the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Geomorphological and hydrological controls on sediment export in earthquake-affected catchments in the Nepal Himalaya
Boris Gailleton
Basanta Raj Adhikari
Bishnu Raj Baral
Abstract. Large earthquakes can contribute to mountain growth by building topography, but also contribute to mass removal from mountain ranges through widespread mass wasting. On annual to decadal timescales, large earthquakes also have the potential to significantly alter fluvial sediment dynamics if a significant volume of the sediment generated reaches the fluvial network. In this contribution, we focus on the Melamchi-Indrawati and Bhote Koshi rivers in central Nepal, which have both experienced widespread landsliding associated with the 2015 Gorkha (Nepal) earthquake. Using a time series of high-resolution satellite imagery, we have mapped exposed gravel along the river from 2012–2021 to identify zones of active channel deposition and document changes over time. Counter to expectations, we show negligible increases in coarse sediment accumulation in both catchments since the Gorkha earthquake. However, an extremely high concentration flow event on 15 June 2021 caused an approximately four-fold increase in exposed gravel along a 30 km reach of the channel with up to 12 m of channel aggradation in the Melamchi-Indrawati rivers; this event was localised and did not impact the neighbouring Bhote Koshi catchment. Based on published reports, new helicopter based photography and satellite data, we demonstrate that this event was sourced from a localised rainfall event between 4500 and 4800 m, and that the majority of the sediment was supplied from sources that were unrelated to the landslides generated by the Gorkha earthquake.
Emma L. S. Graf et al.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2022-1347', Anonymous Referee #1, 30 Jan 2023
This paper asks a very straightforward question regarding what the legacy of earthquake triggered landslides are on the mobilization of sediment in river channels adjacent to landslides in the years after an earthquake. Recent events in Taiwan, China, and New Zealand all highlight the potential for earthquake-triggered landslide debris to cascade into river systems and dramatically change river morphology. This study exploits the 2015 Gorkha (Nepal) earthquake and examines changes in valley bottom gravel storage for two rivers, the Bhote Kosi and the Melachi-Indrwati, as a proxy for sediment storage.
Notably, the study shows that earthquake-triggered landslides don’t influence valley bottom sediment storage in a clear way, despite the fact that many of the landslides are adjacent to river channels. Moreover, a huge flood event in 2021, which did influence the valley bottom sediment storage in a clear and dramatic way on one of the two rivers, transported sediment that was pretty clearly not sourced from earthquake triggered landslides. Rather, very likely, the legacy of previous landslide dams (and importantly the fine sediment they impound) is likely the dominate source of sediment to river channels.
All of this suggests that how earthquakes influence river channels through the landslides they trigger is a bit more nuanced than has been articulated previously. While the legacy of landslides is important, in this example, the importance derives not from the sediment landslides generate directly (which is generally quite coarse) but rather through the sediment that is impounded behind landslide dams (which is by definition mobile in the river and hence finer).
I love this study because it takes a critical and objective look at an issue that I think many have opinions about but few bother to actually explore with data. I just have a one substantial comment below, and then a minor comment.
Well done.
Figure 4-5 (and text pertaining to): why not simply calculate the width of the exposed gravel orthogonal to the river channel as a function of the distance downstream? Plotting area versus distance is confusing because it relies on a discretization, and hence will be more difficult to reproduce in the future. I think a more straightforward thing would be to plot width versus distance, which when numerically integrated in the downstream direction would yield area. There is much less ambiguity about this measurement, and it is much more straightforward to interpret. Also, once you have area, you can easily get to volume with an assumption about aggradation depth.
Figure 6. again, to me thinking about changes in width is much easier to interpret than changes in the area of segments that have arbitrary lengths.
338-344: A similar case was documented in California in the Sierra Nevadas in 2018 and summarized in this study:
Collins, Brian D., et al. "Linking mesoscale meteorology with extreme landscape response: Effects of narrow cold frontal rainbands (NCFR)." Journal of Geophysical Research: Earth Surface 125.10 (2020): e2020JF005675.
