La question des précipitations acides est devenue une des principales préoccupations environnementales de ce siècle. Les dommages engendrés touchent l'ensemble des composantes atmosphérique, aquatique et terrestre de notre écosystème. Dans cet article, l'intérêt est tourné vers le transport des substances polluantes en faisant intervenir le vecteur des précipitations acides: le vent. Il existe une croyance populaire, largement répandue dans l'est du Canada, consistant à admettre que les vents d'est ou du nord sont associés à des précipitations faiblement chargées en éléments acides, contrairement aux vents du sud-ouest qui transporteraient les charges plus contaminées des industries situées dans le Midwest nord-américain. Une confirmation expérimentale de la réalité du phénomène est présentée ici.Une analyse factorielle des correspondances permet de mettre en évidence des relations entre la direction des vents et la composition chimique des précipitations. Des analyses de variance permettent ensuite de montrer la signification de l'effet de la direction des vents sur les concentrations de sulfates et de nitrates, en plus de mettre en évidence un effet saisonnier significatif pour ces deux variables. Les concentrations de nitrates et de sulfates associées aux vents de l'ouest sont respec- tivement de 0,33 mg/l-¹ et 1,73 mg/l-¹ comparativement à 0,24 mg/l-¹ et 1,48 mg/l-¹ pour les vents provenant de l'est. En ce qui concerne l'effet saisonnier, les concentrations moyennes de nitrates sont plus élevées durant les mois de janvier et de mars alors que les concentrations de sulfates sont plus elevées durant les mois d'été.
The problem of acidic precipitation has become an important environmental concern; related damages can affect atmospheric, terrestrial and aquatic components of our ecosystem. This paper focuses on wind-driven atmospheric transport of contaminants, notably niirates and sulfates. Conventional wisdom in eastern Canada holds that winds originating from the east bring precipitation less loaded with acidic components than winds originating from the southwest; the latter carry contarninants generated by industries located in the American Midwest.An experimental confirmation of this phenomenon is presented here. SPERBER (1987) showed that hourly series of piecipitation content and wind direction, measured at a reception site are adequate to represent the lagrangian history of precipitating systems (New York City region). Following this result, we suppose that the northesastern continental atmospheric system is homogeneous enough so that winds measured at our reception site (Québec City region) are representative of the whole system. Thus, our experiment is performed in eulerian coordinates.The data bank used in the statistic alana|yses contains 10 time series: the weekly concentrations of 9 compounds found in the precipitation (H, Ca, Cl, K, Mg, Na, NO3, NH4 and SO4) gathered at the local Montmorency site and a s:ries ofweekly prevailing winds measured at the nearby Québec City airport. The time series contain 312 observations covering a full six year period (December 1981 to December 1987). As the original data bank of corcentrations is episodic, i.e. an obsercvation is available for each day with a significant precipitation event, volumes and loadings are used to derive the average weekly values of concentrations. In contrast, hourly series of direction (projected in 36 directions) and velocity of the prevailing winds are used to build, via a vectorial addition, a weekly series of wind direitions projected on a 12 point wind rose where directions corespond to the nind origin and not its destination.Classic statistical methods are used to deal with this data bank. Principal component analysis studies relationships between series of concentrations in the precipitations, while correspondence analysis highrights the relationships betwlen tile series of precipitation content and the series ofwind direction. The final statistical method, analysis of variance, is used to test the signilïcance of relationships higtrlighted by the correspondence analysis.The principal component analysis shows that all variables were positively correlated with the first component which reflects the fact that a higtrly loaded precipitation event will show high concentrations for each ofthe nine variables. The second component discriminates two groups of variables: one includes NO3, NH4 and SO4, the other Ca, Mg and Cl. The acidity variable, H, is nearer to the acid ion group (NO3, NH4 and SO4) than to the other ion group (Ca, Mg and Cl).The correspondence analysis shows that high concentrations of acidic compounds (NO3, NH4 and SO4) are highly rerated to winds from the W and WSW directions; conversely lower concentrations of the same compounds are associated with winds from the E and ENE directions. The elements Ca, Mg and Cl, reputed to be of oceanic origin, show high functional relationships between high concentrations and E, ENE winds and also between low concentrations and W, WSW winds. These results support the popular belief initially presented that acid precipitation is largely associated with winds from the southwest.The analyses of variance show that precipitation concentrations (NO3 and SO4) are significantly different according to the wind directions. Nitrate and sulfate concentrations associated with winds originating from the west are respectively 0.33-mg.l-l and 1.73 mg.l-l compared, to 0.24 mg.l-l and 1.48 mg.l-l for winds originating from the east. The analyses ofvariance also indicate a significant seasonal effect where mean monthly concentrations in nitrates are hilhest for winter months and early spring whereas sulfate concentrations are highest for the summer months.
