Application of statistical criteria for identifying outliers in the analysis of hydrochemical characteristics of small lakes in Karelia

The hydrochemical characterization of small lakes typical for different landscapes of the Republic of Karelia was compiled on the basis of statistical assessments. The test sample to be studied was formed on the basis of hydrochemical data on 223 most studied small lakes of Karelia, divided into groups based on their belonging to one or another landscape. The possibility of identifying atypical (anomalous) lakes for a certain territory, including unique lakes, has been assessed by applying statistical criteria for identifying outliers in the analysis of hydrochemical data. The search for anomalous values was carried out by classical methods of statistical analysis, among which the graphical method of box diagram and the statistical criteria for finding outliers Chauvenet and Roscher (Thompson). The effectiveness of these methods in detecting anomalies was evaluated. An expert analysis of the obtained sample of lakes with anomalous values of hydrochemical parameters was performed in order to identify the nature and causes of anomalies, as well as the uniqueness of lakes. The ways of further research on the possibility of using mathematical statistics methods to search for anomalies, as well as the development of criteria for the hydrochemical uniqueness of lakes are outlined.

lake ecosystem, lake hydrochemistry, unique lakes, anomaly search, statistical outliers