Schaller et al. present decadal rates of physical erosion and chemical weathering for the Swabian Alb and its foreland. The partitioning of denudation into weathering and erosion is calculated analogous to the chemical depletion fraction, and all data are interpreted in the context of climatic, topographic, biologic, and geologic variables. Their findings reveal generally higher and more variable erosion and weathering rates in the Neckar tributaries compared to the Danube, despite notable scatter in the data. The persistence of this general trend is linked to the higher baselevel of the Danube tributaries. The topic of the study is well suited for ESurf. Particularly, the partitioning of denudation in erosion and weathering is of interest—a subject area where more data from carbonate landscapes are essential to better examine the factors that control this ratio. However, anthropogenic influences, which are currently not considered, may significantly affect decadal-scale erosion rates and potentially influence weathering processes. Additionally, a more detailed description of the methods is necessary for clarity and reproducibility.

Human influence: Currently, the study does not take into account human influences on the decadal erosion and weathering rates. The correlations between erosion/weathering and topographic/climatic/geologic variables are weak and a possible explanation would be that the rates are modified by human activity.

Parts of the study area are highly industrialized and virtually all of the catchments host significant amounts of agriculture. Moreover, all rivers are heavily modified, channelized, contain hydropower plants of various sizes, canals for mills and factories, flood retention channels, etc. All of these anthropogenic factors should have an effect on erosion and sediment connectivity. For instance, it has been shown that suspended sediment rates in agricultural catchments in Europe can be 40 times higher compared to natural conditions, but this effect is highly scale-dependent due to sediment connectivity (Vanmaercke et al. 2011, 2015). The moderately-sized catchments studied here should be susceptible to these effects.

Moreover, sediment connectivity and erosion likely change through time due to changes in land use and river engineering. The study cites very long averaging windows for the data of several decades. It has been shown that river sediment yields are decreasing over the past decades in the Northern Hemisphere, and in Germany in particular (Dethier, Renshaw, and Magilligan 2022; Hoffmann et al. 2023). The problem is that the reasons for this decrease are not entirely clear. Since most dams were built earlier, it might rather be related to changes in agricultural practices and/or urban development decreasing hillslope-channel coupling. However, the unknown driver behind these changes make it challenging to select appropriate predictor variables for simple correlations as used here.

All this to say, the human influence on the physical erosion rates needs to be assessed. I encourage the authors to investigate whether variables that capture human activity, such as the human footprint index, % agriculture per catchment, the connectivity status index of rivers (e.g., Grill et al. 2019), etc. better explain the distribution of physical and total erosion rates.

Similarly, chemical weathering may also be affected by human activity. Soil CO₂ is the main source of acid for the dissolution of limestone, and is modulated by vegetation, such that land use differences between catchments may matter in terms of weathering. The land use effects on carbonate weathering have been shown on a global scale (Zeng, Liu, and Kaufmann 2019).

The use of CDF: The chemical depletion fraction has been based on the ratio of element concentrations (mainly [Zr]) in the bedrock and regolith and assuming steady-state soil formation and denudation (Riebe, Kirchner, and Finkel 2003). As such, a CDF represents the long-term contribution of weathering to denudation reflected in the bedrock-regolith composition difference. However, the authors present contemporary rates of erosion and weathering from river sediment and water chemistry. Therefore, I find the use of CDF a bit misleading because the data do not reflect the time-scale of standard CDF measurements. I suggest referring to denudation partitioning, percentage weathering, or similar.

Just a suggestion, but since the denudation/erosion/weathering rates are low, mm/ka might be the better to unit to report rates and avoid all the zeros.

The use of Ksn: The interpretation of Ksn is a bit tricky for this study area, when comparing mostly non-karst (Neckar) to karst-side (Danube) catchments. The relationship between drainage area and discharge is non-trivial in karst landscapes. Therefore, a comparison in Ksn can be difficult or even misleading. One could either use the discharge stations to check for drainage area-discharge scaling differences, or at least add a short statement outlining the problems of standard Ksn in karst.

More information on the chemical weathering calculation are required. Currently, it is not mentioned how the authors go from ion concentration to a weathering rate.

• What discharge data are being used?
• What kind of mineral dissolution is assumed? Probably best to show the equations used for the calculation.
• How are uncertainties propagated? Fig. 3 shows asymmetric error bars, and almost all data overlap within error. How where these errors computed?

