par De Meyer, Frans;Boute, Carmen
Référence International journal of biometeorology, 31, 4, page (301-322)
Publication Publié, 1987-12
Article révisé par les pairs
Résumé : The numerical data collected daily for the longest series of inorganic chemical tests, carried out in Florence (Piccardi and co-workers, 1951-1972) and in Brussels (Capel-Boute, 1956-1978), have been submitted to a statistical analysis for the purpose of searching an answer to the questions which led to start the collection of long-term series of data with the Piccardi chemical tests in different places. The question was to study the variability in the course of time of various effects observed on aqueous systems, even in the most rigorously standardized conditions, for a chemical precipitation reaction. Since significant long-term perturbations and an annual variation are present in all data sets, the observations cannot be conceived as purely random fluctuations. No common long-term pattern is observed and the measurements are not unambiguously correlated with climatological effects or the solar cycle. The statistical information content of the chemical tests is time-dependent, which implies non-stationarity of the observations. These results suggest the necessity of search for disturbing geophysical and cosmological factors to understand the mechanisms of the interaction. © 1987 Swets & Zeitlinger B.V.