Dear Authors,

The referees have raised several issues that need to be taken into consideration in a revised version of the paper. Generally, your replies to the referee's comments sufficiently consider these issues and I encourage you to revise your manuscript accordingly.
Especially, please address the point of referee #2 with respect to providing some uncertainty/stability estimate of the variogram parameters within each time window (e.g. with a split-sampling test within the time window, or by making the time-windows overlapping): This will be an important prerequisite to evaluate the relevance of changes of the variogram parameters between the time windows.
Also, please pay special attention to the comment of referee #3: make more clear the differences and advantages of your variogram-based approach compared to analysis of time series of alternative signatures derived from moving-windows (as the referee correctly states, the moving-window approach could be applied to any signature). In this context, please also refer to the existing literature on 'classical' approaches of time-series analysis (AR, MA, ARIMA, etc.), which essentially do the same as your variogram approach, i.e. to quantify the temporal autocorrelation structure of a time-series across a range of time-lags. Some related literature is given below.

Yours sincerely,

Uwe Ehret

Chun, K. P., Wheater, H., and Onof, C.: Prediction of the impact of climate change on drought: an evaluation of six UK catchments using two stochastic approaches, Hydrol. Process., 27, 1600-1614, 2013.

Montanari, A., Rosso, R., and Taqqu, M. S.: Fractionally differenced ARIMA models applied to hydrologic time series: Identification, estimation, and simulation, Water Resources Research, 33, 1035-1044, 1997.

Yang, G. X. and Bowling, L. C.: Detection of changes in hydrologic system memory associated with urbanization in the Great Lakes region, Water Resources Research, 50, 3750-3763, 2014.

Box, G. E. P., and G. M. Jenkins, Time Series Analysis: Forecasting and Control, revised ed., Holden-Day, Merrifield, Va., 1976.

Salas, J. D., C. D. Boes, and R. A. Smith: Estimation of ARMA models with seasonal parameters, Water Resour. Res., 18, 1006-1010, 1982.

Dear Authors,

Let me congratulate you to the substantially improved version of your manuscript. The referees and me agree that it is now in a state to be published.

Please consider the following final points:
- The text format is wrong (bold), starting with section 7
- The header of section 8 (conclusions) is missing
- The section with the references should be #9, not #8 (if 8 is the conclusions)

Referee #3 suggested to add in the discussion section some words on the role of data quality issues to the temporal changes of the variogram parameters. This can be an interesting addition, but it is in my eyes not necessary. If you intend to add related text, please let me see the revised version before proceeding with the final publication steps.

Sincerely,
Uwe Ehret