

    dvwnd0    Meridional wind component tendency [m/s/time step]
              [real, dimension(:,:,nlev)]

    rain0     Liquid precipitation (rain) in kg/m2
              [real, dimension(:,:)]

    snow0     Frozen precipitation (snow) in kg/m2
              [real, dimension(:,:)]

input (optional) - for use with the step-mountain (eta) vertical coordinate

    kbot      Index of the lowest model level
              [integer, dimension(:,:)]

input (optional) - for use with prognostic stratiform cloud module

    ql0       Cloud liquid [kg/kg]
              [real, dimension(:,:,nlev)]

    qi0       Cloud ice [kg/kg]
              [real, dimension(:,:,nlev)]

    qa0       Cloud/saturated volume fraction 
              [real, dimension(:,:,nlev)]

output (optional) 

    mc0       Cumulus mass flux (for diagnostic purposes)
              [real, dimension(:,:,nlev)]

    Dl0       Cloud liquid change (for use with prognostic stratiform cloud module)
              [real, dimension(:,:,nlev)]

    Di0       Cloud ice change (for use with prognostic stratiform cloud module)
              [real, dimension(:,:,nlev)]

    Da0       Cloud fraction change (for use with prognostic stratiform cloud module)
              [real, dimension(:,:,nlev)]

-----------------------------------------------------------------

 call ras_bdgt ( precip,   coldT, dtemp, dqvap, duwnd, dvwnd, 
                 pres_int, dql,   dqi )

input

    precip    Liquid precipitation (rain) in kg/m2
              [real, dimension(:,:)]

    coldT     Should the precipitation assume to be frozen?
              [logical, dimension(:,:)]

    dtemp     Temperature tendency [deg k/time step] 
              [real, dimension(:,:,nlev)]

    dqvap     Specific humidity of water vapor tendency [1/time step]
              [real, dimension(:,:,nlev)]

    duwnd     Zonal wind component tendency [m/s/time step]
              [real, dimension(:,:,nlev)]

    dvwnd     Meridional wind component tendency [m/s/time step]
              [real, dimension(:,:,nlev)]

    pres_int   Pressure at half levels in pascals
               [real, dimension(:,:,nlev+1)]

input (optional) 

    dql       Cloud liquid change 
              [real, dimension(:,:,nlev)]

    dqi       Cloud ice change 
              [real, dimension(:,:,nlev)]

Notes:
  * The optional arguments qldt and qidt must be supplied if
    strat_cloud is running to get the actual moisture budget.

NAMELIST

&ras_nml 

   fracs        Fraction of PBL mass allowed to be used 
                by a cloud-type in time DT
                [real, default: fracs = 0.25 ]

   rasal0       Base value for cloud type relaxation parameter
                [real, default: rasal0 = 0.25 ]

   puplim       Upper limit for cloud tops of deepest clouds
                [real, default: puplim = 20.0E2 Pascals]

   aratio       Ratio of critical cloud work function to standard
                value of cloud work function
                [real, default: aratio = 1.4 ]

   cufric       Should Cumulus friction (momentum transport) occur?
                [logical, default: cufric = .false. ]

   rh_trig      Convection takes place only if the relative humidity
                of the lowest model level exceeds rh_trig
                [real, default: rh_trig = 0.0 ]

   alm_min      Min value for entrainment parameter.
                [real, default: alm_min = 0.0 ]

   Tokioka_on   If true, alm_min computed using Tokioka formulation
                [logical, default: Tokioka_on = .false. ]

   Tokioka_con  Constant for alm_min computed using Tokioka formulation
                [real, default: Tokioka_con = 0.05 ]

Tokioka_plim   Tokioka applied only to clouds detraining above Tokioka_plim
                [real, default: Tokioka_plim = 500.0E2 ]

   modify_pbl   If true, mass flux in sub cloud layer varies linearly
                between value at cloud base and zero at surface, and
                tendencies are spread out throughout the entire sub cloud layer.
                [logical, default: modify_pbl = .false. ]

