 Short term and real time prediction models
 Long term prediction models
Research Areas
 Earth’s Magnetic Field
 Ionosphere
 Ionospheric Sounding
 Ionospheric Forecast
 Passive Ionospheric Measurements
 Ionospheric Models
 MiddleUpper Atmosphere
 Paleomagnetism and Rock Magnetism
 Space Weather
 SunEarth Relations: Geomagnetic Phenomena
 SunEarth Relations: Ionospheric Phenomena
 Environmental Terrestrial Physics
 Hydrosphere  Geosphere  Atmosphere Interactions
Long term prediction models
The ionospheric variability of the E and F1 layers can be adequately described by the Chapman theory. This means that the behaviour of these layers can be predicted through very simple formulas. For example, the historical series of measurements of the critical frequencies foE and foF1 collected at the Rome (Italy) observatory, have been used to conduct a statistical regression analysis from which the following two formulas have been devised:
foE= 3.18[(1+8.83*10^{3}*R)cosχ ]^{0.222} (1) 
foF1 = 4.00[(1+1.36*10^{2 }R)cosχ ]^{0.196} (2) 
The formulas (1) and (2) are valid in the Mediterranean area. They are directly derived from the Chapman theory and constitute an example of longterm prediction models. These formulas show that the critical frequencies depend on the solar activity (through the sunspots number R) and on the solar zenith angle
χ
that varies according to the formulacos
χ
= senφ
senδ
+ cosφ
cosδ
cosω
being
φ
the geographic latitude of the observation site, δ
the solar declination (the height of the Sun over the celestial equator, ranging between around –22.27° in winter solstice and +22.27° in summer solstice) and ω
the solar hourly angle in the location and at the time under consideration.Since the F2 region does not follow the behaviour described by the Chapman theory, its variability cannot be predicted through simple formulas. Generally, the prediction models concerning the F region are developed by using parameters as R_{12} (a solar activity index defined as the twelvemonth smoothed mean of sunspot number R; although other kinds of smoothed means could have been used) and the monthly median values of the ionospheric characteristics (foF2, MUF(3000)F2, M(3000)F2, h′F2, foF1, h′F1, foE, h′E) collected by a network of ground ionospheric stations over a long time interval.
Usually, a significant statistical regression analysis that takes into account the huge data records produced is carried out to find the “law” that fits better the measurements.
This “law” is used to predict the ionospheric parameter under consideration for a future period. The ionospheric parameters obtained in this way, are then interpolated to calculate the values in those areas where they cannot be directly measured (for example, deserts and oceans). This approach is used to produce local and regional prediction maps of all the ionospheric characteristics. These maps, usually obtained for a given value of R_{12}, and for a given month and time, constitute the long term predictions, that are valid in the case of undisturbed ionospheric conditions, i.e. in the case of “quiet” ionosphere (figure 1).
Inner pages
