Mesoscale Parallel Experiment Change Log

Experiment Name:  Eta-32 with GFS radiation physics

When Who What
2004/08/02/12 Ferrier
  • A lower limit of 4% is assumed for the standard deviation of total relative humidity in the new cloud cover scheme. The impact will be to increase the occurrence of mostly cloudy and total overcast conditions to be more consistent with preliminary FVS statistics compared with the AFWA total cloud product.
  • 2004/07/06/12 Ferrier
  • Code changes:
    1. RADTN: eliminates the inflation of ozone concentrations in the operational Eta
    2. RADLSW: revert back to an older version of code
    3. SWR95:
      1. Solar constant is reduced to account for the absence of the main ozone layer above the 25-hPa top of the Eta domain, which is the following function of the cosine of the zenith angle (cosz):

        X=1.-.015/sqrt(max(0.1, cosz(i)))

      2. The solar constant is reduced by 1.5% when the sun is at nadir (cosz=1), 2.12% when the sun angle is 60 deg from the vertical (cosz=0.5), 3.35% when the sun angle is 78.5% from the vertical (cosz=0.2), and 4.74% (maximum reduction) when the sun angle is at or greater than 84.3% from the vertical (cosz=0.1). The formula for reducing the solar constant is an approximation to offline calculations by Y.-T. Hou based on ozone concentrations used in the GFS model. The treatment for the optical thickness of snow was changed to be similar to that used for cloud ice crystals, except that a maximum limit of 150 microns is used for the effective radius of snow.
  • 2004/06/25/12 Ferrier
  • Condensate from convective clouds is created only when the cloud fraction from convection exceeds that of grid-scale cloudiness at a particular level. Under these conditions, the mixing ratio of ice from convection is limited to what's in grid-scale cloudiness in association with an assumed maximum ice number concentration of NLImax (which was recently reduced from 20 to 5 per liter). This change should reduce the amount of ice, particularly at upper levels, which should help reduce the warm bias.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/06/23/12 Ferrier
  • The maximum number concentration of large ice particles ("snow") was reduced from 20 per liter to 5 per liter in order to reduce the mass of suspended ice in the hopes of reducing the upper-level warm bias in the parallel.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/06/09/12 Ferrier
  • Changes to AEROS.F, RADTN.F : 1) dramatic reduction of the fractional cloud cover at mid-to-upper levels, creating more occurrences of partially cloudy skies, 2) limit the input relative humidity to a range from 0 to 100% when calculating aerosol optical depths.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/05/28/12 Y. Lin, Ferrier
  • Fixed array bounds errors in ADJPPT.f
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/05/26/12 Ferrier
  • Microphysics changes
    • Temperature at which supercooled water is allowed (T_ICE parameter) was changed from -30C to -25C
    • Cloud droplet number concentration was increased from 75 /cm**3 to 100 /cm**3.
  • Radiation changes
    • Cloud droplet effective radius has an upper limit of 20 microns.
    • Assorted parameter changes were made to the new cloud cover parameterization in order to decrease radiative forcing by convective and grid-scale cloudiness.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/05/21/12 Ferrier, H.-M. Lin
  • In the solar radiation calculations, a fixed effective radius for ice of 75 microns was replaced with more sophisticated calculations using what's parameterized in the grid-scale microphysics. Single scattering albedos for snow were corrected to be consistent with functional relationships originally suggested by Prof. Q Fu (U. Washington).
  • Aerosol scattering effects were put back into clouds.
  • Cloud optical properties assumed for convective clouds were combined with grid-scale microphysics. Code in the radiation driver was restructured and streamlined to consider radiative impacts from prognostic clouds, removing all codes related to diagnostic cloud inputs.
  • Minor changes were made to increase parameterized cloud amounts. More substantial and controversial changes were made to the parameterized water paths for various types of hydrometeors based on the parameterized cloud amounts, and these may be modified or removed based on verification results.
  • The assumed cloud droplet number concentration was increased back to 200 /cm**3 (what's used in operations) and the temperature at which cloud water is completed glaciated to ice was lowered from -10C to -30C.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • 2004/05/13/12 Ferrier
  • Reverted back to operational shallow convection due to higher RMS errors vs. raobs over the CONUS compared to Alaska below 200 mb, and an increasing dry bias (esp. RH at 850, 700 mb) with time.
  • EDAS restarted from 32-km (ETAV) control EDAS.
  • Page Last Modified: May 13, 2004