doc file - University Corporation for Atmospheric Research

advertisement
INCORPORATING GLOBAL REAL-TIME SURFACE FIELDS INTO MM5
AT THE AIR FORCE WEATHER AGENCY
George A. Gayno and Jerry Wegiel
Air Force Weather Agency
Offutt AFB, NE 68113
1. Introduction
The Air Force Weather Agency (AFWA) has
been running its agriculture meteorology model
(AGRMET) operationally for over ten years.
AGRMET produces global agricultural-weather
data, such as soil moisture and temperature, in nearreal time for use by the Air Force and the U.S.
Department of Agriculture to predict global grain
production. Recently, AFWA has been working to
initialize the land-surface module in its operational
MM5 system (MM5-V3-R2) with AGRMET data.
Because AFWA runs MM5 grids throughout the
world, the impact of AGRMET data is being
evaluated for a variety of climate regimes.
This note describes the AGRMET model and
the process of ingesting AGRMET data into MM5
at AFWA. Experiments using AGRMET data to
initialize MM5’s land surface module are being
compared to production runs using MM5’s simple,
climatological treatment of the ground. These
results will be presented at the workshop.
2. AGRMET Description
AFWA runs its AGRMET model in an
uncoupled mode.
In other words, the soil
hydrology module is “forced” with analyses of
precipitation, short and longwave radiation at the
ground, and shelter height fields of temperature,
moisture and wind speed. Forcing land-surface
modules (LSMs) with analysis fields eliminates the
errors associated with coupling them to
atmospheric models, which often have biases in
their prediction of precipitation and other fields.
AGRMET runs on two polar stereographic
grids that resolve all major land areas of the
northern and southern hemispheres. These grids
have i/j dimensions of 512 by 512, are true at 60
_________________________________________
*Corresponding author: George Gayno, 106
Peacekeeper Dr, Ste 2N3, Offutt AFB, NE 68113.
Email: george.gayno@afwa.af.mil Phone: (402) 2943863, Fax: (402) 294-2246.
degrees, and have a horizontal grid spacing of 48
km. AGRMET is vertically configured with four
soil layers with thicknesses of 10, 30, 60 and 100
cm. These layers were chosen to coincide with
those used by MM5’s LSM. AGRMET is run
twice a day - at 00 and 12 UTC - and uses one-hour
time steps.
a.
Soil Hydrology Module
The core of the AGRMET model is the NOAH
soil hydrology module (Chen et. al, 1996). The
acronym “NOAH” represents the four agencies
that developed it – the National Center for
Environmental Prediction, Oregon State University,
the U.S. Air Force, and the Office of Hydrology.
The NOAH module is similar to MM5’s LSM
(Chen and Dudhia, 2000) with the following
exceptions: (1) The drag coefficient for heat is a
function of a flow-dependent thermal roughness
length (Chen et. al, 1997). (2) It uses different
algorithms for calculating the thermal diffusivity
and conductivity of the soil (Peters-Lidard et. al,
1998). (3) It includes frozen soil physics (Koren et.
al, 1999).
b.
Preprocessing
AGRMET derives hourly shelter height
analyses of temperature, relative humidity and wind
speed from surface observations and U.S. Navy
one-degree NOGAPS MVOI analyses. First, it
interpolates NOGAPS analyses at 00, 06, 12 and 18
UTC to the AGRMET grid. AGRMET then
temporally interpolates these six-hourly fields to
obtain hourly first guess fields, which it then blends
with surface observations using a Barnes (1964)
technique. These shelter height fields are used by
the soil hydrology module to calculate parameters
such as the sensible and latent heat fluxes at the
ground, and the drag coefficients for heat and
momentum.
AGRMET calculates the net short and
longwave radiation at the ground, including the
effects of clouds and snow cover, using the
methods of Shapiro (1987) and Idso (1981)
respectively. It retrieves analyzed cloud amounts
and types for four levels in the atmosphere from
AFWA’s real-time nephanalysis model (RTNEPH)
(Hamill et. al, 1992). The RTNEPH produces a
global cloud analysis from visible and IR satellite
data every three hours at a horizontal grid spacing
of 48 km. The surface albedo is a function of the
observed snow cover as analyzed by AFWA’s snow
depth analysis (SNODEP) model (Kopp and Kiess,
1996). SNODEP is run once per day at a horizontal
grid spacing of 48 km. It uses surface observations
and satellite data to produce a global analysis of
snow depth and sea ice. Snow depth is also an
important input to the soil hydrology module.