In this work there is weather radar available that demonstrates exactly what you are speculating about here - a local zone of extremely intense precipitation. In the case of the Sierras, the culprit is a Narrow Cold Frontal Rainband (NCFR). I don’t know if these also occur commonly in the Himalaya, but regardless, this paper provides support for the basic idea you are proposing so I think it’s worth citing.
Citation: https://doi.org/10.5194/egusphere-2022-1347-RC1 -
RC2: 'Comment on egusphere-2022-1347', Alexander Densmore, 08 Feb 2023
This is a very nice manuscript that tackles an important and topical issue in Earth surface processes – the mobility and fate of coseismic landslide debris after large earthquakes. I read it with interest and I think it will make a good contribution to the journal. The manuscript is well-written and illustrated overall and won’t require much work to make it suitable for publication. I do have some comments and queries for the authors, which are detailed in an annotated PDF. I won’t repeat all of those here, but will just touch on a couple of the more substantial points.
Most importantly, I don’t think the introduction really motivates the specific study that the authors have carried out. The text to line 42 is pretty much a factual summary of the literature around coseismic and postseismic landsliding and the potential issues associated with that sediment. The authors then pivot to state what they will do, but without having articulated any outstanding issue or problem, it’s hard to see what specific knowledge gap they want to fill. I can guess at this, but it would be helpful to make it explicit. This will also help to address another wider point, which is that – as currently written – the manuscript feels like two loosely connected pieces of work: one on changes in gravel area in two specific parts of the Gorkha rupture area, and another on documentation of a particular flood event in the Melamchi Khola in 2021. It’s not clear, from the introduction or anywhere else, why the authors have chosen to focus on the Melamchi Khola (as opposed to other events in the two catchments – e.g. 2019 and 2020 flows in the Bhote Kosi), or how that links with the first part of the manuscript. Again, I can guess at the link but it would be good to spell it out for the reader. Line 50 introduces the idea of pre-conditioning, but without explaining what that means, and I think that lack of clarity creates an issue later in the manuscript which I'll come to below.
The methodology is straightforward and clearly documented, although I wondered why aggradation was represented by changes in bare-sediment area over a 0.5 km2 tile, rather than by changes in width at a set of closely-spaced stations. There is some double-counting that goes on (which the authors do acknowledge in the SI) with their chosen method. I also wasn’t sure why they then needed to apply a smoothing function; if I understand the method correctly, then aren’t the areas already smoothed over c. 10 cells centred on each raster cell? Why smooth it again? I also would have liked to see some consideration of how much change in area they would expect to see in confined reaches like those along the Bhote Kosi, which yields ‘valley widths’ of <100 m over much of their study section. In a lot of places I’m familiar with, there could be considerable aggradation that would not lead to appreciable widening of the visible water + sediment. It’s perhaps not surprising, therefore, that there is a broad correlation between areas where sediment area has increased and areas where valley width is larger in Figs 6 and 7 (especially for the Bhote Kosi, where panel b looks like a stretched version of panel d), and not much change visible in between. Do the authors have a sense of how much aggradation could be hidden from this approach?
The analysis of landslide connectivity to the channel network feels like a bit of an add-on, and it’s only after the results are shown (lines 375-377) that the rationale is made somewhat more clear – essentially, to see if ‘low connectivity’ can explain the lack of direct evidence for aggradation. The authors don’t really explain what they’re judging that against, though, other than a quick reference to the work by Gen Li and colleagues in Wenchuan. What are they looking for? How low does connectivity have to be to fit with the lack of aggradation that they observe? It’s not clear why they need to use two different inventories or to repeat the analysis that’s already been done by Roback et al. (2018), and Fig 8 isn’t really utilised. I’ve made some suggestions for how this might be better linked with the rest of the manuscript, perhaps by better signposting up front, by taking a similar approach to Li et al. and looking at connectivity as a function of landslide size (since they are making a direct comparison with Wenchuan), and leaving the inventory and buffer size sensitivity analyses out of the main manuscript.
