Glossary of data terms

data assimilation combines recent observations with a previous weather forecast to obtain our best estimate of current atmospheric conditions. This is evolved forward in time by the forecast model to produce the next forecast.

average weather and its variability over a period of time, ranging from months to millions of years. The World Meteorological Organization standard is a 30-year average.

a change in the climate's mean and variability for an extended period of decades, or more.

a mathematical representation of the climate system based on its physical, chemical and biological components, in the form of a computer programme. The computer climate models used at the Met Office Hadley Centre are detailed three-dimensional representations of major components of the climate system. They are run on the Met Office supercomputer.

coupled climate models combine representations of various components of the earth system such as atmosphere, ocean, sea ice and land surface. In coupled models each 'coupled' area of the earth system is influenced by other areas, as well as evolving independently.

a single best-guess forecast, which does not attempt to represent uncertainty. A probabilistic forecast can be produced through statistical analysis or ensemble methods.

the computer model is run a number of times from slightly different starting conditions. The complete set of forecasts is referred to as the ensemble, and individual forecasts within it as ensemble members.

a statistical analysis which can be employed to estimate the theoretical maximum or minimum value attainable by infinite 'trials' of any particular system. For example, maximum wave height at a particular location.

a forecast is an estimate of the future state of the atmosphere. It is created by estimating the current state of the atmosphere using observations, and then calculating how this state will evolve in time using a Numerical Weather Prediction computer model.

weather parameters mapped onto a regular geographical grid e.g. output from any of the Numerical Weather Prediction model or radar. Gridded data is useful when you need to analyse the weather across a wide area.

a hindcast is a numerical model integration of a historical period where no observations have been assimilated. This distinguishes a hindcast run from a reanalysis.

measurements or model data that were generated in the past. This data is stored for several years, so it can be accessed in the future e.g. satellite or radar images, rain gauge data, weather station observations, Numerical Weather Prediction model outputs.

a computer simulation of the processes in the Earth's atmosphere, land surface and oceans which affect the weather. Once current weather conditions are known, the changes in the weather are predicted by the model.

a measurement of actual weather conditions by instruments such as weather stations, buoys, radars or satellites.

the values obtained through observations.

collective term for any measurement of weather recorded at a single point (rather than across an area).

a range of techniques applied to data to make it suitable for a specific purpose e.g. averaging values, filtering, quality control.

a forecast which represents the estimated uncertainty of the prediction. This uncertainty can be estimated statistically through assessing past performance, or using an ensemble forecast.

precipitation recorded by a radar across an area, which measures the reflectivity of rain drops before they hit the ground.

a device for measuring the amount of precipitation (rain and snow) at a specific point over a period.

the process of applying modern data assimilation techniques to historical periods. A consistent scheme is applied to periods of decades (for climate monitoring and associated products) or longer (for climate change studies). Reanalyses are important since they provide weather and climate information across the region, not just where there are observations. They give a more complete and coherent picture of the weather than can be obtained from observation data alone.

climate models split the Earth's atmosphere and ocean into a finite number of grid boxes (similar to the pixels on a digital camera) - the higher the number of grid boxes, the higher (or finer) the spatial resolution. For example, a model with a horizontal resolution of 1 degree would have 360 (latitude) x 180 (longitude) = 64,800 grid boxes. The height of the atmosphere, and the depth of the ocean are split into distinct layers - so the number of these layers determines the vertical resolution of the model.

a forecast indicating when a weather parameter (or a set of weather parameters) exceeds a given value e.g. a wind speed of over 80 knots which may lead to a bridge closure.

the numerical model developed and used at the Met Office. Unlike most other organisations that use Numerical Weather Prediction models, we apply a seamless modelling approach for both weather and climate prediction.

a range of statistical techniques to determine how the weather affects a business outcome, positively or negatively e.g. impact on footfall or demand by sometimes subtle changes in temperature or cloud cover. WSA is used by organisations to improve operational planning.

a collection of instruments which measure surface weather conditions at a specific location. Instruments can include, amongst others, thermometers, anemometers, rain gauges, and radiometers. Automatic Weather Stations monitor and record observations automatically, whereas manual weather stations rely on meteorologists.