IBM demonstrates regional weather
forecasting system at the Supercomputing 1997 conference
The capabilities developed and utilized for the Olympics, AMS and CMA have
since been employed in other operational forecasting settings. For example,
at the Supercomputing '97 (November
17-21, 1997 in San Jose, CA), this capability was replicated in the IBM
booth as part of the conference's technical exhibition. The system was
adapted to the San Jose area as shown below. For this and any of
the subsequent images, you can view a higher-resolution version by simply
clicking on it. You can also interact with this map via a scene in
PanoramIX or simplified VRML.
New 24-hour mesoscale forecasts were produced at 5 km resolution in
a region roughly 425x425 km in extent during the conference on a three-hour
update cycle. The computation took place on eight 160 MHz P2SC thin nodes
of a much larger SP that was in the booth. One 135 MHz P2SC wide
node was used for I/O. Two workstations (an IBM RS/6000 and an IBM
Intellistation) and two laptops (IBM RS/6000 860 and IBM Thinkpad 760E)
were available in the booth to interact with the model and analyze results
in a fashion similar to what was used at the Olympics. Additional workstations
were configured for use in the NOAA/FSL booth at the conference.
In this case, raw observations were received regularly over the internet
at FSL in Boulder, CO, where the LAPS pre-processor step was executed.
The LAPS results were then networked to the SP via the internet from FSL
for ingest into RAMS. A 24-hour forecast over the 425x425km domain at 5x5km
took about 2 hours on the eight SP nodes.
Output from RAMS every 10 minutes of forecast time were provided for
browsing visualization. Animations were produced routinely as the primary
mechanism to evaluate the model output through the procedures proven at
the Olympics. However, this Data Explorer-based application was significantly
enhanced over the earlier implementations to include integrated flyover
(key-frame animation), and creation of new products for web distribution
besides animations and image "snapshots". Image-based rendering of
three-dimensional scenes via PanoramIX
and geometric descriptions of three-dimensional scenes via VRML
incorporating simplified
geometry were available. The system also permitted simple tracking
of the simulation, so that the interactive tools could be utilized while
the model was running. If there were problems in the model run, then the
execution could be terminated and the model restarted with new input observations.
This operational numerical forecasting coupled with routine, interactive
three-dimensional visualization was unprecedented at this conference.
In addition, interactive visualization applications for the analysis
of post-processed model results and as well as the output
from LAPS were introduced operationally. All of these
new capabilities are illustrated below.
The following image is from one of the animations
produced during the conference. The 145-frame animation shows predicted
rainfall along the coast and in the mountains.
It shows a terrain map, pseudo-colored by precipitation overlaid with
coastline, county and river maps. Predicted winds are illustrated by arrows,
colored by speed. Topographic effects are seen on both the surface winds
and on clouds, which are visualized as a white, translucent isosurface
of cloud water density. This particular time step can also be examined
via a flyover animation.
This model run and subsequent ones that day predicted heavy rainfall
(about 20 mm) along the Pacific coast and in Santa Cruz with little or
no rainfall in San Jose late Tuesday evening, November 18 and early Wednesday
morning November 19, which were clearly illustrated in the visualization.
The previous example was derived from a run initiated for 8:00 AM PST on
November 18. The following examples are from a later run initiated
7 hours later, which show in more detail the distribution of predicted
rainfall and a front that passed through the region, and are shown in a
145-frame MPEG animation. The results from
these simulations and the exhibition of the system at the conference were
reported by the local media, particularly the San
Jose Mercury-News and the local ABC network affiliate, KNTV.
The Mercury News featured an article in the business section, "Call
this supercomputer Deep Thunder" on November 19 with a short follow-up
article the next day.
Compared to the previous image (and animation), reflectivity surfaces
at 30 dBz are also shown, corresponding to rain shafts. This particular
time step can also be examined via a flyover
animation, simplified VRML geometry and
a PanoramIX scene, as well as in the closer
view below. Here the reflectivity data and annotation are omitted,
and the wind are now visualized as streamlines draped over the topography
and pseudo-colored by speed. The predicted rainfall along the coast
and in the Silicon Valley area is shown in more detail.
This rainfall prediction was in contrast to the official forecast which
indicated a fairly uniform pattern of rain over the entire valley, including
San Jose. The actual rainfall amounts were confirmed by checking
the rain gauge data at specific sites along the coast the next day (1/4
to 1/2 inch). San Jose only had some drizzles in a few areas.