Significantly more information are required on the calculation of decadal erosion rates. There are a lot of choices to be made to go from a sediment rating curve to a sediment yield estimate:

• The authors should report the equation fitted and the a and b values (assuming Q_s = a * Q_w^b) as these also hold additional information about sediment transport.
• If the authors fit the rating curve in log-space, what log-transformation bias correction was applied?
• What was the average number of measurements per rating curve? And what r² values did you get?
• Were quality controls applied, such as minimum amount of data points, minimum r² value?
• Did the authors check for hysteresis in the Qw-Qs data? Frequently, many measurements are taking around flood events, which can bias rating curves due to hysteresis.
• The authors cite long time spans for the station records. However, rating curves change over time. Are the entire time periods used for fitting the rating curve? And do rating curves for different river represent different time intervals? Usually rating curves change through time because of human-induced effects on sediment connectivity (Dethier et al. 2022; Warrick 2015) and need to be re-fitted based on a window of interest. For instance, in Germany suspended sediment loads have been decreasing over the past decades (Hoffmann et al. 2023). Therefore, one cannot directly compare erosion rates from rivers with rating curves based on different time intervals or average long-term suspended sediment load. One could decide on a specific decade for analysis or show that temporal trends are negligible.
• Most of these questions are also relevant to the other erosion calculation methods, such as the one with average values.

Working my way through the results, I suggest that the authors use more informative names for their rate estimates instead of numbering them. I found it almost impossible to remember what CWR 1,2,3 , PER 1,2,3, TDR 1,2,3, and CDF1,2,3 mean. Can this be simplified to min, max, best guess? Or PER_bed, PER_TDS etc.

I find CWR3 a bit distractive in the manuscript. I think it’s good that the authors acknowledge the fact that secondary calcite precipitation is common in the Swabian Alb and that may lead to an underestimation of weathering if only looking at river chemistry. However, the rates are unreasonably high and the plots with CWR3 and CDF3 could be moved to the supplement. Many studies have calculated chemical weathering rates based on spring water chemistry, such as Hoenle 1991. These data should not be biased by secondary calcite precipitation and the average for the Swabian Alb was 52 mm/ka. Therefore, my recommendation would be to give these maximum estimates less exposure in the manuscript making the paper easier to follow. Also, the authors could compare their rates to the ones previously calculated from spring water chemistry to estimate the amount of secondary calcite precipitation.

Line comments:

L10: If we talk about contemporary rates, anthropogenic influences need to be considered.

L52: No need for granitoid lithologies to measure ¹⁰Be. Please, replace with quartz-bearing lithologies or similar.

L57-59: I don’t follow. A CDF is based on the ratio of two concentrations (bedrock vs saprolith, saprolith vs soil). No need for corrections. Are the authors referring to the need to correct cosmogenic nuclide-derived denudation rates with CDF measurements for weathering below the production zone? Please, clarify or correct this statement.

L 69-70: Please, make the time-scales more general. River dissolved loads can integrate over hours; cosmogenic nuclide rates can integrate over hundreds of years depending on the rate. Maybe use powers of ten. For instance, 10³ to 10⁴ for cosmogenic nuclides.

Fig 1. Make sure it is clear that the right-hand side is only that complicated in the case of weathering. In an arid environment, cosmogenic nuclide measurements would be tracking denudation.

L96: Please, add references.

L102: Please, add references.

L106: It’s not just a regional drainage divide. It’s part of the continental drainage divide.

L 119-124: Can you please add the methods used for estimating these rates?

L121-124: Studies that are missing are for instance the overview work of Hönle 1991 that constrains the average chemical weathering rate of the Swabian Alb to 52 mm/ka, and works from University of Tübingen (!) such as Poppe (1993) and Bauer (1993). Probably, there’s more buried in the German literature.

L 160: Any justification for 0.45? Many studies use 0.45. However, since the authors are using TopoToolbox, one could use the ‘mnoptim’ function to check what works best for steady-state channel sections in the area. Though, I understand that heterogeneous lithology might create difficulties here. Nevertheless, please justify the choice of mn.

L165-172: Would be good to state how the NDVI time window compares to the measurement time window of erosion and weathering rates.

L173-179: Can you explain on what parameter the geologic binning is based on? Is this based on assumed weatherability, erodibility, or something else? I am confused as to why the basement rocks in the Black Forest end up in the same category as the Keuper evaporites, and Jurassic marls.

Table 3: I assume that the categorical variables for geology are converted to % of catchment area? Can you please clarify this in the table or caption, otherwise it is hard to follow what a correlation coefficient of -0.4 between CWR and Lower Triassic is supposed to reflect.

Also Table 3: It just says CWR/PER/TDR. Can you please add the number qualifiers to the column headings? I had to search in the text, which numbers are being displayed.

