IWRA Proceedings

AbstractINTRODUCTION Deviations of climatic elements from average levels caused by, e.g., changes in the atmospheric circulation modify the conditions in which the streamflow forms. The river regime is controlled by both, precipitation and air temperature, whose magnitudes show a significant dependence on the intensity of zonal circulation. A simple indicator characterising the atmospheric circulation over the north Atlantic is the North Atlantic Oscillation Index (henceforth, NAO). In winter, the discharges of north European rivers are positively correlated with the winter NAO index, and those of Southern Europe are correlated negatively. While in Northern Europe this relation can be accounted for by an increase in precipitation in a positive stage of the NAO and in Southern Europe by its decrease, in Central and Western Europe the dependence of winter precipitation on the NAO is rather weak. A strong relation does exist, however, between the NAO indices and air temperatures. Temperature controls the magnitude of water loss through evaporation in summer and the development and disappearance of snow cover in winter. In Central Europe during a warm winter (NAO+) snow cover dwindles, hence meltwater floods are rare and carry small discharge volumes. In a negative NAO stage, in turn, winters are severe with a thick snow cover, thus contributing to high and large-volume flood waves. The aim of the research was to examine the extent to which the North Atlantic Oscillation affects the discharges of Polish rivers. METHODS The analysis embraced annual, seasonal and monthly discharges of 100 Polish rivers taking into consideration Hurrell's winter NAO index and Rogers' monthly and annual NAO indices from the years 1951-2000. The selected rivers are distributed evenly throughout Poland and represent a diversity of environmental conditions. Differences were calculated between the monthly and seasonal flows in the years with high (>2) and low (<-2) values of the NAO index and their statistical significance was determined. To examine the relations between the discharges and the North Atlantic Oscillation, linear correlation analysis was employed. Finally, a classification was made of the rivers in terms of the similarity of strength of the correlation between their monthly discharges and the NAO index. In the classification procedure use was made of Ward's method of hierarchical grouping. RESULTS It follows from the analysis of the coefficients of correlation between the winter NAO index and monthly discharges that the correlations are positive on the rivers of entire Poland in January and February. The strongest statistically significant positive correlations are observed in January on mountain rivers (p<0.05) and in January and February on the rivers of northern (p<0.05) and north-eastern Poland (p<0.001). In March positive correlations keep dominating on the rivers of northern Poland, although they are less significant statistically, while in central Poland statistically significant negative correlations start to predominate. In April the relations change. Throughout the country negative correlations can be observed, with the strongest on the rivers of north-eastern Poland (p<0.01) and central and south-western Poland (p<0.05). The situation is similar in May, with the significance of the negative correlation coefficients increasing in central Poland and decreasing in the south-east. In the remaining months, no statistically significant relations can be observed on the majority of the rivers under study. An exception is August when negative statistically significant correlations (p<0.05) are recorded for mountain rivers. The obtained results show that the North Atlantic Oscillation has the greatest effect on river discharges in the winter-spring months, i.e. in the period when the most abundant water resources are being formed in Poland. Hence, in a positive NAO stage one might expect a considerable dwindling of the water resources almost throughout the entire territory of the country. The diversity of environmental conditions in the catchments of the analysed rivers gives the observed relationships a different character, which was confirmed by a typological classification of the rivers by strength of the correlation between their monthly discharges and the NAO index. The results of the grouping made it possible to distinguish ten regions in Poland differing in the strength, statistical significance, and temporal pattern of the relations between the discharges and the NAO index.