In addition, the visualization accurately showed the movement of the front
that led to the rain, which can be seen in the animation.
Another example is shown below from a forecast initiated at 9:00 AM
on Wednesday, November 19, and in a 78-frame MPEG animation.
It shows a terrain map, pseudo-colored by precipitation overlaid with
coastline, county and river maps. Predicted winds are illustrated as streamlines
draped over the topography and pseudo-colored by speed. Topographic
effects are seen on both the surface winds and on clouds, which are visualized
as a white, translucent isosurface of cloud water density. Remnants
of the front discussed earlier with previous forecasts can be seen moving
east in the animation. This particular
time step can also be examined via a flyover animation,
simplified VRML geometry and a PanoramIX
scene.
Another example is shown below from a forecast initiated at 1:00 PM
on Thursday, November 20, and in an 115-frame MPEG animation
(or at higher resolution). It was one
of the last model runs completed at the conference.
It shows a terrain map, pseudo-colored by filled, color contours of
temperature, overlaid with coastline, county and river maps. Predicted
winds are illustrated as flags which point in the direction of the wind
and are pseudo-colored by speed. This particular time step can also
be examined via a flyover animation, simplified
VRML geometry and a PanoramIX
scene.
This visualizations shown above are from a browser application.
It has the ability to view and interact with the data in a latitude-longitude
(from the model's stereographic grid)-altitude (from the model's height
calculations) coordinate system. The view is annotated with base
maps. The focus of this application is qualitative assessment (browsing)
of the model output. The features of this application are described
below, including the capabilities to create the visualization already discussed.
You have the ability to create both time-based and key-frame (flyover)
animation. There is a Primary Controls panel, which gives you the
ability to select specific surface and upper air variables to visualize
with pre-defined methods. It also gives you control over the creation
of key-frame animation. The Input and Outputs Controls panel allows
you to select the particular RAMS run of interest, including one that may
still be executing. It also allows you to create products for web
pages -- images, VRML and PanoramIX. Additional information about
these options is available via the Help button in each control panel.
This is a screen dump of the application.
There are two other windows of interest. The primary one is the
Image window, within which you may view and directly interact with the
model output. There are several options available, including changing
viewing modes (Options pull-down, View-Control) such as rotation or pan/zoom,
and saving/printing images/animations (File pull-down). The coordinates
can be optionally annotated with an axes box be turning on AutoAxes from
the Options pull-down.
The other window is the Sequence Control, a graphical widget with the
appearance of a VCR. It gives you the ability to specify a time step
or frame within the model run you are currently examining. You can
move forward or backward in sequence, single step, pause, loop continuously
or loop back and forth. The default settings for the Sequencer are
to have the program operate on each individual time step from the model
as a frame in the animation. If you click on the box in the upper
right of the Sequencer, you will notice that you have the ability to change
the increment setting. By reducing the size of the increment, for
example, you will lengthen the animation by skipping time steps from the
simulation. You can also manually control which time steps are available.
The Primary Controls panel in the browse application allows you to select
specific visualization techniques and data. The visualization coordinate
system (latitude, longitude, altitude) is annotated with base coastlines,
political boundaries, river and topographic maps. The vector/line
maps are draped over a topographic surface and registered in this three-dimensional
coordinate system with the model output.
You may choose one of several surface scalar parameters to show on the
map, which include temperature, dew point, relative humidity, precipitation,
heat index and pressure. The data are shown as a continuous pseudo-color
field. Most variables (e.g., temperature) are shown with a continuous
colormap from cyan to red. Variables related to moisture (e.g., humidity
and precipitation) are illustrated with a continuous colormap from brown
(implying dry) through green (for modest moisture) to deep blue (implying
wet). The data are shown as a shaded colored surface that is deformed
vertically by local topography, and overlaid with the vector maps.
Alternatively, you can select topography, then the surface will be colored
like a topographic map, where color is a redundant way of indicating altitude
with the height of the surface. You also have the ability to adjust
the scaling range for the variable that you have selected via a pair of
stepper widgets.
This panel also allows you to optionally visualize surface winds through
one of several techniques, which will be pseudo-colored by speed.
The maximum value for scaling may also be set. The continuous colormap
ranges from a deep violet for very calm winds to white for the maximum
speed specified. These techniques include vector arrows of fixed
size draped over the topographic surface, which point to indicate the direction
of the wind, streamlines with directional arrows, which show the continuous
flow of wind at the particular time step, or flags distributed over the
domain of the model run. The flags are rigid, which point to indicate
the direction of the wind.
The model contains several parameters related to clouds. You may
select an isosurface of total cloud water (sum of ice and liquid) to be
generated, which can correspond to cloud boundaries. The isosurface
is colored white with variation in darkness and opacity according to the
data. Keep in mind that these surfaces are three-dimensional analogues
of contour threshold levels and thus, do not necessarily represent a true
"cloud surface". If the isosurface threshold value is low, then the
surface can often be considered an outer cloud "boundary". You may
indicate one or more threshold values in kg/kg (water/air) via a widget.
Alternatively, the total cloud water density can be visualized via direct
volume rendering, in which the larger predicted values are mapped to increased
opacity and darker shades of white in a continuous representation.
Forecasted reflectivities in dBz can also be selected using the same options
as available for cloud water density, except cyan is used for coloring
instead of white. Isosurface representations of reflectivity inside
a cloud isosurface can correspond to internal rain shafts.
The panel also allows you to optionally visualize surface low and high
pressure regions. This is accomplished by marking the low and high
areas with an L and H, respectively. In addition, you have the ability
to mark several major cities in the region, and control the display of
basic annotation.
You have the ability to create a flyover key-frame animation along a
path that you can specifiy interactively with the mouse. To change
the path, go to Cursors Mode (Options pull-down in the Image window, View-Control).
You will see a set of little probe points in the volume. With the
left mouse button, you can drag individual points around, which will show
coordinates in the upper left. You can indicate a new location by
double clicking with the left mouse button at the desired place.
You can also delete one of the extant locations by pointing at it with
the cursor and double-clicking the left mouse button. The locations
that have been specified will serve as control points for the animation.
When you want to view the animation, select the Flyover button and hit
Execute (or Ctrl-O). Another image window will pop up and be the
view of the model output of the current time step along the flight path.
The main window will show the model output with the flight path and a little
cursor corresponding to the location being viewed in the animation window,
which is shown in the screen dump.
After each RAMS execution, all of the results
are collected and reorganized into a form that can be used by standard
meteorological analysis tools as provided by FSL, for example. This
post-processed data were made available for interactive three-dimensional
visualization and analysis via a new Data Explorer-based viewer application.
This includes all of computed variables from the model, but at hourly resolution
unlike the browser application that worked with a subset of variables but
at six times the temporal resolution. Here is a some sample image
and animation created with this application
for a forecast initiated on Wednesday, November 19 at 2100 UTC (1 PM PST).
A surface variable (lifted index) has been selected for display as
pseudo-color, which is overlaid on a topographic map. Rivers (blue)
and coastlines (black) are draped on the surface. An upper air variable
(reflectivity) has been selected for display via surface extraction.
The surface at 15 dBz is requested in translucent brown. It is not
visible in this image but can be seen in the animation
from this model run. Another field (specific humidity) has been selected
to show as a vertical slice, which is pseudo-color contoured. Any
of the three-dimensional fields available from the model can be visualized
with either of these methods. The upper air wind data can be seen
along four vertical profiles, which are specified interactively.
The direction of the model wind field along these "virtual soundings" are
shown via vector arrows pseudo-colored by horizontal wind speed.
The length of the arrows also corresponds to the horizontal speed.
The profile is realized as a pseudo-colored tube, which is contoured by
the variable selected for isosurface realization (i.e., reflectivity).
No tubes are seen in this image because the model does not predict any
reflectivity. They can be seen in the animation.
Here is a another sample image and animation
created with the RAMS viewer for a forecast initiated on Thursday, November
20 at 2100 UTC (1 PM PST).
A surface variable (lifted index) has been selected for display as
pseudo-color, which is overlaid on a topographic map. Rivers (blue)
and coastlines (black) are draped on the surface. An upper air variable
(relative humidity) has been selected for display via surface extraction.
The surface at 65% is requested in translucent tan. Another field
(temperature) has been selected to show as a vertical slice, which is pseudo-color
contoured. Another field (vertical wind speed) has been selected
to show via a horizontal, isobaric slice at 300 mb. Any of the three-dimensional
fields available from the model can be visualized with any of these methods.
The upper air wind data are shown via streamribbons. The seed points
are along vertical profiles, which are specified interactively. The
direction of the model wind field along this "virtual sounding" is shown
via vector arrows. Both the arrows and ribbons are pseudo-colored
by horizontal wind speed. The length of the arrows also corresponds
to the horizontal speed. The profile is realized as a pseudo-colored
tube, which is contoured by the variable selected for isosurface realization
(i.e., relative humidity).
