About Opentopia
Opentopia Directory Encyclopedia Tools

Numerical weather prediction

Encyclopedia : N : NU : NUM : Numerical weather prediction

An example of 500 mb geopotential height prediction from a numerical weather prediction model
Enlarge
An example of 500 mb geopotential height prediction from a numerical weather prediction model

Numerical weather prediction is the science of predicting the weather using "models" of the atmosphere and computational techniques. Manipulating the huge datasets and performing the complex calculations necessary to do this on a resolution small enough to make it accurate requires some of the most powerful supercomputers in the world.

Contents

Physical Overview

The atmosphere is a fluid, the basic idea of numerical weather prediction is to sample the state of the fluid and use the equations of fluid dynamics to estimate future states at any time in the future.

Models

A model, in this context, is one or more computer programmes that output meteorological information for future times. Weather forecasting organisations often develop their own models; comparing the output of different models can indicate the uncertainty of the forecasts (see Ensembles below)- if several models give the same result, the forecast has a higher likelihood of occurrence, but if different models give different forecasts for the same place at the same time, the forecast uncertainty is greater. Models often have their own quirks and known biases, and human intervention and interpretation is often needed to remove these from forecasts prior to dissemination.

Models typically represent meteorological variables in terms of position and altitude as they evolve in time. Global models are generally based on a two-dimensional horizontal coordinate system with a fixed-difference vertical coordinate, dividing the Earth's atmosphere into a 3D grid with discrete cells. Regional models are generally finite-difference in all three dimensions, and typically use finer grids to explicitly resolve smaller meteorological phenomena.

A 96-hour forecast of 850 mb geopotential height and temperature from the Global Forecast System
Enlarge
A 96-hour forecast of 850 mb geopotential height and temperature from the Global Forecast System

The forecasts are computed using mathematical equations that describe the physics and dynamics of the atmosphere. The model calculates how the atmosphere will change in each cell with time, and how each cell affects its neighbours, building a forecast. Models are "initiated" using observational data, mostly collected by weather satellites. This data is then processed by the computer program. Data assimilation and objective analysis are methods of realigning scattered observations onto a ordered grid that the computer can understand and ensuring quality control. The data is then sent to the model to create a forecast. Commonly, the set of equations used are those known as the Primitive Equations. These equations are initialized from the analysis data and, based on the distribution of forces, time tendencies are determined. The equations are stepped forward in time through the complete forecast length. The time step is generally dependent upon the spatial resolution of the model. Time steps for climate models may be on the order of minutes to days, while the time step for regional model may range from a few seconds to several minutes.

Numerical weather prediction models may be global or regional. Some of the better known ones are:

Global models

Regional models

Ensembles

As proposed by Dr. Edward Lorenz in 1963, it is impossible to definitively predict the state of the atmosphere more than approximately 10 days in advance, owing to the chaotic nature of unevenly heated rotating fluid bodies. Furthermore, existing observation networks have limited spatial and temporal resolution, especially over large bodies of water such as the Pacific Ocean, which introduces uncertainty into the true initial state of the atmosphere which is required for a numerical weather prediction model. To combat this uncertainty, stochastic or "ensemble" forecasting is used, involving numerous forecasts created with different model systems, different physical parameterizations, or varying initial conditions. The ensemble forecast is usually evaluated in terms of the ensemble mean of a forecast variable, and the ensemble spread, which represents the degree of agreement between various forecasts in the ensemble system, known as ensemble members. When the ensemble members are in relatively high agreement, the ensemble spread is low and the forecast confidence high. Conversely, when the ensemble members diverge, there is a large ensemble spread and poor forecast confidence.

Useful references

See also

External links

 

From Wikipedia, the Free Encyclopedia. Original article here. Support Wikipedia by contributing or donating.
All text is available under the terms of the GNU Free Documentation License See Wikipedia Copyrights for details.

Search Titles
0123456789
ABCDEFGHIJ
KLMNOPQRST
UVWXYZ?

E-mail this article to:

Personal Message: