First, Weather Measurements are Made: Twice each day, radiosondes
(weather balloons) are launched all over the world at approximately the same time.
They ascend up through the troposphere (the lowest 7-10 miles of the atmosphere, where
our weather occurs), and measure temperature, wind, air pressure, and wind speed and direction
(through radio tracking or GPS tracking). This information is continuously transmitted back
to the ground station, then sent to the National Centers for Environmental Prediction (NCEP)
in Washington, D.C. These measurements constitute a global “snapshot” of our weather at one
point in time, every twelve hours. Additional measurements are made at surface observing
stations (there is probably one near you), on ships at sea, from commercial aircraft, and
by weather radars of different types. While different forecasting centers might provide up
to four updated forecasts during the day, it is this twice-daily measurement system that leads
to only two significantly improved weather forecasts each day.
Next, These Measurements are Put Into Weather Forecast Models
All of these measurements are also transmitted to a different weather modeling centers – in
the U.S., England, Germany, Australia, and other countries. These modeling centers have
very fast supercomputers that have been programmed with equations that describe the
interactions between temperature, wind (both horizontal and vertical), and pressure;
as well as equations for water vapor transport, the formation of clouds and precipitation;
and for the absorption and reflection of solar radiation, and the absorption and emission of
infrared (heat) radiation. Together, all of these equations run as a computer program are
called a "numerical weather forecast model".
Finally, the Interpretation of Model Output: The Forecaster’s Role
For many years, weather fcorecast models were not good enough to compute what the weather
at the surface would be like…they could only compute the future weather at different levels
up through the atmosphere. What happens at the surface (where we live) is very complicated
because so much energy is exchanged there, and in a variety of complex ways. For instance,
the presence of vegetation greatly cools the surface due to evapo-transpiration--a jungle
is much cooler than a desert (but much more humid, too). Forecasters would examine the
output from the computer models and would make an educated guess of what the model forecast
meant in terms of surface weather. Also, the models were typically biased—under certain
conditions, they might habitually make storms too intense, not produce enough rain, etc., --
so these biases had to be factored into the forecaster’s thinking as well.
|