Turned back on use of Level 2.5 NEXRAD data in 3DVAR, only sites with good lat/lon data (about 15-20% of original total) are included in the data dump.
Absorption coefficient for ice is made to be the same as in operations.
Reset the maximum number concentration of large ice back to 20 per liter as in operations.
Increased the minimum longwave emissivity for snow from 0.9 to 0.95.
||Bender, Keyser, Rogers
Revert back to ops radial wind data due to 75-85% of
the Level 2.5 supob reports in /dcom/us007003/*/b006/xx002 having
a missing superob longitude (SUPLON) and a superob latitude (SUPLAT)
equal to 65.530 (for every report with a missing SUPLON). It
appears that something is corrupt in the incoming Level 2.5 raw
files from which SUPLAT and SUPLON are generated.
||Ferrier, Ek, Mitchell
Changed the standard deviation of RH for grid-scale condensation from 2% to 1%
Changed "RHsat=.5*(RHgrd+H1)" to "RHsat=RHgrd", which will have the effect of an
earlier onset for partial cloudiness, which will be mitigated by the change in the first bullet.
These changes make the cloud cover more "binary"
Added the effects of snow emissivity in the calculation of effective snow-ground
sfc temperature. This was done by changing the value from 1.0 to 0.9.
Adopted only one of the EtaL changes by decreasing the
maximum number concentration of precipitation ice particles from 20 to 10
per liter. Also, the solar absorption by ice is a little less than a half of
what's parameterized in operations. These changes are expected to reduce
solar absorption by upper-level ice clouds and to increase the transmission of
solar radiation into the lower troposphere, in which the goal is to reduce the
warm (cold) bias at upper (lower) levels. A slight change was made in the order
that solar zenith angle is called at the beginning of the radiation driver will
speed up the code slightly but should not impact the forecast.
Changed the assumed number concentration of cloud droplets
from 100 back to 200 /cm**3, which increases the threshold cloud water
mixing ratio for autoconversion to rain from 0.838. Also increased the
maximum number concentration of precipitation ice particles from 10
back to 20 per liter.
Changed cloud water absorption coefficients back to operational
Eta values. The absorption coefficients for ice are unchanged and are
slightly less than what's currently used in operations.
Decreased the assumed number concentration of cloud droplets from
200 to 100 /cm**3, which reduces the threshold cloud water mixing ratio
for autoconversion to rain from 0.838 to 0.419 g/m**3. Also decreased
the maximum number concentration of precipitation ice particles from 20
to 10 per liter. Both changes are intended to lower the amount of
suspended condensate in the atmosphere, which should help reduce the
cool bias that has developed below clouds in the parallel.
Fixed a minor bug in ADJPPT1 that limits the amount of cloud water
from exceeding the autoconversion threshold
A new cloud cover scheme is introduced in order to increase
the presence of forecast partial cloudiness. Two adjustable
parameters in the scheme have been tuned to match the AFWA total
cloud cover product. The first parameter (STSDM) is the assumed
standard deviation of total grid-scale relative humidity in the grid
box, and it is set to 2%. The second parameter (RHsat) is the total
relative humidity associated with an assumed cloud fraction of 50%,
and it is set to the average of the threshold relative humidity for
the onset of grid-scale condensation (RHgrd) and a value of 100%
(i.e., 98.3% for the 32-km runs).
Increase minimum stomatal conductance from 40s/m to 70s/m for
cropland, pasture, and grassland vegetation types in order to decrease
In calculating longwave emissivities in cloudy layers, the
optical depths for cloud water are obtained using the downwelling
longwave relationships of Smith and Shi (1992). The LW absorption
coefficient for ice to be the same as what's currently used in the
In calculating solar absorption, the optical depths for
cloud water and for ice (cloud ice and snow) are obtained using the
relationships described by eqs. (5.2), (5.3), and Table 9 from
Hou et al. (2002), assuming a constant effective radius of 10 microns
for cloud water and 75 microns for ice. For cloud water optical depths,
equal weighting is assumed for absorption of UV-VIS and near IR radiation.
In the operational radiation, a minimum optical depth is
assumed for grid-scale liquid water clouds consistent with a minimum
mixing ratio of 0.1 g/kg. This lower limit has been removed, and this
change has been found to have the biggest impact.
Replaced GFS radiation with ops Lacis-Hansen scheme with
modified water & ice absorption coefficients that are
more consistent with those in the GFS radiation scheme
Revert back to operational version of cloud microphysics
Restarted cycled EDAS from ETAZ
Changed minimum size for precipitation ice particles back to 100
microns, as well as using the same temperature dependence for
precipitation ice, reverting back to the formulation prior to
Increased the temperature at which liquid water is assumed to
glaciate completely to ice from -25C back to -10C, reverting back to
what's used in operations.
Leaf area index (LAI) parameter in LSM physics changed from 1.0 to 2.0
||Ferrier, Ek, Mitchell
The solar constant is reduced by at most 2.12% rather than by 4.74%.
The minimum effective radius for cloud water is increased from 5 microns to 10 microns.
The minimum size assumed for precipitation ice particles is increased from 100 microns
to 150 microns. The temperature dependence assumed for the size of these particles
was also changed so that at a temperature of -40C the mean diameter increased from
117 microns to 201.9 microns.
Removed all references to cloud water and ice (cloud ice + snow) paths for convective
cloudiness in RADTN, RDLWSW, and SWR95. Instead, cloud water and ice paths for
convective clouds are merged with grid-scale cloudiness, but only if the convective
cloud fractions are larger than the grid-scale cloud fractions at each model level.
Cloud water mixing ratios for convection are assumed to be 0.05 g/kg, and ice mixing
ratios for convection are assumed to be the minimum of 0.05 g/kg or the ice mixing
ratio associated with a large particle concentration of 5 per liter.
In Noah LSM, LAI_DATA=1.0, SMHIGH=3.0
EDAS restarted from ETAZ
Begin use of new 3DVAR analysis which uses a 2dvar surface analysis
to assimilate surface temperature data.
Minimum size for precipitating ice particles is increased from
50 microns to 100 microns. This change will reduce the absorption of
solar radiation by upper-level ice clouds based on tests in the EtaL.
Change "LAI_DATA" in LSM physics from 4.0 to 3.0
Change "LAI_DATA" in LSM physics from 5.0 to ops value of 4.0
Revert to operational version of shallow convection
parameterization in CUCNVC
Restarted cycled EDAS from ETAV atmospheric states and ETAZ soil states
Revert to operational versions of PRODQ2 and MIXLEN, removing
changes to these routines which removed mixing
length limits under stable conditions.
New precipitation assimilation scheme (routine ADJPPT) modified
to account for the fact that the variable F_rain is actually the fraction
of liquid condensate that is rain, not the fraction of total condensate that
is liquid rain