I also struggled a bit to understand the unique contribution that is made by section 6, which largely summarises some reports on the Melamchi event. Part of my confusion might link back to the overall point of investigating that event in detail, and so addressing the motivation may help with this query as well. As I understand it, the main contribution that the authors are trying to make here is that there is evidence in satellite imagery for multiple sediment sources upstream of Bremthang, rather than just the failure of a moraine dam in the Pemdang Khola as argued for by some of the previous work. This is based on Fig 10, which shows bare sediment in a number of tributaries feeding into the Melamchi Khola after the event, and the authors interpret this as evidence of sediment excavation from multiple sources (although there’s some confusing wording on that point). I agree with them, although Fig 10 could be enlarged and annotated to make that easier for the reader to see. I’ve also looked at the evidence for that event, and I struggled to see what they were trying to show in the figure, so a reader could be forgiven for not picking it up. If I’ve missed the point of this section then it would be good to clarify what the unique contribution is that they are trying to make.
Related to this, the authors then argue that there is no evidence for ‘pre-conditioning’ of the event by the Gorkha earthquake. Here I think they need to be more explicit about what they mean by pre-conditioning, because this is a term that is used (or mis-used!) in lots of different ways. I think their argument is that the sediment sources visible in their image don’t match up with coseismic landslides in the Roback et al. inventory, and therefore the Melamchi event wasn’t just caused by remobilisation of coseismic landslide debris. That’s fine, and I agree, but they don’t actually demonstrate that here in any of the figures; it would be useful to overlay the image with the coseismic inventory just to make that point more clearly. But at the same time, I don’t think – on the basis of the evidence that they’ve shown here – that they can rule out what Maharjan et al. (2021) were referring to as pre-conditioning, as in the weakening of hillslopes by shaking in the Gorkha earthquake, which then failed under heavy rainfall in 2021. The assertion that the hillslopes were weakened by the earthquake and then failed is very hard to test; we looked for that kind of effect in a pair of earthquakes in New Zealand (Parker et al. 2016) and it’s really difficult to identify unequivocally. Equally, however, I don’t see any evidence from the authors that that has NOT happened. So I’d suggest that this section, and the conclusions, be reframed in places to link more closely with the evidence that they’ve shown. Similarly, the assertion in line 449 that Melamchi-type events pose a greater risk to populations than coseismic landslides certainly could be true, but it’s not been demonstrated by anything in the manuscript, and I’d also suggest that that is removed or modified.
The figures are generally clear and have good, informative captions. The photos in Fig 2, while definitely compelling, don’t really add much to the focus of the manuscript. I also wasn’t sure why the steepness (long profile or k_sn) and width data needed to be shown in both Fig 6 and Fig 9, so a clearer rationale for why both figures are necessary would be helpful.
Alex Densmore
-
RC3: 'Comment on egusphere-2022-1347', Oliver Francis, 14 Feb 2023
Review of Graf et al 2023; Geomorphological and hydrological controls on sediment export in earthquake-affected catchments in the Nepal Himalaya.
Summary:
In this interesting and thought-provoking study Graf et al track the evolution of gravel storage and export in 2 large catchments in the Nepal Himalaya. Using a unique dataset of gravel area through time and across catchment length, they determine the 2015 Gorkha Earthquake has had a minimal effect on the channel morphology and that only the largest hydro-meteorological events can cause significant alterations of the channel network. I believe this is a significant finding and would make an excellent addition to literature concerning the impact of extreme events on landscape evolution. As a result, this manuscript would make an excellent contribution to the journal subject to several revisions.
The revisions I propose are focused on more closely aligning the 3 main analyses (the analysis of the gravel area, landslide connectivity, and the morphology of the catchments) towards the stated objectives of the manuscript. I will discuss the main points of these revisions here and provide more specific line by line comments afterwards.
