CM1 version 1 release 2:  17 February 2003.

Summary of changes.

-------------------------------------------------------------
1.  New features.

- Dynamic memory allocation for water/ice/turbulence arrays.
  The program no longer allocates memory for arrays it does
  not use.  Can save significant memory resources.

- Ability for user to choose what data to output:  namelist.output

- Ability for user to choose what stats to output:  namelist.stats


-------------------------------------------------------------
2.  New options.

- Added ability to do restarts.  To save restart files, set rstfrq
  (restart frequency) switch.  Then, to restart a simulation, set
  irst=1 and set proper filename in subroutine read_restart.

- Added option to use either 5th order advection or 6th order advection
  (via advorder switch).  The 5th order scheme is implicitly diffusive:
  the diffusion is proportional to wind speed, and inversely proportional
  to grid spacing.  The 6th order scheme is not diffusive, but its use
  requires an additional artificial filter for stability (idiff option).

- Added switch for artifical diffusion term (idiff=0,1).  When idiff=1,
  user must set difforder to 2 or 6, for 2nd order or sixth order, as
  well as the diffusion constant via kdiff2 or kdiff6.

- Added Smagorinsky turbulence closure (iturb=2).  The TKE closure is
  still available (iturb=1).

- Replaced dts variable from namelist.input:  user must instead
  specify a new switch, nsound, the number of small time steps per
  large step.  Only used if kwss=1.

- Parallalized ice microphysics scheme.  Added support for new equation
  set (neweqts=1) when ice microphysics is used.

- Added switch ihail to namelist.input.  Only used when ptype=2.
  For ihail=1, the large ice category is hail (fall velocities are
  10-20 m/s).  For ihail=0, the large ice category is graupel (fall
  velocities are 2-5 m/s).

- Added new saturation adjustment scheme when using ice microphysics.
  The new scheme calculates a saturation mixing ratio that is 
  weighted between saturation wrt ice and saturation wrt water at
  temperatures between about -40 and 0 Celsius.

- Added simple surface fluxes of heat and moisture (isfcflx=1) and
  simple surface drag on winds (idrag=1).

- Separated ewbc into ebc and wbc, and nsbc into nbc snd sbc.

- Added isothermal base state sounding (isnd=2).

- Added constant lapse rate base sounding (isnd=3).

- Added an idealized tropical cyclone initialization (isnd=6, iinit=6).

- Added option to choose balance assumption for initial 3D pressure
  field.  Option 1 is hydrostatic balance, option 2 is anelastic
  balance (based on the buoyancy pressure perturbation for an
  anelastic atmosphere).

- Added domain motion into namelist.input (imove, umove, vmove switches).

- Added namelist.output.  User can select what date to output.  New
  variables are:

      - potential temperature perturbation
      - pressure perturbation
      - nondimensional pressure (pi, or Exner function)
      - nondimensional pressure perturbation
      - density
      - density perturbation
      - turbulence coefficient for momentum
      - turbulence coefficient for scalars
      - water vapor perturbation
      - u perturbation
      - v perturbation
      - u interpolated to scalar points
      - v interpolated to scalar points
      - w interpolated to scalar points

- Added namelist.stats.  User can select what stats to output.  New
  variables include:

      - max/min pressure perturbation
      - max/min surface potential temperature perturbation
      - max/min relative humidity with respect to ice
      - max/min surface equivalent potential temperature
      - max/min surface pressure
      - max/min surface wind speed
      - max vertical vorticity at:  sfc, 1 km, 2 km, 3km, 4km, and 5km


-------------------------------------------------------------
3.  Bug fixes.

- Added support for new equation set with ice microphysics scheme
  (neweqts=1 and ptype=2).

- Added OpenMP parallelization with ice microphysics scheme.

- Stats related to moisture are no longer computed when imoist=0.

- Max/min cloud top now considers ice field when ice microphysics
  is utilized.