IBM supports regional weather
forecasting at the China Meteorology Administration
Based upon the success of the approach at the Olympics, several meteorological
agencies world-wide have determined that it fulfills a critical need in
their operational requirements. For example, the China Meteorological Administration
(CMA) began to utilize a version of
this system in May 1997, adapted for operation in Beijing area. Essentially,
the system used in operation by NOAA/NWS at the Olympics in 1996 and demonstrated
at the AMS conference in 1997 was replicated and adapted into operations
to meet CMA's requirements. In less than two weeks, personnel from IBM
and FSL were able to complete this task.
A capability to ingest available weather observations in the FSL LAPS
pre-processing step over a domain roughly 2000x3000 km (at 10 km resolution)
covering eastern China was established. The image below illustrates the
extent of this region. For this and any of the subsequent images,
you can view a higher-resolution version by simply clicking on it.
The output of LAPS was used to define initial conditions for the forecast
model, RAMS. Boundary conditions were established from a larger domain,
lower-resolution model operating on CMA's Cray. RAMS would run in an arbitrary
850x850 km (at 10 km resolution) window within the LAPS domain, with the
user having control over how many SP nodes would be used (at least 20 were
recommended), resolution in time for output for visualization (10 to 30
minutes), and length of the run (typically 24 hours). A new run would be
initiated every three hours.
Weather visualizations from CMA
You can view a number of visualizations of RAMS output generated at CMA.
Each example is accompanied with an in-line image. Click on the image to
see a full-size image of the same data. There are also MPEG animations
for these examples, and a VRML geometry corresponding and PanoramIX scene
for two of them.
Forecast for the Beijing area
on May 22, 1997
The image below shows a region over 830x830 km in size centered about
Beijing. You can see a terrain map, pseudo-colored by predicted precipitation
overlaid with coastline, national boundary and river maps. Predicted winds
are illustrated by directional arrows, colored by speed. Upper air data
are visualized as cloud water density isosurfaces and predicted reflectivity
surfaces corresponding to rain shafts. An animation
illustrates the formation of rain storms in this vicinity. The effect of
the sea breezes in the Gulf of Chihili can be seen.
You can also further examine this particular time step via a fly-through
animation or a simplified
VRML geometry or a PanoramIX scene.
Forecast for southern China on May 22, 1997
The image below shows a region over 830x830 km in size west of Hong
Kong (in the far right-hand side). It shows the Kwangtung and Kwangsi Chuang
regions north of Hainan. The image below shows a terrain map, pseudo-colored
by predicted precipitation overlaid with coastline, national boundary and
river maps. Predicted winds are illustrated by directional arrows, colored
by speed. Upper air data are visualized as cloud water density isosurfaces.
An animation illustrates the formation of rain
storms in this vicinity. A convergence zone along the coast of the Gulf
of Tonkin and South China Sea can be seen.
Forecasts for Hong Kong on May
22-23, 1997
The image below shows a region over 830x830 km in size centered east
of Canton, north of Hong Kong. It shows the Kwangtung region. The image
below shows a terrain map, pseudo-colored by predicted precipitation overlaid
with coastline, national boundary and river maps. Predicted winds are illustrated
by streamlines with directional arrows, colored by speed. Upper air data
are visualized as cloud water density isosurfaces and predicted reflectivity
surfaces. An animation illustrates the formation
of a cluster of thunderstorms in the vicinity. (The animation
can also be viewed at higher resolution, but the file is three times bigger.)
A convergence zone along the coast of the South China Sea can also be seen.
This same region was examined in a subsequent simulation for the period
12 hours later. You can compare the results via animation.
You can also further examine this particular time step via a fly-through
animation or a simplified
VRML geometry or a PanoramIX scene.
The LAPS/RAMS/DX weather forecasting system at CMA
is a joint project between IBM, the Forecast
Systems Laboratory of the National Oceanic
and Atmospheric Administration and CMA.
lloydt@watson.ibm.com
