APPLICATION OF THE METHODS OF STATISTICAL ANALYSIS OF PSEUDO-RANDOM SIGNALS TO THE INTERPRETATION OF FIELD TRIALS ON THE ATMOSPHERIC TRANSPORT OF ACTIVE PRODUCTS AND TO ESTIMATE OF CONFIDENCE LIMITS
An attempt is made at identifying the most important factors which introduce difficulties in the analysis of results from tests on pollutant dispersal: the unsteadiness of the phenomenon, the effect of external uncontrollable parameters, and the inherent complexity of the problem itself. The basic models for prediction of dispersion of passive contaminants are discussed, and in particular a Lagrangian approach which seems to provide accurate results. For the analysis of results many problems arise. First the need of computing for the results the statistical quantities which describe them: the mean, the variance and higher order moments are important. It is shown that there is no easy solution if the duration and/or the number of independent events to be analyzed are too limited. The probability density function provides the most useful information, but is not easy to measure. A family of functions is recalled which predict reasonably well the trend of the pdf. Then the role of intermittency is shown in some detail. Its importance cannot be underestimated and its relationship to pdf and the effects on measurements are shown to be rather complex. Finally, an example is made to show the effects of the variance of external factors.
Bibliographic Reference: EUR 9536 EN (1984) MF, 135 P., BFR 240, BLOW-UP COPY BFR 675, EUROFFICE, LUXEMBOURG, POB 1003
Availability: Can be ordered online
Record Number: 1989123031300 / Last updated on: 1987-01-01
Available languages: en