PubAg

Main content area

Dealing with artificial reservoirs and diversions in alpine rainfall-runoff forecasting

Author:
Wesemann, Johannes, Holzmann, Hubert, Schulz, Karsten, Herrnegger, Mathew
Source:
Österreichische Wasser- und Abfallwirtschaft 2018 v.70 no.9-10 pp. 485-496
ISSN:
0945-358X
Subject:
engineering, hydrologic models, runoff, time series analysis, uncertainty, watersheds, Austria
Abstract:
A large number of the Austrian runoff gauging stations is influenced by reservoir management or artificial diversions. The measured time series are of limited use for the design and calibration of hydrological rainfall-runoff models since they can usually only reproduce natural flow systems. Especially in alpine headwater catchments, the loss of the natural dynamics and the seasonal shift of the discharge can be challenging. Furthermore, small scale information on the subbasin level is usually missing. Despite the missing discharge data, calibrated hydrological rainfall-runoff models are required for practical uses and engineering problems like reservoir management.In this study, it is demonstrated how the natural discharge can be derived for very small subbasins from secondary information such as reservoir water levels, measurements at water intake structures and diversions for the highly disturbed Stubache catchment in Salzburg. This data is further used to calibrate the rainfall-runoff model COSERO. For subbasins with reservoirs, the local discharge is derived by balancing all inflows and outflows and the change in stored volume. Partly, the time series from the water intakes, the diversions and the pressure lines can be used to estimate the flows from the feeding catchments. Uncertainties in the measurements and time shifts are being compensated by temporal aggregations leading to time series with 12- and 24-hours resolution. The hydrological simulations show mainly good results. For the Nash-Sutcliffe Efficiency and the Kling-Gupta Efficiency top values of 0.79 (0.79) and 0.85 (0.79) and mean values of 0.64 (0.57) and. 0.71 (0.58) could be achieved for the calibration (validation) period.
Agid:
6158888