Numerical Weather Prediction - PowerPoint PPT Presentation

1 / 15
About This Presentation
Title:

Numerical Weather Prediction

Description:

Computer models, numerical models attempt to predict the weather using 7 ... Planetary, synoptic, meso, and micro (or storm) scales ... – PowerPoint PPT presentation

Number of Views:93
Avg rating:3.0/5.0
Slides: 16
Provided by: adamp7
Category:

less

Transcript and Presenter's Notes

Title: Numerical Weather Prediction


1
Numerical Weather Prediction
  • EAS 211
  • Spring 2005
  • 04/22/05

2
NWP
  • Computer models, numerical models attempt to
    predict the weather using 7 fundamental equations
    that govern how the basic parameters change in
    the atmosphere with time.
  • The study of how we can physically and
    mathematically represent processes in the
    atmosphere is known as dynamics.

3
Computer models cannot perfectly predict the
weather, obviously, but what makes them
inaccurate?
  • Errors.
  • Initial conditions
  • Intrinsic to the model
  • Errors perpetuate with time and the model becomes
    less and less accurate.
  • Parameterizations.
  • Methods for dealing with physical processes that
    are too small to include in the model.

4
Why do we need parameterizations? How can a
process be too small?
  • Planetary, synoptic, meso, and micro (or storm)
    scales
  • The model equations used in the model are only
    used for one scale, so we eliminate all other
    scales.
  • But often, the scales we eliminate are large
    contributors to the scale we are examining.
  • Smaller scale motions not captured by the surface
    or upper air observing networks are still
    important to the larger scale.
  • So we parameterize the effects of the smaller
    scale on the larger scale.

5
What is parameterized?
  • Convection
  • MicrophysicsMoisture
  • Turbulence
  • Soil Moisture
  • Radiation

6
WE NEED BETTER OBS.
  • Improvements in the density of our surface and
    upper air observing networks to better capture
    the small scale motions would lead to improvement
    in the model results.

7
Current NWP models
  • Run daily at the National Center for
    Environmental Prediction (NCEP)
  • NGM
  • ETA
  • AVN
  • ECMWF
  • RUC
  • MM5
  • WRF
  • GFS
  • NAM

8
NCEP Models
  • NGMNested Grid Model
  • 80 km
  • Grid point model, meaning obs are converted to
    values at various points that are evenly spaced
    so the computer program can plug them into
    equations.
  • Frozen model.
  • Being phased out.

9
NCEP models
  • Eta
  • Doesnt stand for anything.
  • Grid point model using either a 22 or 12 km grid.
  • Accounts for changes in terrain as a series of
    steps each of which corresponds to a level of the
    models vertical coordinate, eta. (Eta is
    analogous to pressure).
  • NAM North American Mesoscale Model

10
NCEP models
  • AVN, MRF, GSM
  • Global-spectral models
  • Converts data into a large number of mathematical
    waves which the model uses for its calculations
    and then returns the waves in a manner that will
    produce a forecast map.
  • These models go out farther in time than the
    regional models.
  • AVN-160 hours

11
NCEP Models
  • ECMWF (European Center for Medium-Range Weather
    Forecasts) and Canadian
  • Spectral models used for medium range forecasts
    (7-10 days)
  • Considered by many to be the most advanced model
    in the world.

12
NCEP Models
  • MM5Mesoscale Meteorological Model Version 5.
    Penn. State and NCAR.
  • WRFWeather Research and Forecasting Model (NCAR)
  • MM5 is designed to run on the mesoscale.
  • MM5 and WRF are primarily used for research
    purposed, but there are some available MM5 and
    WRF models running throughout the US.

13
Ensemble Predictions
  • Method uses a collection of two or more forecasts
    for the same time. These forecasts either use
    different initial conditions or are based on
    different forecasting procedures.
  • The various forecasts represent possibilities of
    how the weather might change in the future.
  • An estimate of the probabilities of various
    events as well as an average (consensus)
    forecast.
  • This means if several forecasts all give the same
    result, this result is more likely to occur, we
    have more confidence in our forecast.

14
Advice on NWP
  • Models should not be used without first
    consulting the observed data and applying
    knowledge of atmospheric processes to determine a
    forecast.
  • To use only model data is to succumb to what is
    called meteorological cancer
  • Try to use your time wisely when creating a
    forecast rather than trying to determine which
    model to use. You will find that no model will
    predict as well as an experienced, knowledgeable
    human forecaster.

15
Information on Models
  • NCEP
  • http//www.ncep.noaa.gov
Write a Comment
User Comments (0)
About PowerShow.com