Firstly, the manuscript is lacking a clear statement of the research gap the study is trying to fill. The introduction provides a good overview of the study sites and analysis completed by the study does not clearly demonstrate the motivation of the study.
Secondly, the analyses of landslide connectivity and valley morphology feel underdeveloped and lack connection to the main gravel area dataset. In the results and discussion, the authors highlight the ineffectiveness of landslide connectivity in its current definition for predicting changes in the sediment storage in the valley. I feel that this is an important result which requires more analysis and exploration. I suggest that the authors conduct another analysis investigating landslide connectivity along the length of the study catchments. This could potentially be done using the boxes used to analyse the gravel area distribution. This study would allow for the correlation (or lack thereof) between the change in gravel area and landslide connectivity following the earthquake. I also feel a similar additional analysis is required for the valley morphology. Currently it is not clear what the objective of the steepness and width analysis is or how it connects to the gravel area dataset. By explicitly comparing the gravel area dataset with the morphology studies the influence of valley shape on sediment residence time will become more obvious.
Finally, I think the influence of the other hydrological events identified in section 3.2 needs to be discussed as well. Currently the manuscript only focuses on the 2021 Melamchi disaster and as a result the wider picture of the impact of these types of events is lacking. By discussing the complete timeline of events that affected both catchments it will be easier to interpret figures 6&7 and provide an estimate of the magnitude an event needs to be before it affects the catchment.
Line by line changes:
Lines 28&29: Earthquakes are followed by increased landsliding due to 1) remobilisation of coseismic landslide deposits and 2) weakening on the substrate by shaking (damaging the bedrock and breaking cohesive bounds in the soil profile).
Line 33: Should consider the findings of (Jones et al., 2021) as they directly discuss how the earthquake affected landsliding rates in the Nepal Himalaya.
Lines 48 – 52: Currently this statement of intent is not well situated in the literature review discussed in the previous paragraphs. While these paragraphs offer a good coverage of the impact coseismic landslides have on channel evolution they do not discuss sediment export which this statement suggests is the main purpose of the manuscript. I suggest an additional paragraph on post earthquake sediment budgets and the role of meterological and hydrological events is required (these exist for the Chi-Chi (Chen et al., 2011; Dadson et al., 2004; Hovius et al., 2011), Wenchuan (Fan et al., 2019; Francis et al., 2022; F. Zhang et al., 2019; S. Zhang et al., 2016) and various New Zealand (Howarth et al., 2012; Parker et al., 2015; Wilkinson et al., 2022) earthquakes) to root this statement in the research gap that is of interest.
Line 60: (Marc et al., 2016) would be considered a more up to date reference here.
Line 80/figure 1: Kathmandu could be labelled on the figure inset to provide better context.
Lines 81-87: Referencing the figure throughout this paragraph would help the reader to follow the description of the rivers. i.e. “Elevations in the Indrawati catchment range from around 600 m at Dolalghat (D in Fig. 1b) to >5000 m, and the climate spans temperate to polar tundra environments.”
Section 3.2: This section currently reads as a series of events that are not well versed into the context of sediment generation, transport and storage. In particular the observed, or expected (if not recorded) impacts on the sedimentary system, are not discussed.
Line 125: The location of figure 2 should be labelled on figure 1b.
Figure 2 c&d: The viewpoints and directions of a&b should be labelled to help the reader recognise the locations.
Line 152: Figure 3 should be referenced here.
Lines 160-163: It is not clear why this procedure is being used rather than a direct comparison of the shapefiles through time.
Lines 160 - 173: The procedure for choosing the parameter values described in A2 should be moved to here as currently it is not clear how these values were chosen in the main text. I also feel the relationship between ds and dp and the overall area recorded needs further exploration as there is significant overlap between the windows.
