The present study associates observed changes in streamflow patters in four large basins of the Midwestern Unites States to changes in precipitation, evapotranspiration and human interventions, particularly the increasing cultivation of soybean and corn enhanced by artificial drainage. It analyses 79 years (1935 - 2013) of precipitation, streamflow, artificial drainage, and cultivation data, to provide a extensive statistical time series analysis, identifying breakpoint years that evidence the hypothesized changes. Additionally, the changes in basin storage are associated to changes in climate or in land-cover using a simple water budget. The paper concludes that artificial drainage as part of large agricultural development in the Midwestern US amplifies the changes in rainfall-runoff response because of an increased hydrologic connectivity and a reduced storage in the basin.
This version is a minor revision of the original submission that I reviewed a few months ago; therefore, I do not see significant changes to address successfully all of the referee comments. Most of the improvements have been made on the form of the text and in a lesser extent to address the several criticisms on the document provided by the different reviewers. I would be more keen to see a thorough revision of the manuscript than a justification of why such changes have not been assimilated.
Particularly, with respect to those provided in my revision, I keep the main major comments as unsolved:
- It is not a stylistic request to divide results from discussion. It is indeed difficult to separate them because of the tendency to confuse result description from interpretation. This is a skill necessary to develop in order to be objective at drawing conclusions and remove speculations or interpretations that are not completely based on the current data and results.
- The study uses 7 metrics to analyze the streamflow regime that might be redundant as these are related to flow magnitude mainly. In the previous revision, I have recommended to include other hydrological indices that relate to the different flow attributes (magnitude, frequency, duration, timing, and flashiness). The authors could perform a correlation analysis between the metrics, gauges, and catchments, or a Principal Component Analysis to either demonstrate that the indices they are using are indeed independent, or to identify more informative hydrological indices that characterize streamflow more comprehensively.
I judge the paper could be suitable for publication after these revisions. |