Precipitation is by far the most important input
to AGRMET’s soil hydrology physics. AGRMET
uses a blend of rain gauge observations and several
estimates to create global three-hourly precipitation
analyses. Rain gauge observations are used first.
However, in many parts of the world AGRMET
must rely solely on estimates. The methods used to
estimate precipitation are described below and are
listed in order of accuracy or “rank.” In lieu of a
rain gauge report, each grid point will be assigned
the highest-ranking estimate available.
1) SSM/I Estimate – AGRMET has access to
Special Sensor Microwave/Imager (SSM/I) derived
rain rates from three DMSP satellites. It uses
SSM/I estimates only in the tropics.
cloud cover exceeds climatology by a specified
threshold, AGRMET assumes precipitation has
fallen. This precipitation amount is based on
climatology.
5) Climatological Estimate – If there is no rain
gauge observation at a grid point, or none of the
higher-ranking estimates is available, AGRMET
will use a pure climatological precipitation value.
c.
AGRMET specifies nine soil types after the
one-degree Goddard Institute for Space Studies
(Columbia Univ.) database of Zobler (1986), and
thirteen vegetation types after the one-degree
database of Dorman and Sellers (1989). Albedo is
a function of the observed snow cover, and varies
between the snow-free value (one-degree seasonal
NCEP database) and a satellite-derived deep-snow
value specified after Robinson and Kukla (1985).
AGRMET simulates the annual growth and decay
of plants through the use of climatological green
vegetation fraction data (Gutman and Ignatov,
1998). This monthly, 0.15-degree greenness data is
temporally interpolated to the current day at the
start of each AGRMET cycle. AGRMET specifies
its lower thermal boundary condition by assuming
an invariant slab temperature equal to the annual
mean air temperature.
d.
2) Present Weather Estimate – Occasionally, a
surface observation will not report precipitation or
will report it as “missing.”
In these cases,
AGRMET estimates an amount based on the
current weather or the weather that occurred during
the past 6 or 12 hours. This estimate depends
largely on the weather type (thunderstorm, drizzle,
etc.), but it is also a function of the season and the
observation’s latitude and WMO block number.
3) Geostationary Satellite Estimate – AGRMET
determines an initial rain rate from the sensed
brightness temperature using a modified version of
the NESDIS Flood Homepage algorithm (Vicente,
1997). This initial rain rate is increased/decreased
if the sensed point is colder/warmer than its
neighbors or if it has cooled/warmed over time.
AGRMET has access to data from four
geostationary satellites that provide nearly global
coverage.
4) RTNEPH Estimate – AGRMET compares the
analyzed RTNEPH cloud amount with the
climatological RTNEPH amount. If the analyzed
Static fields
Output data
AGRMET outputs the following fields, every
three hours, for use in initializing MM5’s LSM:
- Soil temperature (Kelvin)
- Soil moisture (volumetric)
- Canopy moisture content (m)
- Snow liquid equivalent (mm)
The data are written in GRIB format for processing
by MM5’s PREGRID utility.
3. Ingesting AGRMET data into MM5
AGRMET outputs volumetric soil moisture,
which is intimately related to the specified soil
type. Currently, AGRMET and MM5 use different
soil type databases. This inconsistency precludes
the use of a simple interpolation of volumetric soil
moisture from the AGRMET to MM5 grids, as in
done in REGRID. Therefore, for these tests the
standard REGRID approach for interpolating soil
moisture was slightly modified.
First, the AGRMET soil moisture was
converted from volumetric to “relative” units using
the maximum value (porosity) and wilting point
(both specified according to the soil type) as
follows:
wrel 
wvol  wwlt
wmax  wwlt
This relative soil moisture field, which varies
between 0 and 1, was then mapped to the MM5 grid
by REGRID using its simple four-point
interpolation. After interpolation by REGRID, the
relative soil moisture on the MM5 grid was
converted back to volumetric units using the
maximum and wilting values as specified according
to MM5’s soil type database.
REGRID interpolated AGRMET’s soil
temperature, canopy moisture content and snow
liquid equivalent to the MM5 grid with no
modifications.
