Although stage–discharge relationships are crucial for discharge estimations and hydrological analyses, few efforts have been taken to assess their temporal alterations in the context of dam regulation. Here, the upper Yangtze River basin serves as an example to demonstrate the influence of hydraulic structures on stage–discharge relationships evolution. Daily records of water level and river discharge from 1950 to 2013 at Yichang hydrometric station were grouped and analyzed. Back-propagation artificial neural network was used to model the stage–discharge relationships. The obtained curves revealed substantial shifts since the Gezhouba Dam (GD) and Three Gorges Dam (TGD) were put into practice sequentially. In low flow scenarios, the decline of water levels due to GD and TGD regulation were variable with river discharge, whereas in normal flow scenarios, the rating curves indicate equilibrium state with almost the same slopes regardless of GD and TGD influence. In high flow scenarios, the rating curves representing natural condition, GD, and TGD regulation intersect with each other. Moreover, the detected changes in stage–discharge relationship were mainly in response to dam regulation, channel erosion and sand exploitation, while irrelevant to precipitation variability. The contribution of sand mining, GD regulation, and TGD regulation on rating curve variations at Yichang station were 36%, 11%, and 53%, respectively.
- artificial neural network
- rating curve
- upper Yangtze River
- First received 13 January 2015.
- Accepted in revised form 25 April 2015.
- © IWA Publishing 2016