Answer to Reviewer 1

Abstract

Printer-friendly version Discussion paper for the ACC*w calculation method. The reasons for the ACC*w resulting in lower RMSE compared to ACC or w only, is now described in detail in the methods section. Overall, the results (correlations, RMSE, NCS, ∆RBTcalc) changed only so slightly, that the scientific conclusion and the key messages were not influenced. This also visible in the attached track changes version of the manuscript. GENERAL COMMENTS The manuscript presents a study of short term and long term changes in river temperature and investigates the influence of natural and anthropogenic drivers of these changes which is interesting and generally within the scope of HESS. River temperatures at various monitoring locations along the river Rhine as well as industrial production and nuclear power plant activities are analyzed. The authors further develop a novel approach of calculating a catchment-wide average air temperature which is used in the linear regression relationship between air and river temperature. Overall, the scientific approach and the methods appear to be valid. However, there are some points which need further clarification: (1) The relationships between river temperature and its drivers are investigated using multiple linear regressions separating the so-called Rhine base temperature (i.e. the river temperature without influences of air temperature and discharge) and air temperature and discharge influences on river temperature. More information on the multiple linear regressions for each location is required for the reader to be able to evaluate the robustness of this approach Comment: The RMSE and the NCS information is provided for every measurement station. In addition, the data is now included in the supplement. We used the year 2001 as a test year and regressed Tw using Tc and Q just for this year. Then the 2001 regression coefficients were used to calculate a modelled Tw for the years 2000 to 2018. The RMSE and NCS show better results compared to the long term regression, which is sensible for a shorter regression period.