Statistical simulation of variations in medical characteristic during heliogeophysical disturbances

Abstract

Heliobiological studies, although recently very active, have failed to provide a full understanding of the Sun–biosphere interaction mechanisms. This is because the forecasts of bio tropic effects of the active Sun, like geophysical forecasts, are mostly determined by empirical patterns and are of a probability character. To obtain probability predictions, an adequate statistical model should be prepared. This is what is treated in this work. The time series of data from medical studies have undergone the trend treatment. The samples were sorted by the state of heliogeophysical conditions, and statistical distributions and invariants have been calculated down to the forth order. The analysis has shown that during heliogeomagnetic disturbances the obtained distributions differ from the Gaussian law. Therefore, an attempt at a statistical description of medical characteristics was made on the basis of the Poisson model. For our samples, the characteristic function of the exponential kind was chosen under the assumption that time series are the superposition of some, determined or random, process. Through the Fourier transform, the characteristic function was transformed into a nonholomorphic expressiveasymmetrical function of the probability density. The statistical distributions, calculated for the studied samples over perturbation periods, were compared with the obtained model function of distribution. The probability of coincidence for a posteriori distributions by the criterion χ² is Р ~ 0.7–0.9. Our analysis has allowed a conclusion about the applicability of a model based on the Poisson random process for statistical description and appraisals of the probability of changes in the number of vascular bouts and strokes during heliogeophysical perturbations.