Discussion 5.1.3. Why are these rates only put into global context, and not with regional studies? And what would these rates mean for long-term landscape evolution of the Swabian Alb foreland How do they compare with previous estimates?

Figure 6: I think this figure would be easier to read if the horizontal bars are removed. The information could be displayed or categorized differently. For some categories different colors refer to geographic location (Danube vs Neckar) for other symbols the color refer to the references. I recommend to revise the layout of this summary figure to make it easier for the readers to follow.

Feel free to contact me in case there are any questions about this review.

Richard Ott.

References

Bauer, Michael. 1993. “Wasserhaushalt Und Losungsaustrag Im Wutachgebiet.” Pp. 189–202 in Eintiefungsgeschichte und Stoffaustrag im Wutachgebiet (SW-Deutschland), edited by G. Einsele and W. Ricken. Tübingen: Tübinger Geowiss. Arbeiten (TGA).

Dethier, Evan N., Carl E. Renshaw, and Francis J. Magilligan. 2022. “Rapid Changes to Global River Suspended Sediment Flux by Humans.” Science (New York, N.Y.) 376(6600):1447–52. doi: 10.1126/SCIENCE.ABN7980/SUPPL_FILE/SCIENCE.ABN7980_SM.PDF.

Grill, G., B. Lehner, M. Thieme, B. Geenen, D. Tickner, F. Antonelli, S. Babu, P. Borrelli, L. Cheng, H. Crochetiere, H. Ehalt Macedo, R. Filgueiras, M. Goichot, J. Higgins, Z. Hogan, B. Lip, M. E. McClain, J. Meng, M. Mulligan, C. Nilsson, J. D. Olden, J. J. Opperman, P. Petry, C. Reidy Liermann, L. Sáenz, S. Salinas-Rodríguez, P. Schelle, R. J. P. Schmitt, J. Snider, F. Tan, K. Tockner, P. H. Valdujo, A. van Soesbergen, and C. Zarfl. 2019. “Mapping the World’s Free-Flowing Rivers.” Nature 569(7755):215–21. doi: 10.1038/s41586-019-1111-9.

Hoffmann, Thomas O., Yannik Baulig, Stefan Vollmer, Jan H. Blöthe, Karl Auerswald, and Peter Fiener. 2023. “Pristine Levels of Suspended Sediment in Large German River Channels during the Anthropocene?” Earth Surface Dynamics 11(2):287–303. doi: 10.5194/esurf-11-287-2023.

Poppe, R. 1993. “Karstsystem Und Lösungsaustrag Im Oberen Jura Des Aitrachtals.” Pp. 181–88 in Eintiefungsgeschichte und Stoffaustrag im Wutachgebiet (SW-Deutschland), edited by G. Einsele and W. Ricken. Tübingen: Tübinger Geowiss. Arbeiten (TGA).

Riebe, Clifford S., James W. Kirchner, and Robert C. Finkel. 2003. “Long-Term Rates of Chemical Weathering and Physical Erosion from Cosmogenic Nuclides and Geochemical Mass Balance.” Geochimica et Cosmochimica Acta 67(22):4411–27. doi: 10.1016/S0016-7037(03)00382-X.

Vanmaercke, Matthias, Jean Poesen, Gerard Govers, and Gert Verstraeten. 2015. “Quantifying Human Impacts on Catchment Sediment Yield: A Continental Approach.” Global and Planetary Change 130:22–36. doi: 10.1016/j.gloplacha.2015.04.001.

Vanmaercke, Matthias, Jean Poesen, Gert Verstraeten, Joris de Vente, and Faruk Ocakoglu. 2011. “Sediment Yield in Europe: Spatial Patterns and Scale Dependency.” Geomorphology 130(3–4):142–61. doi: 10.1016/j.geomorph.2011.03.010.

Warrick, Jonathan A. 2015. “Trend Analyses with River Sediment Rating Curves.” Hydrological Processes 29(6):936–49. doi: 10.1002/HYP.10198.

Zeng, Sibo, Zaihua Liu, and Georg Kaufmann. 2019. “Sensitivity of the Global Carbonate Weathering Carbon-Sink Flux to Climate and Land-Use Changes.” Nature Communications 10(1):1–10. doi: 10.1038/s41467-019-13772-4.

