Auerswald et al. HESS

General comments:

The authors present a provocative piece that positions land and soil use processes at the landscape scale as important drivers of the climate change effects on floods and droughts. The manuscript reviews the relevant processes and makes a convincing argument. Only at the beginning is the piece set up in a way that suggests landscape factors to be more important than CO2-driven climate change effects. I’m not convinced by this framing, especially as the relevant evapotranspiration argument is not fully explained. I suggest positioning the two strands of impacts – via CO2 and via landscape factors – as complementary, and maybe the landscape processes have been overlooked. But the manuscript doesn’t disentangle the two in my opinion. It would also be good to triangulate the arguments with timeseries data, specifically ET data from eddy-flux towers or lysimeters, soil sealing timeseries, soil compaction timeseries. I know especially the latter two are hard to come by, but maybe for the case study in Bavaria.

Specific comments:

L23: The “imbalance” suggested here needs more explanation.

L49-61: In this section, land use and soil use seem to be conflated. I suggest to be clear about the focus.

L127-129: The argument for a small effect of rising T on ET is crucial and hence needs more explanation. Maybe unpack the Penman-Monteith equation to pinpoint all the influences of T on ET and then convincingly show how these may be smaller than conventional wisdom holds.

Figure 3 also misses the ET argument.

L142-143: How to disentangle CO2-driven climate change from land use-driven climate change?

L150-151: What about open water evaporation from ponding on sealed surfaces?

Some of the reference in section 3.1 are also quite old. Do recent studies confirm these findings?

And how do the timelines of CO2 in the atmosphere and soil sealing compare? Does a comparison support the argument of a soil sealing-driven climate change?

L244-248: What about outflow from the river to the groundwater?

L264-266: How does soil compaction look over time?

L342: What is meant by “unfavourable behaviour” here?

This contribution aims to argue that factors such as land use and soil management are more important than increasing temperatures when it comes to their impact on floods and droughts. This is an interesting and perhaps controversial viewpoint that can help to stimulate the debate on the origins of hydrological extremes. While I can appreciate what the authors are trying to achieve, and I second many of the points made, the contribution currently contains numerous strong claims that are insufficiently rooted in either scientific evidence or logic. The contribution also suffers from a lack of clear definitions of what a flood or drought is according to the authors, and at what scale processes are relevant. As a result, their argumentation suffers. As an example: if flooding is defined as the length of (over)saturation of soils at a given location (which is perhaps not totally unreasonable), than artificial drainage will locally lead to a reduction of flooding, and not by definition to an increase downstream. This is in fact the main reason that much of Europe has seen the introduction of intense drainage. The contribution has similar weak argumentation in other places, which in my view leads to a situation in which the main proposition (Land use, soil management, and landscape hydrology are more significant drivers than increasing temperatures) is not sufficiently backed up by evidence or strong arguments. Therefor I believe the contribution should either be toned down (Focusing on the question are Land use, soil management, and landscape hydrology are more significant drivers than increasing temperatures?), or stronger arguments should be provided to back up the claims.

Detailed comments:

Line 24 and onward: Here, the authors attack the use of common statements by arguing they are solely based on correlations and not causality. This attack fails to convince. It is general knowledge that the evaporation process (the phase transitioning from liquid to gaseous phase) depends on temperature and vapor pressure deficit. This is not an example of circular reasoning. Atmospheric and soil (moisture) states are both part of a complex (and open) feedback system, and many statements on direction of feedbacks implicitly assume timescales associated with these processes. The argument on recycling, which I find weak, is a good example of mixing of scales by the authors. Soils can be dry because it hasn’t rained locally, but indeed this drying might contribute to rainfall thousands of kms away. Not sure what the authors exactly try to prove here.

Line 70 and onward: Here the authors argue that because summer rainfall does not show a clear trend, rainfall cannot explain any increases in floods. This is a false comparison. A fair comparison would be to analyse event precipitation that induced flooding. It is well possible, and in line with the arguments provided on erosivity, that extreme precipitation is showing a clear increase whereas summer rainfall in itself does not. Please provide a fair comparison.

The authors use the study of Davin et al. (2014) to argue that land management would have contributed to a shorter, less intense drought. There are several problems with this paragraph. First, the authors claim that the change of albedo would have lowed temperature country-wide by 2 K. This claim is not supported by the work of Davin et al. Yes, during some individual days during the 2003 heatwave, their simulations showed a reduction in temperature maxima of close to 2K, but only averaged over pixels with more than 60% cropland (their Fig 4). These are likely not the pixels where the high death toll mentioned by the authors took place, and the effect of cropland management on these would likely have been very small. The authors also argue that this 2K difference during the warmest days would have lead to a reduction in soil drought – seemingly in contradiction with a later statement that 2K warming does not significantly affect evaporation. The authors also seem to have done some selective shopping for arguments: in Davin et al. it is also shown that for much of France (NE, see their Fig 3) the warming impact of NOTILL for lower temperature quartiles is nearly the same as the cooling impact for higher quartiles. Please study this reference (and other references) again to make sure statements are in line with evidence provided in these works.

Line 117 and onwards: Here the authors argue that “As the droughted area heats up more than neighboring well-watered areas, the higher temperature spreads to these nearby areas and causes them to transpire more until they also run short of water. Thus, the area with reduced evapotranspiration grows and may finally spread over an entire continent, described as "event self-propagation" by Miralles et al. (2019).” Several things are wrong here. First of all, this self-propagation is never the only process operating. Please also elaborate on the role of atmospheric dynamics. And this statement seems to conflict with the arguments on recycling provided earlier. The more evaporation and drying of soils in one place, means more precipitation in other places. Again please be clear in the argumentation about spatial of temporal scales at which processes occur.

Line 130 and onwards: Here, the authors claim based on the work by Roderick et al. (2014) that a 2 K temperature rise would increase evapotranspiration by only 5 %. This is misleading. In my understanding of the Roderick study, this is based on long-term global average values including feedbacks with precipitation. For a smaller region like Bavaria and the focus on individual droughts, the sensitivity is likely much higher. I did a quick analysis based on data for De Bilt, where potential ET (by the Makkink method) has been measured since 1957. Comparing August values for 1957-1966 and 2014-2023 results in 82.9 mm and 106.6 mm, respectively, or an increase of over 28%. The same periods show a warming of 2.67 K, resulting in a sensitivity of nearly 11 %/K, much much higher than the 2.5 %/K mentioned by the authors. This analysis took me only 5 min. I expect similar results will be found for Bavaria. Please investigate a bit further than simply picking a global number not relevant to regional drought from a 10-year old study.

Line 228: “In consequence, the remaining precipitation …” -> What is remaining precipitation? Relative to what? In natural systems there will also be a considerable amount of drainage, perhaps even equal to that of drained lands (where storage on average is lower, but this does not necessarily have a strong impact on average partitioning).

Section 3.4: This is a good example of selective argumentation. When discussing the study of Davin, the authors argued that a small reduction in warming would help to counteract drought. Using the same reasoning, hedgerows, which as the authors correctly point out have a warming impact on their environment (besides many other advantages!) should lead to an increase in drought because of the higher temperatures. I personally don’t think this effect is very strong (and if so it is likely beneficial), but for this contribution it is important that arguments are consistent.