On the meteorological interpretation of the chemical composition of monthly precipitation samples

Abstract

A methodology is presented for studying the influence of synoptic weather patterns on the chemical composition of monthly precipitation samples. The data selected for illustration are the sulfur and chloride monthly depositions at two stations in southern Sweden, Plonninge (coastal) and Flahult (inland), the observations extending from 1951 to 1969. The monthly values have been grouped in various ways into subsets of about 24 members, and the median has been used as an index of central tendency in each subset. An initial grouping according to monthly precipitation amounts yielded curvilinear relations between deposition and precipitation. Another grouping by 2-year time periods revealed a marked increase in sulfur deposition at both stations in the early 1960's: chlorides, on the other hand, were highest in the middle 1950's, and the secular variations at Plonninge were not in phase with those at Flahult. The variations in deposition could not be explained by variations in precipitation amounts. In order to study the influence of meteorological factors, a weather classification was developed based on the direction of the surface geostrophic wind over southern Sweden during periods of precipitation. For sufficiently wide sectors and for medians of large samples of data, this proved to be a useful predictor of deposition, being an index of upwind trajectory. All precipitation events during the 18 years of record were thus classified, and linear relations were established between median depositions and median monthly frequencies of precipitation-bearing SE-SW-NW flows (sulfur) and SSW-NW flows (chloride). For chloride, most of the secular variations at Plonninge could be explained with this index. The method was not so successful at Flahult, probably because the upwind are was too wide at that station: the principal source for chlorides is the sea, and SSW winds have not generally had a recent over-water trajectory before reaching Flahult. For sulfur, the secular variations in the frequencies of precipitation-bearing SE-SW-NW flows were very closely in phase with median sulfur depositions 1951–1969. There was evidence, however, for a residual upward trend in sulfur depositions. A linear regression analysis that included both weather frequencies and time as dependent variables yielded a best estimate for this trend of 2 to 3 % per year, a figure not in disagreement with estimates of the rise in emissions of anthropogenic sulfur. DOI: 10.1111/j.2153-3490.1971.tb00542.x