General comments

In their manuscript ‘Spatiotemporal denudation rates of the Swabian Alb escarpment (Southwest Germany) dominated by base-level lowering and lithology’ Schaller et al. investigate decadal-averaged chemical weathering and physical erosion rates for a range of river systems draining both sides of the Swabian Albs into the Danube and Neckar rivers. In general, most catchments are dominated by chemical weathering over physical erosion, and total denudation rates (chemical weathering + physical erosion rates) are on average two times higher on the north side of the escarpment draining into the Neckar compared to tributaries draining southwest to the Danube. The authors interpret higher rates in the north to be caused by ongoing escarpment retreat and drainage capture events (caused by base level differences). To better understand spatial variability in erosion and weathering rates, the authors compare rates to topographic, climatic and lithologic catchment parameters and consider lithological differences as an important control parameter on weathering rate differences.

While a detailed disentanglement of rates and control factors for denudation processes is useful and timely, and the authors put quite some effort into compiling a large dataset, I have three important comments on the current version of the manuscript. I will mention them briefly here and explain them in more detail below. I suggest that these comments will be addressed before publishing this study in ESurf. First, the study should be streamlined more by matching the focus of the introduction with the rest of the manuscript or adjusting the introduction accordingly. In particular, a more detailed motivation, a clear research question and a discussion of the wider implications of the results would strengthen the relevance of the study. Second, the manuscript does not contain sufficient documentation of the methods used. Therefore, reproducibility cannot be guaranteed, nor can it be assessed in all cases how reliable the different approaches are, especially when calculating weathering and erosion rates. Third, the data on water discharge, total suspended solids (TSS) and total dissolved solids (TDS) collected over several decades are susceptible to anthropogenic influences, e.g. agriculture, river dams and discharge controls at sluices. The anthropogenic influences on these time series, and the erosion and weathering calculated from them, are largely neglected. Points 2 and 3 were also addressed by the other reviewer, who gave very constructive and detailed feedback on point 3 in particular. I fully agree with his comments and will therefore focus on points 1 and 2 below.   

Specific comments

Structure

• The introduction largely focuses on the different methods to disentangle erosion and weathering, but the rest of the paper mainly focuses on the specific case study of the Swabian Albs. After reading the introduction, I expected a more methodology focused paper that tests a new method to differentiate erosion and weathering rates. However, the remaining paper reads more like a case study investigating denudation patterns of the Swabian Albs. I suggest to streamline the paper by making clear from the beginning which direction this study is going. If it is supposed to be a more method- based paper, then the proposed approach should be evaluated in the discussion against other available methods mentioned in the introduction. However, if the general idea is to investigate spatial (and temporal) patterns of weathering, erosion, and denudation in the Swabian Albs, including potential control factors, then I suggest to rewrite the introduction with a clearer focus on such study and also discuss the broader implications of the findings. What can we learn from the Swabian Albs about other sites, will the data help us to improve our process understanding, will the data help us to improve numerical landscape evolution models? The broader impact of the work should also be mentioned at the end of the abstract and in the conclusions.

• Related to the point above, I had difficulties following the introduction about the different available methods on separating erosion and weathering rates (mainly lines 46-67). For example, ‘decadal-scale catchment-wide denudation rates (D) can be determined by making use of river discharge (Q) and total suspended and dissolved solids (TSS and TDS, respectively), allowing the determination of physical erosion (E) and chemical weathering rates (W).’ But how does this work? Without the extra explanation, it is complicated to follow the individual approaches presented here. Also, the text refers to figure 1, however many parameters mentioned in figure 1 are not explained at all (e.g. Zircon-related parameters, erosion and weathering in the saprolite). In fact, the entire box model shown is not really explained as a model as a whole. On the other hand, not all parameters mentioned in the text show up in the model. For clarification, I suggest to rewrite this part of the introduction (if it stays) and make sure that the text better links to the figure and to point out what the general research aim of this study is (e.g. testing the box model, or finding a different approach for the equations in panel B).

Methodology

Individual steps of the analyses need further explanation, as they are not reproducible in the current state.

• Calculation of weathering and erosion rates (lines 187-197): How do you deal with the fact that measurements of discharge, TTS and TDS were taken at different locations when calculating chemical weathering and physical erosion rates? Are these values somehow projected to the point locations giving in the tables?
• Water discharge values play a crucial role in the calculations of chemical weathering and physical erosion rates. Discharge values were collected over a few decades (line 191). One crucial information that is missing here is to what degree the discharge in those rivers is anthropogenically controlled (dams, sluices, etc.).
• Which datasets of discharge, TSS and TDS were used for the calculations? Just saying that the data comes from various online available databases is not sufficient (lines 187-189).
• The three different approaches for calculating decadal-averaged chemical weathering rates (lines 198-214) and physical erosion rates (lines 215-220) are not described with enough detail to ensure reproducibility. Please enhance those descriptions by providing all necessary information. Also, it will help to briefly summarize the main assumption behind every approach. For example, the values derived by CWR3 are about 3 times higher compared to those of CWR1 and CWR2, implying that the underlying assumptions of each approach play a major role on the final values. The assumptions will help to get a better feeling for how reliable each of the datasets is. Another example is PER2, where it says that ‘PER 2 is estimated from abundant discharge values using the empirical relation between individual discharge and suspended solids’ (lines 217-218). How does the discharge and suspended solid data look like, and what is the equation for the empirical relationship? Further below, it is mentioned that a power-law function is used (line 270), but the whole approach should be explained in the method section and the analysis potentially shown in the supplement.
• Lines 212-214: Provide a very brief summary of how the method of Campbell et al. (2022) works, such that the reader can follow the approach here without reading the referenced paper.
• Lines 228-230: Please briefly justify why you pick those approaches out of the longer list of options.
• Lines 232-235: Provide a short summary of how escarpment retreat rates were calculated instead of only referring to previous literature. At least mention the main underlying assumption for the transformation. Also, which locations were chosen for calculating the according catchments? In table 2, only river names are listed. Please add the location used for calculations to ensure reproducibility.
• Section 4.2: What kind of statistical tests were performed here? In particular, which model was used underlying the tests? Simple linear regressions? If so, does it make sense? Giving that for example denudation rates and mean catchment hillslope gradient or mean ksn are known to correlate non-linear, a simple linear regression is not sufficient. Also, it might help to have different color schemes for positive and negative R2 values in table 3, to clearly distinguish between positive and inverse correlations.    

 Technical corrections

•  Line 11: Unclear what ‘these rates’ refer to. Is it about disentangling erosion and weathering rates or the impact of tectonic, lithology, climate and biota on these rates? Please clarify.
• Line 17: ‘…published longer-term rates…’ What type of rates? Denudation rates?
• Line 17: ‘…evaluate how these differences…’ What differences, between short and long timescales or between weathering and erosion?
• Line 22: Shouldn’t it be ‘dominant denudation process’ instead of ‘erosion’?
• Line 23: What are chemical sedimentary rocks? Please clarify.
• Line 46: What does the ‘these’ relate to? Please clarify.
• Figure 2: The figure contains a lot of information using different colours, which are not so easy to see. For example, the Neckar river is hard to recognize. Maybe making the elevation colours more transparent will improve the recognition of the study sites and analysed catchments. In addition, I suggest to add a topographic cross section of the Swabian Albs to the figure, ideally with the underlying lithological units. This will help to reader to follow the description of the study site.
• Line 147: What does ‘these’ refer to? Which metrics exactly?
• Lines 168-172: NDVI is mentioned in the method section, but no data is presented later in the study. Consider removing.
• 2. I suggest to mention somewhere early in this section the number of stations that are analysed. Some numbers are presented later, but it would be helpful to get a good idea of the dataset early on.
• Line 238: Which part of the study area is considered as the plateau? Could it be indicated in one of the maps?
• Line 267: What is meant by left-side tributaries of the Neckar river? Those draining the foreland and not the Albs?
• Line 270, point 6: In line 219 the approach is described differently, please clarify.
• Line 350: Misplaced reference to figure 4, or at least unclear what figure 4 is referenced for.
• Line 371: Did you mean to refer to figure S4? Figure 4 shows no information on lithology.
• Line 372: I don’t follow why Figure S5 is reference here, the figure shows no information in Opalinus clay.
• Line 381-384: This should be moved to the method section.
• Line 387-389: Is the mismatch between Swabian Albs retreat rates and global retreat rates due to a special setting of the Alb sir rather a methodological bias, based on the way the catchment area is calculated?
• Figure 6: Numbering (A,B,C,…) in caption is misleading, as there are no sub-plots. Looks like some error bars are of different colours than the according symbol. (C), what kind of DEM analysis is used to calculate surface lowering rates?
• Line 428: ‘…to being dominated by physical erosion…’ This phrasing is misleading as the CDFs barely reach values below 0.5. I suggest rephrasing.   
• Lines 434-435: Where are the other reported values from geographically? What can be learn from comparing the values of the Swabian Albs with other sites globally?
• Line 475-492: Currently, the conclusion is only a summary of the main finding, but lacks any broader implication of the study’s finding. I suggest to add a few sentences about what we can learn from this study beyond erosion and weathering rates in the Swabian Albs to enhance to relevance of the study for a more general community.
• Figure S4: Instead of only reporting the geologic units (e.g. Lower Jurassic), I suggest to also provide additional information on the dominant lithologies of the units.