Numeric Weather Prediction (NWP) has resort to the calculation capability of computers to estimate the future state of the atmosphere by using the so-called "prediction models". These are based on a set of equations that interpret the physical laws that describe the hydrodynamic behaviour of the atmosphere.
The models performance by means of supercomputers is only possible through highly complex computer codes that supply predictions for different variables of the atmosphere, such as temperature, pressure, wind, precipitation, among many others. In his daily work, the meteorologist interprets the data thus obtained, checking them against other meteorological information sources to prepare the weather forecast.
The quality of the numeric information is, however, conditioned by uncertainties inherent to the knowledge of the state of the atmosphere at a given moment and also by limitations imposed to the prediction models. The uncertainties inherent to the observation of the atmosphere and to the use of observations that are irregularly distributed over the globe are error factors introduced into the analysis; the approximations applied to equations and the space-time discretization conditioned by the capability of the computers used are examples of error factors introduced into the models.
Numeric prediction activity is a recent one when compared with other scientific activities in Meteorology. It began in mid 20th century, when the work of scientists as renowned as Rossby (1898-1957) was joined with the invention of the electronic computer. The first successful experiment in NWP is ascribed to Jule Charney (1917-1981) and took place in the USA in April 1950. Nowadays, NWP is an integral part of the operational activities in most of the meteorological services all over the world. Global forecasts, for a period of about one week, are prepared only at some world centres, but regional or local ones, for much shorter periods, are issued in many countries. In functional terms, the "timely" factor of the availability of meteorological information requires that highly specific activities should be maintained, which implies the availability of human resources with high technical and scientific skill.
The current NWP working group of IPMA began in 1993, resuming this activity in the national meteorological service in an integrated way within the European efforts towards the development of short- and medium-range numeric prediction, Portugal being a member of the European Centre for Medium-range Weather Forecasts.
One of the main products resulting from this area of activity at IPMA is the operational version of the AROME limited area model, which has a horizontal resolution of 2.5 km. Several products derived from this model have been developed, as well as from the ECMWF global model, and the results are verified/validated in a systematic way.