About NCAR-RAP Upper-Air data page:

Winds/Temperatures on pressure levels Every day at 0000 and 1200 UTC, weather stations around the world launch helium-filled balloons with miniature weather instruments inside a package the size of a milk carton. These are called radiosondes or rawinsondes and are launched at the sites shown on the map found on the Upper-Air data page (as well as from other sites around the world not displayed on the map). These balloons carry sensors for measuring temperature, pressure and humidity (moisture). The measured data are relayed via radio signals and received by the ground station where winds are computed based on these signals or by newer GPS technologies. From these data meteorologists obtain "soundings" or profiles of temperature, moisture, and winds with altitude.     The graphics shown on the top of the Upper-Air data page plot these data on a map of North America with temperature (°C) to the upper left, dewpoint depression (°C) below it, altitude of the pressure level (decameters) to the upper right, and a wind barb representing the wind speed and direction. These data are fed into supercomputers at the National Centers for Environmental Prediction which analyze the data into a regular grid of data before predicting weather into the future. The contours of temperature and heights are provided by one of these numerical models (called the Rapid Update Cycle) which is updated every 3 hours using not only rawinsondes (which are only launched every 12 hours) but also automated aircraft reports (ACARS) and wind profiler data (see below). This web site is very different than others in that the contours of temperature, height, and wind speed update every 3 hours while the plotted station data reflect the 12 hourly rawinsonde data. However, this means that noticeable discrepancies may exist between the contour data and station data - use with this in mind.

Skew-T/Log-P Diagrams The next section of the Upper-Air page provides the rawinsonde data on a diagram meteorologists call "Skew-T/Log-P" plots. This name comes from plotting Pressure in a logarithmic fashion on the Y-axis and Temperature as skewed lines on the X-axis (running from lower left to upper right). The diagram is truly a tool for meteorologists and can reveal many key aspects of the atmosphere above a single point on earth - it is most often used to judge the amount of instability or thunderstorm potential. For simplicity, many of the severe weather indices are pre-computed and placed along the top of the graphic (Lifted Index, Sweat, CAPE, and many others).     At some point I'd like to write an online guide for how to interpret a Skew-T plot but that's down the road. For some help in the meantime, I often recommend the Univ of Illinois' Weather World 2010 website. My plots are somewhat different than others and for those who want to know what the colored circles are on the hodograph (the inset diagram with circles in the upper-left), search no further. The colored gray and green rings and blue dot represent the work of Rasmussen and Blanchard (see References at end of this page) and are an attempt to predict the motion and type of supercell thunderstorms based on a climatology study of supercells. I suggest reading the full journal reference for all details but the short answer follows: The blue dot is the estimated supercell storm motion computed by taking 60% of the magnitude and 8 m/s to the right of the boundary-layer to 4 km shear vector. This blue circle represents a 4 m/s error estimate and is also indicated by the heading "CELL" along the top and is only valid for developing to mature supercell thunderstorms not necessarily all general thunderstorms. When the blue dot lies within the green ring, then no supercells are anticipated. When the blue dot lies on the green ring, then HP, or "High-Precipitation" supercells are expected. When the blue dot lies on the gray ring, then "Classic" supercells are expected and when the blue dot lies outside the gray ring, then LP, or "Low-Precipitation" supercells are expected.

Wind Profiler plots A series of vertically-pointing radars exists in the Central Plains states as a "demonstration network". Unlike typical weather radars (like those displayed on the radar page) which measure energy reflected off precipitation (or birds or insects), wind profiling radars have a much longer wavelength (about 1-5 m instead of 5 or 10 cm) which measures energy after reflection due to atmospheric density changes. Upon analyzing these radar data, the speed and direction of the wind can be computed from near the ground to a height of approximately 11 km. On the Upper-Air page, these data are plotted horizontally at different altitudes as well as singular plots for each station as time-series vs. altitude. Wind barb notation is used for both types of plots.

References Rasmussen, E. N., and D. O. Blanchard, 1998: A baseline climatology of sounding-derived supercell and tornado forecast parameters. Wea. Forecasting, 13, 1148-1164.