Thursday, October 13, 2011

Forecast Tools: The Utah Synoptic Diagnostic


Meteorologists have thousands of ways to slice and dice the output from numerical weather prediction models.  The four-panel Utah Synoptic Diagnostic (above), available at weather.utah.edu for several regions covering the United States and Europe, presents some of the most common levels and diagnostics used by meteorologists in a single plot for quick analysis and forecasting of the weather.

At lower right is the surface analysis/forecast, which includes the sea-level pressure (red contours every 3 mb), surface wind (i.e., for 10-m above the ground), and accumulated precipitation since the last available forecast time (liquid equivalent in hundredths of an inch).  The background is the model topography.

Wind barbs displayed here and on other charts extend from the model grid point or observation location toward the direction from which the wind is blowing.  A half-barb denotes 5 knots (2.5 m/s), a full barb 10 knots (5 m/s), and a flag 50 knots (25 knots), and you simply add them up to get the wind speed.  Some examples are given below.

Source: NOAA/HPC, Wikipedia Commons
At lower left is the 700-mb analysis/forecast.  Meteorologists like to work on pressure levels rather than constant height levels for a variety of reasons.  Pressure decreases with height.  It is near 1000-mb at sea level, but 700-mb at about 10,000 feet.  Thus, this panel provides information about the atmosphere at a level near the crest height of the Wasatch Mountains and includes temperature (red contours every 2ºC), wind, and relative humidity (color fill following scale at right, averaged for the 800–500 mb layer).  This is essentially a free-atmosphere forecast, so actual winds and temperatures in the Wasatch are going to vary compared to what is presented here depending on local effects.

At upper right is the 500-mb analysis/forecast.  This is a pressure level that is widely used by meteorologists and sits at about 18,000 feet above sea level.  Included are contours of geopotential height (green, every 60 meters), vertical velocity (every 3 cm/s, red=rising motion, yellow=sinking motion, averaged for the 700–500 mb layer), and absolute vorticity (x10e-5 per second, following scale at right).  For non meteorologists, this is one of the more obtuse panels.  You can think of the geopotential height contours as being the equivalent of height contours on a topographic map.  Areas with low values are troughs, high values ridges.  The flow is generally nearly parallel to the geopotential height contours with lower heights on the left, so you can infer the flow from the contours.  It is also inversely proportional to the distance between the contours (i.e., closer together=faster flow).

The vertical velocity just tells you how fast the air is rising or sinking as predicted by the model.  Keep in mind this does not include detailed topographic effects.  Finally absolute vorticity is a measure of circulation density, or the circulation per unit area, including the effects of the rotating Earth.  Locally high values are associated with strong cyclonic flow curvature or horizontal shear and locally low values are associated with strong anticyclonic flow curvature or horizontal shear.  Often, important features in the upper-levels are easier to see using absolute vorticity than geopotential height contours.

At upper left is the dynamic tropopause analysis/forecast.  The dynamic tropopause is an imaginary surface that separates the troposphere from the stratosphere.  The jet stream is usually located at or very near this level.  Included are the pressure of the dynamic tropopause (mb color filled following scale at right), wind, and wind speed (contoured every 12.5 m/s).  One can easily find the polar and/or subtropical jet streams using this diagnostic.

The time stamp in all of these plots is presently the year, month, day, and hour that the forecast is initialized (i.e., started) in UTC (Coordinated Universal Time).  For example, 2011101306 means that the forecast was initialized at 0600 UTC 13 October 2011.  One needs to subtract 6 hours for mountain daylight time or 7 hours for mountain standard time, so this corresponds to 0000 MDT 13 October 2011 or 1100 MST 12 October 2011.  The number following the F is the forecast hour.  54 means 54 hours after initialization or 1200 UTC 15 October 2011 (0600 MST 15 October 2011).  Many people have complained that I don't stamp the time the forecast is valid for on these panels.  Perhaps I'll get to this in a future life...

The topography in all of these plots is based on the terrain provided by the model grid, which isn't always the same as that of the native model.  For example, the four panel above is based on the NAM-212 grid.  212 just designates the region and resolution of the model grid we download from the National Weather Service.  Although the NAM is run at 12-km grid spacing, the 212 grid is only 40-km grid spacing, which means we don't need to use as much disk space to store it. Similarly, all the fields (sea-level pressure, precipitation, etc.) are also on this somewhat degraded grid.  If you want something closer to the native NAM, look at the NAM-218 grid, which has 12-km grid spacing that is very close to that of the actual NAM.

1 comment: