|
|
Weather Forecast Model Component
The AOAWS will feature a numerical
weather prediction capability provided for by the
Pennsylvania State University/NCAR Mesoscale Model 5 (MM5). The MM5 will generate
regional weather forecasts for the CAA and these will be available at both CAA and
CWB. The MM5 is a mesoscale prediction model with a history of both real-time
forecasting and research applications. In the Taiwan and east Asia region, it has been
used for real-time operations in Taiwan and Hong Kong. Research studies employing the MM5
have addressed key phenomena of the region such as typhoons, Mei-Yu fronts, mesoscale
convective systems, and flows over complex topography.
For the AOAWS, the MM5 will provide regularly-updated
forecasts on a range of temporal and spatial scales. It will allow forecasters to see the
large-scale changes over East Asia and the Western Pacific over 2-day periods, while
providing detailed information on conditions over the Taiwan FIR at half-hourly
resolution.
Taiwan FIR domain
(1,2,3) region.
The Model
The MM5 is the central component of
a modeling system which acquires observations and gridded analyses, processes the data for
model ingest, generates a numerical forecast and displays the forecast output. The MM5 is
a grid-point model capable of variable horizontal and vertical resolution and has
been used with horizontal grid sizes under 200 m. The MM5 employs a terrain-following
sigma vertical coordinate, with vertical resolution typically ranging from 20 to 40 m near
the surface and decreasing with height. A nested-grid model, the MM5 permits 10 levels of
sub-domains, as well as moving and overlapping grids. In nested configurations, domain
interaction can be either 2-way or 1-way. In the former, predicted fine grid information
is overwritten on the coarse grid, while boundary conditions are provided by the coarse
grid to the fine grid. In the latter, boundary conditions are provided by the coarse grid
to fine grid, but the fine grid product does not get overwritten on the coarse grid; the
grids run separately. With respect to computation, the MM5 is multi-tasked. Its
parallelism was originally based on a shared memory paradigm, although work has begun on
its adaptation to distributed-memory architectures. |
The MM5 is based upon the set of
primitive equations for a fully compressible, nonhydrostatic atmosphere in a rotating
frame of reference. The model's predicted variables include pressure perturbation,
horizontal and vertical wind, temperature, water vapor, cloud water and rainwater. The
model's prognostic equations are written in flux form.
The MM5 simulates grid-resolved and
sub-grid-scale precipitation. The former involves the explicit treatment of moist
processes, for which the model offers a variety of packages. These feature prognostic
equations for variables such as cloud water, cloud ice, rain water and snow. For
subgrid-scale moist processes, the MM5 currently offers six cumulus parameterizations.
These include the popular Grell, Kain-Fritsch, and Betts-Miller parameterizations.
|
|
Advanced Operational Aviation
Weather System (AOAWS) in Taiwan 