Line 178: Does the change in resolution of the satellite images have an impact on the area recorded?
Line 183: Not sure what is meant by “The 75th percentile of the collection of standard deviations”
Line 184: Is this the total uncertainty of the entire profile or per box?
Section 4.2: Connectivity is not discussed in detail prior to this section.
Line 188: Is the location of the landslide defined by its origin/scar location or its deposit?
Line 193: This is the first discussion of why connectivity is being analysed in this manuscript and it does not seem to be linked to the overall stated aim of investigating sediment transportation in the Nepal Himalaya.
Section 4.3: It is not clear how these metrics are used to identify areas of sediment storage.
Section 5.1. Currently it is difficult to follow the section as the reader is required to flick repeatedly between figures 6&7. I think this section should focus first on figure 6 and the Melamchi-Indrawati rivers before discussing the Bhote Koshi. The comparison between the 2 rivers can be done at the end or in the discussion section.
Figure 6&7a are not referenced in the text, perhaps should be relabelled.
Figures 6&7. Currently there is too much information on this figure and the key dates (the earthquake and the hydrological events) are not immediately clear. Perhaps the yearly surveys can be combined to bins based upon the timeline of the events discussed in section 3.2. Another panel which may be useful is a figure showing the absolute cumulative change through time, this would have the potential to draw out the most dynamic areas for further discussion.
Line 237: It is not clear which panel of figure 3 is being referred to here.
Lines 241 – 257: A lot of this section seems better fitted in the discussion rather than the results section. Most of the description of the Melamchi disaster has been discussed in the introduction and is not relevant here.
Line 260: Is the connected area/volume the total area of any landslide identified as connected or just the area within the buffer?
Line 266 - 270: This section would be better placed in the discussion rather than here. The points raised in this section are important, particularly when combined with the lack of a clear signal of gravel mobilisation in the river profile. A further analysis comparing the volume of connected landslides to the change in gravel area along the river length would allow for a conclusion to be drawn about the usefulness of this definition of connectivity in the study area.
Section 5.3: I think this section needs to be more explicitly linked to the gravel area data. I.e. is there a correlation between the channel slope and width with the channel area? This can then be compared with the connectivity of coseismic landslide volume to offer a further opportunity to explore whether the morphology of the channels sets the location of sediment storage or if it is the location of the sediment source.
Lines 305-320: It is not clear from this reporting how much time there is between the cascading events. Is this a continuous event or the result of a series of unfortunate events spanning several days or longer?
Line 344: (Jones et al., 2021) also identifies similar events causing widespread sediment mobilisation.
Section 7.2: This section should also include analysis of the history of events identified in section 3.2 to provide a better understanding of how frequent large events are and their influence on sediment storage.
Lines 392-395: Why can you not calculate the connected volume of landslides in the area affected by the Melamchi disaster? This seems like an important piece of data for this study.
Section 7.3: This section only focuses on how the morphology of the valleys may have influenced the probability of the 2021 disaster occurring. It does not discuss in detail how morphology controls the storage and export of sediment. If this extreme event had not happened would there be any significant differences in the evolution of the gravel area of the 2 catchments?
