Forecast, Weather

a scientifically substantiated hypothesis regarding future changes in the weather, compiled on the basis of an analysis of large-scale atmospheric processes.

Weather forecasts are classified as short-range (covering from a few hours to one or two days), moderately long-range (three to ten days) and long-range (one month or more). Forecasts are made for territories (districts, regions, countries, and ocean areas), as well as for particular populated areas, airports, air routes, highways, and railroad lines.

There are specialized weather forecasts, which are intended for various sectors of the national economy, and forecasts for the general use of the population. Specialized weather forecasts include warnings of dangerous weather phenomena (cyclones, thunderstorms, fog, snowstorms, strong winds, dust storms, and frosts), which may cause difficulties or damage in various sectors of the national economy or threaten the safety of the population. Anticipated weather conditions are described in greater detail in short-term forecasts and warnings than in long-term forecasts. For example, weather forecasts for aviation include the expected weather conditions at flight altitude (type and quantity of clouds, wind direction and speed, and the presence of dangerous phenomena such as bumpiness, icing, or lightning storms), as well as conditions at the landing field (cloud height, visibility, wind direction and speed, and atmospheric temperature).

Moderately long-term weather forecasts, which describe the weather for the coming period in more general terms than short-range forecasts, predict the predominance of clear or cloudy weather, the possibility of precipitation, the range of daytime and nighttime temperatures, sharp changes in the weather, and the dominant direction and speed of the wind. Weather forecasts for a month include the sign and magnitude of the expected deviation from normal of the average temperature and precipitation, as well as an indication of the periods of the most marked changes in the weather (cold and hot periods, and transitions from dry to stormy weather). In addition to monthly forecasts, the USSR uses weather forecasts for “natural synoptic seasons,” with an average duration of about two months. These forecasts include the general temperature and precipitation pattern.

Weather forecasts are compiled accordingly to the methods of synoptic meteorology, which require the preparation of synoptic weather maps for various levels of the atmosphere, to an altitude of 30 km from the earth’s surface. Synoptic weather maps are based on data from the observations recorded at meteorological and aerological stations. Information obtained from meteorological satellites is also widely used. Analysis of data from satellites makes it possible to show large atmospheric formations on weather maps (air masses and the atmospheric fronts separating them, cyclones, and anticyclones). The motion and evolution of these atmospheric formations are related to major changes in the weather. Increasingly, weather forecasting is turning to numerical methods, which make it possible to use electronic computers to solve equations for the hydrothermodynamics of the atmosphere and for the temperature at various levels of the atmosphere, as well as to calculate the quantity of precipitation several days in advance.

All of these methods provide an idea of the general weather background, which forecasters analyze in further detail for local conditions. Long-term weather forecasts use various types of statistical links between the past and future development of atmospheric conditions and the state of the weather.

Within the limits of the period for which they are compiled, all weather forecasts decline in accuracy with time. The practicality of any method of weather forecasting is evaluated by comparing its success with that of inertial forecasts, which predict the maintenance of the principal weather pattern for the period of the forecast. On the average, more than 80 in every 100 short-range or moderately long-range forecasts prove to be correct. The main reasons for the greatest errors in forecasts are inaccuracies in calculating the baric field and the direction, speed, and evolution of cyclones and atmospheric fronts. These inaccuracies are due to the imperfection of methods of calculation and the absence of sufficient information from ocean areas, from sparsely populated regions, and particularly from the upper layers of the atmosphere.

The use of numerical methods for weather forecasting is limited by the availability of high-speed computers capable of storing hundreds of millions of words of information and handling it at a rate of several hundreds of millions of operations per second. The methods of long-term weather forecasting are still in the developmental phase. Long-term forecasts do not yet meet the necessary standards for quality. The compilation of sufficiently accurate long-term weather forecasts is one of the most difficult problems in modern science.