Comparative analysis of the quality of prediction for fluences of relativistic electrons of the outer radiation belt of the Earth at different phases of the solar activity cycleтезисы докладаЭлектронная публикацияТезисы
Дата последнего поиска статьи во внешних источниках: 9 марта 2017 г.
Аннотация:The forecast of radiation environment at geosynchronous orbit (GEO) near
the boundary of the Outer Earth’s Radiation Belt (OERB) is very important
due to the large number of satellites populating this region. Relativistic
electrons of the outer ERB are even called "killer electrons" since they can
damage the electronic components of spacecraft. The flux of the electrons
with the energy >2 MeV at GEO orbits is very unstable within time
intervals about days and even hours. It shows a strong temporal dependence
on epoch of solar activity (SA) relative to the onset of geomagnetic
disturbances (AE-min and AE-max models). Any generally accepted theory
that explains all the features of the experimentally measured behavior of
outer ERB is not created, so the statistical methods are usually less
dependent on the selected physical models, it is useful to have an effective
forecasting model based on statistical methods. Artificial neural networks
(ANN) model that would work as good or better than any other statistical
model. The intensity of the OERB depends on the values of the components
of interplanetary magnetic field, on solar wind, and on geomagnetic storms.
It is known that all of these values are strongly dependent on SA cycle
phase. We have already created an ANN prediction model for hourly
average values of the electrons of OERB at GEO with prediction horizon
from 1 to 12 h (R2
=0.98 and 0.86) using data without dividing it by the
level of SA. The hypothesis that separate networks trained for different
phases of SA would be more successful than the single network using the
whole data for the overall cycle has been checked in presented study.
This study was supported by the Russian Science Foundation, grant № 16-
17-00098.