With the viewer visualization and analysis application, you have the
ability to view and interact with the data in a latitude-longitude (from
the model's stereographic grid)-pressure coordinate system. The coordinates
are annotated with an axes box and base maps. The focus of this application
is the analysis (visualization and interrogation) of the model output.
There is a control panel, which gives you the ability to select any
of the surface and upper air variables to visualize with pre-defined methods.
Another control panel, which is not visible, allows you to select the RAMS
run of interest. Additional information about these options is available
via the Help buttons in the control panels. This is a screen
dump of the application.
There are three other windows of interest. The primary one is
the Image window, within which you may view and directly interact with
the model output. There are several options available, including
changing viewing modes (Options pull-down, View-Control) and saving/printing
images/animations (File pull-down).
A third window is the Sequence Control, which is not visible.
It is a graphical widget with the appearance of a VCR. It gives you
the ability to specify a time step or frame within the model run you are
currently examining. You can move forward or backward in sequence,
single step, pause, loop continuously or loop back and forth. If
you hit the button in the upper right, you can manual control the time
steps available as well as the increment between steps to be examined.
The control panel in the viewer application allows you to select specific
visualization techniques and data. The visualization coordinate system
(latitude, longitude, pressure) is annotated with an axes box and base
coastlines, river and topographic maps. The vector/line maps are
draped over a topographic surface and displaced below the lower vertical
boundary. The surface may be pseudo-colored by any one of over two
dozen surface, scalar, two-dimensional variables produced by the model.
This may also include topographic height.
The model contains several upper air, three-dimensional fields.
For each of 14 upper air scalar fields, you may choose to realize the data
as an isosurface, vertical slice and horizontal (isobaric) slice.
For the isosurface, you may choose a specific threshold value. The
isosurface is colored according to a segmented colormap. When you
change the variable of interest, the default value for the isosurface is
the mean. You may indicate the desired value via a stepper widget.
For the vertical slice, you may select a grid position, and whether
the slice is meridional or zonal. The slice is color-filled, pseudo-colored
with a segmented colormap and line contoured. For the horizontal
slice, you may choose what pressure level is to be shown. The slice
is pseudo-colored with a continuous colormap.
You may probe the volume for specific values at selected locations within
the model output. If the probe button is pushed, the variable that
was selected for isosurface representation will be interrogated.
The results will be displayed in a dialog box that will pop up on the screen.
To change the location, go to Cursors Mode (Options pull-down in the Image
window, View-Control) and select probe_volume. You will see a little
probe point in the volume. With the left mouse button, you can drag
the point around, which will show coordinates in the upper left.
In execute-on-change mode, when you release the mouse button, the results
will be shown. An example of this feature is shown in the image.
The upper air three-dimensional wind velocity is visualized via interactive
marking of geographic locations of interest. You may define one or
more geographic locations for "virtual soundings" within the model
atmosphere. This is also done in Cursors Mode. You should select
profilers. You can either grab and move one of the locations already
set with the mouse or indicate a new location by double clicking with the
left mouse button at the desired place. You can also delete one of
the extant locations by pointing at it with the cursor and double-clicking
the left mouse button. At the locations that have been specified,
a vertical profile is extruded through the entire model atmosphere, which
is realized as a tube. The sounding location is used to derive information
about wind velocity. The wind velocity along the profile is shown
by a set of vector arrows that point in the direction of the wind.
Horizontal speed at these points are indicated by the color and length
of the arrows. Optionally, the locations on each "virtual wind profiler"
can be used for seed particles for particle advection, which is realized
as streamlines. These lines, which are also pseudo-colored by horizontal
speed, indicate the instantaneous direction of the modelled wind from these
locations. If a variable has been selected for realization as an
isosurface, then the values along each profile of that variable are also
shown as pseudo-colored, filled contour bands using the same segmented
colormap as is employed for the isosurface. These profiles are also shown
via a conventional pressure-profile plot as shown in a separate image window.
As part of the effort in developing forecasts, the results
from the LAPS pre-processing, assimilation step are analyzed as well as
RAMS output. Tools similar to those for RAMS were also developed for the
study of LAPS results. A discussion of this capability and example
results are available for you to examine.
lloydt@watson.ibm.com
