Optimizing Discharge Gauging Station Placement Using Flow Regime Correlation Analysis

Stream gauging stations are artificial installations used to monitor streamflows in rivers. They provide fundamental informations about and water availability for human activities and river ecosystems. Flow-measuring stations, however, provide pontwise estimates of water resources, making sometimes difficult to chacarterize the spatial distribution of flow regimes. This is particularly true if one considers that flow-measuring stations are often unevenly distributed along river networks, and may provide redundant or poorly representative information on discharge variations. In this context, adopting streamflow correlation as an indicator of similarity between recorded streamflow series can represent a proper method to identify suitable river branches for the installation of new gauging stations. In this thesis the maximum streamflow correlation between a new test section and the already placed flow-measuring stations is used to identify optimal discharge gauging station placement. To this aim an analytical model for predicting discharge correlation is adopted. This model, proposed by Betterle et al. 2017, confronts the precipitation regime contributing to multiple river sections and quantifies the cross correlation between streamflows at two arbitraty river sections even in the absence of streamflow time-series. Considering frequency of joint (i.e. synchronous) and disjoint (i.e. non-synchronous) runoff-producing rainfall events and precipitation depth correlation of joint events in the contributing areas of confronted sections, the distribution of streamflow correlation between multiple test sections and existing stations is obtained. The application presented here refers to the Argen catchment in Germany. Therein correlation thresholds are used to identify suitable river branches for the installation of new flow-monitoring stations on the river network, and inadvisable locations where a new station would provide limited additional information on flow regimes. Results indicate that the northwestern area of the basin has persistently low correlation with already monitored river sections, which makes it a suitable location for installing a new flow-measuring station expected to provide complementary and non-redundant information on local flow-regime dynamics. The method provides a parsimonial operational tool for the installation of low redundancy streamflow gauges.