4. Impact of AGRMET data on MM5 forecasts
The MM5 test runs were configured similarly
to AFWA’s production runs: 45 km resolution,
MRF pbl scheme, Grell cumulus scheme, Reisner
mixed-phase explicit moisture microphysics, and
Dudhia cloud radiation. The production runs,
however, used the five-layer soil model with
climatological surface characteristics specified
according to the 24-category PSU/NCAR land-use
database.
Runs from polar, mid-latitude and
tropical locations were compared. In particular,
predictions of precipitation, radar reflectivity, and
low-level temperature, wind and humidity were
evaluated and will be presented at the workshop.
In summary, the coupled modeling system
does a better job of reproducing rainfall
distribution, low-level cloudiness and surface fields
of temperature, dew point and relative humidity in
drought stricken and snow covered regions.
Testing and evaluation at much finer resolutions are
continuing at this time. The target implementation
date for running the coupled LSM-MM5 modeling
system in production mode at AFWA is 30
November 2000.
REFERENCES
Barnes, S., 1964: A technique for maximizing
details in numerical weather map analysis, J.
Appl. Meteor., 3, 396-409.
Chen, F. and J. Dudhia, 2000: Coupling an
advanced land-surface/hydrology model with
the Penn State/NCAR MM5 modeling system.
Part I – Model description and implementation.
Submitted to Mon. Wea. Rev.
Chen, F., Z. Janjic, and K. Mitchell, 1997: Impact
of atmospheric surface-layer parameterizations
in the new land-surface scheme of the NCEP
mesoscale ETA model. Bound-Layer Meteor.,
85, 391-421.
Chen, F., K. Mitchell, J. Schaake, Y. Xue, H.L.
Pan, V. Koren, Q.Y. Duan, M. Ek, and A. Betts,
1996: Modeling of land surface evaporation for
four schemes and comparison with FIFE data. J.
Geophys. Res., 101 (D3), 7251-7268.
Dorman, J.L. and P.J. Sellers, 1989: Climatology of
albedo, roughness length and stomatal resistance
for atmospheric general circulation models as
represented by the simple biosphere model
(SiB). J. Appl. Meteor., 28, 833-855.
Gutman, G. and A. Ignatov, 1998: The derivation
of green vegetation fraction from
NOAA/AVHRR for use in numerical weather
prediction models. J. Remote Sensing, 19, 15331543.
Hamill, T.M., R.P. d’Entremont and J.T. Bunting,
1992: A description of the Air Force real-time
nephanalysis model. Wea. Forecasting, 7, 288306.
Idso, S., 1981: A set of equations for the full
spectrum and 8- to 14- micron and 10.5- to 12.5
thermal radiation from cloudless skies. Water
Resources Research, 17, 295-304.
Kopp, T.J. and R.B. Kiess, 1996: The Air Force
Global Weather Central snow analysis model.
Preprints, 15th Conf. on Weather Analysis and
Forecasting, Norfolk, VA, Amer. Meteor. Soc.,
220-222.
Koren, V., J. Schaake, K. Mitchell, Q.Y. Duan, F.
Chen, and J.M. Baker, 1999: A parameterization
of snowpack and frozen ground intended for
NCEP weather and climate models. J. Geophys.
Res. - Accepted.
Peters-Lidard, C.D., E. Blackburn, X. Liang, and
E.F. Wood, 1998: The effect of soil thermal
conductivity parameterization on surface energy
fluxes and temperatures. J. Atmos. Sci., 55,
1209-1224.
Robinson, D.A., and G. Kukla, 1985: Maximum
surface albedo of seasonally snow-covered lands
in the Northern Hemisphere. J. Climate. Appl.
Meteor., 24, 402-411.
Shapiro, R., 1987: A simple model for the
calculation of the flux of direct and diffuse solar
radiation through the atmosphere. AFGL-TR87-0200, Air Force Geophysics Lab, Hanscom
AFB, MA.
Vicente, G., 1987: Satellite rainfall estimates in real
time through the internet for applications to
flash flood watches, warnings and heavy
precipitation forecasting. Conf. on Earth
Observation and Environmental Information
(EOIE’97), Alexandria, Egypt, 13-16 Oct.
Zobler, L., 1986: A world soil file for global
climate modeling. NASA Technical
Memorandum #87802.
Download