Elucidate the interaction between catchment and in-stream processes by using high frequency multivariate and multisite data

Hydrological processes perform a fundamental role in ecosystem dynamics. Understanding the complex temporal variations of relevant parameters is essential for the effective management of catchments and watercourses. However, when faced with large data sets and a patterns to disentangle, combined with uncertainty about which variables are causing the most significant impact, valuable insights can be lost in the complexity. The inability to distinguish significant patterns and key variables can limit a thorough understanding of the analyzed phenomenon, thus limiting informed decision making. The primary objective of this research was to explore the application of high-resolution multivariate data, collected from diverse locations, in discerning the complex interplay of watershed and in-stream processes within stream ecosystems. To achieve this goal, Principal Component Analysis (PCA) of time series, also referred to as empirical orthogonal functions, was employed to identify prevalent patterns among the various variables under consideration. The dataset used in this study was derived from five monitoring stations located within the Bode river basin in Saxony-Anhalt, Germany. Six key variables were continuously measured at each station, i.e., electrical conductivity, pH, NO3 concentration, turbidity, water temperature, and water discharge. Consequently, a total of thirty time series were analyzed. Measurements were carried out at 15-minute intervals spanning the temporal range from 2013 to 2020. The first principal component explained 46% of the total variance and described the typical effect of stream discharge fluctuations on water quality compounds. In contrast the second component, explaining 13% of the variance, grasped discharge of upwelling saline groundwater in the downstream reaches of the river during stream low-flow. The third component, contributing 7% of the variance, was closely related to diurnal oscillations of pH. As other effects on pH have been factored out by the first two principal components this pattern could be assigned to the photosynthetic activity of algae. It is remarkable that it did not only show seasonal pattern but also substantial short-term variability as well as large variability between different stream reaches and between years. In particular the hot years of 2019 and 2020 stand out in this regard. The application of high-resolution multivariate data and the utilization of PCA proved to be helpful in disentangling complex interactions within stream ecosystems. These findings emphasize the importance of advanced analytical techniques in unraveling complex hydrobiogeochemical dynamics. As freshwaters continue to face environmental challenges, further studies employing similar methodologies can enhance our ability to predict and mitigate the impacts of anthropogenic and natural factors on aquatic ecosystems.