 --- Evaporation of convective scale precip 

   evap_on      Turn on evaporation if true
                [logical, default: evap_on = .true. ]

   cfrac        Fraction of grid box assumed to be covered by convection
                [real, default: cfrac = 0.05 ]

   hcevap       Evap allowed while q <= hcevap * qsat
                [real, default: hcevap = 0.80 ]

 --- Partitioning of cloud liquid water into precip, detrained vapor & liquid

   rn_ptop      rn_frac_top of parcel liquid water converted to precip 
                for cloud top pressures above rn_ptop
                [real, default: rn_ptop = 500.0E2 Pascals]

   rn_pbot      rn_frac_bot of parcel liquid water converted to precip 
                for cloud top pressures below rn_pbot  
                (linear profile in between)
                [real, default: rn_pbot = 800.0E2 Pascals]

   rn_frac_bot  Fraction of parcel liquid water converted to 
                precip for cloud top pressures below rn_pbot
                [real, default: rn_frac_bot = 0.8 ]

   rn_frac_top  Fraction of liquid water converted to
                precip for cloud top pressures above rn_ptop
                [real, default: rn_frac_top = 1.0 ]

 --- Cloud order specification

   ncrnd        Number of random cloud-types between krmin and krmax
                to be invoked in a single call to ras
                [integer, default: ncrnd = 0 ]

   iseed        Integer seed used in generating random numbers
                -- used only when ncrnd > 0
                [integer, default: iseed = 123 ]

   krmin        Index of the top most level to which random clouds may
                be invoked
                [integer, default: krmin = 2 ]

   krmax        Index of the bottom most level to which random clouds 
                may be invoked. krmin should be <= krmax. 
                If ncrnd is specified as zero, then all cloud-types 
                below the level krmax will be called sequentially.
                [integer, default: krmax = 2 ]

   botop        A logical variable -- .true. if sequential clouds are 
                called from bottom to top and .false. if top to bottom.
                Level krmax will be called sequentially.
                [logical, default: botop = .true. ]

Notes:

   * For parameters listed in the "Cloud order specification" section above,
     very limited testing has been done with values other than the defaults.

   * Diagnostics of RAS are controlled by the namelist to moist_processes
     module.

DIAGNOSTIC FIELDS

Diagnostic fields may be output to a netcdf file by specifying the
module name ras and the desired field names (given below)
in file diag_table. See the documentation for diag_manager.

Diagnostic fields for module name: ras

   field name   field description
   ----------   -----------------
   tdt_revap    temperature tendency from re-evaporation of convective
                precip [deg_K/s]
   qdt_revap    specific humidity tendency from re-evaporation of convective
                precip [kg/kg/s]
  prec_revap    total convective precip evaporated [kg/m2/s]
  snow_revap    frozen convective precip evaporated [kg/m2/s]
  prec_conv_3d  3D precipitation rate  [kg/m2/s]
  pcldb         Pressure at cloud base from RAS

CHANGE HISTORY

ERROR MESSAGES

Fatal errors in ras:

    ras_init has not been called
       You have not called ras_init before calling ras.

Fatal errors in ras_bdgt:

    RAS BUDGET CHECK AT i,j =
       The enthalpy, and/or moisture, and/or momentum budgets did 
       not balance. This usually happens for one of two reasons:
       1) "bad" data is being passed into ras; 2) a code error ( or 
       inconsistency ) has been introduced into the ras module. 
          

REFERENCES

   Moorthi, S., and M. J. Suarez, 1992: Relaxed Arakawa-Schubert:
         A Parameterization of Moist Convection for General Circulation
         Models. Mon. Wea. Rev., 120, 978-1002.

KNOWN BUGS

     There are no known bugs.

NOTES

   Developed by Shrinivas Moorthi & Max J. Suarez
   Original code dated: 04 August 1992.
   First Fortran 90 version created by Joe Sirutis in April 1997.

FUTURE PLANS