References discussed in this review:
Chen, H., Lin, G. W., Lu, M. H., Shih, T. Y., Horng, M. J., Wu, S. J., & Chuang, B. (2011). Effects of topography, lithology, rainfall and earthquake on landslide and sediment discharge in mountain catchments of southeastern Taiwan. Geomorphology, 133(3–4), 132–142. https://doi.org/10.1016/j.geomorph.2010.12.031
Dadson, S. J., Hovius, N., Chen, H., Dade, W. B., Lin, J.-C., Hsu, M.-L., Lin, C.-W., Horng, M.-J., Chen, T.-C., Milliman, J., & Stark, C. P. (2004). Earthquake-triggered increase in sediment delivery from an active mountain belt. Geology, 32(8), 733. https://doi.org/10.1130/G20639.1
Fan, X., Scaringi, G., Korup, O., West, A. J., van Westen, C. J., Tanyas, H., Hovius, N., Hales, T. C., Jibson, R. W., Allstadt, K. E., Zhang, L., Evans, S. G., Xu, C., Li, G., Pei, X., Xu, Q., & Huang, R. (2019). Earthquake-Induced Chains of Geologic Hazards: Patterns, Mechanisms, and Impacts. Reviews of Geophysics. https://doi.org/10.1029/2018RG000626
Francis, O., Fan, X., Hales, T., Hobley, D., Xu, Q., & Huang, R. (2022). The Fate of Sediment After a Large Earthquake. Journal of Geophysical Research: Earth Surface, 1–19. https://doi.org/10.1029/2021jf006352
Hovius, N., Meunier, P., Lin, C. W., Chen, H., Chen, Y. G., Dadson, S., Horng, M. J., & Lines, M. (2011). Prolonged seismically induced erosion and the mass balance of a large earthquake. Earth and Planetary Science Letters, 304(3–4), 347–355. https://doi.org/10.1016/j.epsl.2011.02.005
Howarth, J. D., Fitzsimons, S. J., Norris, R. J., & Jacobsen, G. E. (2012). Lake sediments record cycles of sediment flux driven by large earthquakes on the Alpine fault, New Zealand. Geology, 40(12), 1091–1094. https://doi.org/10.1130/G33486.1
Jones, J. N., Boulton, S. J., Stokes, M., Bennett, G. L., & Whitworth, M. R. Z. (2021). 30-year record of Himalaya mass-wasting reveals landscape perturbations by extreme events. Nature Communications, 12(1), 6701. https://doi.org/10.1038/s41467-021-26964-8
Marc, O., Hovius, N., Meunier, P., Gorum, T., & Uchida, T. (2016). A seismologically consistent expression for the total area and volume of earthquake-triggered landsliding. Journal of Geophysical Research: Earth Surface, 121(4), 640–663. https://doi.org/10.1002/2015JF003732
Parker, R. N., Hancox, G. T., Petley, D. N., Massey, C. I., Densmore, A. L., & Rosser, N. J. (2015). Spatial distributions of earthquake-induced landslides and hillslope preconditioning in the northwest South Island, New Zealand. Earth Surface Dynamics, 3(4), 501–525. https://doi.org/10.5194/esurf-3-501-2015
Wilkinson, C., Stahl, T., Jones, K., Fujioka, T., Fink, D., & Norton, K. P. (2022). Delayed 10Be dilution in detrital quartz following extensive coseismic landsliding: A 2016 Kaikōura earthquake case study. Earth and Planetary Science Letters, 581, 117392. https://doi.org/10.1016/j.epsl.2022.117392
Zhang, F., Jin, Z., West, A. J., An, Z., Hilton, R. G., Wang, J., Li, G., Densmore, A. L., Yu, J., Qiang, X., Sun, Y., Li, L., Gou, L., Xu, Y., Xu, X., Liu, X., Pan, Y., & You, C.-F. (2019). Monsoonal control on a delayed response of sedimentation to the 2008 Wenchuan earthquake. Science Advances, 5(6). https://doi.org/10.1126/sciadv.aav7110
Zhang, S., Zhang, L., Lacasse, S., & Nadim, F. (2016). Evolution of Mass Movements near Epicentre of Wenchuan Earthquake, the First Eight Years. Scientific Reports, 6(December 2015), 1–9. https://doi.org/10.1038/srep36154
Citation: https://doi.org/10.5194/egusphere-2022-1347-RC3 - AC1: 'Comment on egusphere-2022-1347', Emma Graf, 07 Apr 2023
Emma L. S. Graf et al.
Emma L. S. Graf et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 488 | 185 | 14 | 687 | 5 | 6 |
- HTML: 488
- PDF: 185
- XML: 14
- Total: 687
- BibTeX: 5
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1