Weather Forecasting - II

1
Weather Forecasting - II
2
Review

• The forecasting of weather by high-speed
  computers is known as numerical weather
  prediction. Mathematical models that describe how
  atmospheric temperature, pressure, winds and
  moisture will change with time are programmed
  into the computer. The computer then draws
  surface and upper-air charts, and produces a
  variety of forecasts called progs.
• After a number of days, flaws in the computer
  models atmospheric chaos and small errors in
  the data greatly limit the accuracy of weather
  forecasts.
• Available to the forecaster are a number of tools
  that can be used when making a forecast,
  including surface and upper-air maps, computer
  progs, meteograms, soundings, and satellite
  information.
• Ensemble forecasting is a technique based on
  running several forecast models (or different
  versions of a single model), each beginning with
  slightly different weather information to reflect
  the errors in the measurements (similar to BCS
  computer averages).

3
Ensemble forecasting

• A technique to improve medium-range forecasts
• Run different models
• Run the same model with slightly different
  initial conditions
• Superimpose all results on the same graph
• Spaghetti plot
• The most common outcome is probably the most
  likely one

4
Forecasting Tools

• AWIPS Advanced Weather Interactive Processing
  System
• Meteogram collection of observations at a given
  station over some time
• Sounding vertical profile of the temperature,
  due point T and wind.

5
Forecasting Methods (1)

• Persistent forecast (short term forecast) when
  there is no reason for the weather conditions to
  change.
• Steady state forecast (trend method). If we know
  the speed and the direction of the weather
  system, we can extrapolate to its future location
  assuming same speed.
• Example a cold front, located 90 km due west
  from here, is moving eastward at 30 km/hr.
  Prediction a cold front will pass through in
  exactly 3 hours.
• Analogue forecast (pattern recognition) I have
  seen these atmospheric conditions before and
  based on what happened back then, I can predict
  tomorrows weather.
• Statistical forecast the forecast is made based
  on
• Numerical models forecast for weather elements
  (humidity, cloud cover, wind direction,
  temperature).
• Statistically weighted analogue forecasts based
  on those

6
The chance of rain is 60

• It will rain over 60 of the forecasted area. NO
• There is a 60 chance that it will rain somewhere
  within the forecasted area. NO
• It will rain 60 of the time. NO
• There is 60 chance that any random location
  will receive a measurable amount of rain. YES
• If you stay at one place there is a 60 chance
  that you will see rain. YES

7
(No Transcript)
8
Forecasting Methods (2)

• Climatological forecast Based on information for
  the typical weather conditions at a given
  location for a given season.
• Probability forecast based on historical data.
  What is the chance to snow in Gainesville on
  Christmas?

9
Forecasting Methods (3)

• Weather type. Uses general criteria such as the
  position of the subtropical highs, upper-level
  flow, prevailing storm track

• In general For a forecast to show skill, it must
  be better than a persistence forecast or a
  climatological forecast

10
The weather over the weekend
11
Type of Forecasts

• Very short-range forecasts (up to 6 h) based on
  satellite images and Doppler radar. Not likely to
  include numerical models.
• Short-range forecast (2-3 days) variety of
  techniques - satellite images, Doppler radar,
  wind and pressure maps of the surface and aloft
  also numerical models usually beyond 12 hrs.
• Medium range forecast (3-8 days) Mostly based on
  computer generated weather predictions.
• Long-range forecast (up to 16 days). It is not
  very accurate. Gives mostly the trends of the
  weather and the global expectations.
• Extended forecast one that goes beyond 3 days
• Outlook how the average weather conditions for
  a particular month or a season compare to normal
  ones.

12
90-day outlook for (a) precipitation and (b)
temperature
Above normal
Below normal
Below normal
Above normal
13
Predicting the weather from local signs.

• Signs of an approaching cold front hot
• and humid air, pressure quickly dropping.
• Signs of an approaching warm front cool weather,
  high clouds that are thickening and lowering,
  decreasing pressure, precipitation starting with
  snow (if cold enough) changing to sleet, freezing
  rain and rain.
• Signs of an approaching tropical
• storm decreasing pressure,
• increased cloud cover, the
• direction and the speed of the
• cloud motion (winds).

14
(No Transcript)
15
Weather folklore

• When halo rings the moon or sun,
• rain's approaching on the run
• When windows won't open,
• and the salt clogs the shaker,
• The weather will favor
• the umbrella maker!
• A cow with its tail to the West
• makes the weather best,
• A cow with its tail to the East
• makes the weather least
• Cats and dogs eat grass before a rain.

TRUE
TRUE
TRUE
FALSE
16
The Station Model Appendix B

• Temperature77 F
• Dewpoint 68 F
• Pressure (mbar) 998.
• Pressure tendency the last 3 h.
• The pressure here has fallen .3 mb
• the last 3 hours.
• Cloud cover it's mostly black showing that this
  station is mostly cloudy.
• Wind barb it points to where the wind is coming
  from. The two ticks represent 20 knot winds.
• Present weather a thunderstorm is occurring at
  the station.
• Pressure tendency the 1st hour the pressure was
  steady, then fell the last two hours.
• Past weather (one hour ago). In this case it was
  a light rain shower.

17
Example
18
Using surface maps to predict the weather.
Pressure Tendency
Current weather
Predicted front position
Upper level map
19
What actually happened

• Skip the last sections A forecast for six
  cities and A meteorologist makes a prediction