the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
17 Nov 2021
17 Nov 2021
Natural and human-induced landslides in a tropical mountainous region: the Rift flank west of Lake Kivu (DR Congo)
- 1Centre de Recherche en Sciences Naturelles, Department of Geophysics, Lwiro, DR Congo
- 2Université catholique de Louvain, Earth and Life Institute – Environmental Sciences, Louvain-La-Neuve, Belgium
- 3Royal Museum for Central Africa, Department of Earth Sciences, Tervuren, Belgium
- 4University of Liège, Department of Geography, Liège, Belgium
- 5F.R.S.-FNRS, Brussels, Belgium
- 6KU Leuven, Department of Earth and Environmental Sciences, Belgium
- 7Vrije Universiteit Brussel, Department of Geography, Brussels, Belgium
- 1Centre de Recherche en Sciences Naturelles, Department of Geophysics, Lwiro, DR Congo
- 2Université catholique de Louvain, Earth and Life Institute – Environmental Sciences, Louvain-La-Neuve, Belgium
- 3Royal Museum for Central Africa, Department of Earth Sciences, Tervuren, Belgium
- 4University of Liège, Department of Geography, Liège, Belgium
- 5F.R.S.-FNRS, Brussels, Belgium
- 6KU Leuven, Department of Earth and Environmental Sciences, Belgium
- 7Vrije Universiteit Brussel, Department of Geography, Brussels, Belgium
Abstract. Tropical mountainous regions are often identified as landslide hotspots with particularly vulnerable populations. Anthropogenic factors are assumed to play a role in the occurrence of landslides in these populated regions, yet the relative importance of these human-induced factors remains poorly documented. In this work, we aim to explore the impact of forest cover dynamics, roads and mining activities on the occurrence of landslides in the Rift flank west of Lake Kivu in the DR Congo. To do so, we compile an inventory of 2730 landslides using © Google Earth imagery, high resolution topographic data, historical aerial photographs from the 1950’s and extensive field surveys. We identify old and recent (post 1950’s) landslides, making a distinction between deep-seated and shallow landslides, road landslides and mining landslides. We find that susceptibility patterns and area distributions are different between old and recent deep-seated landslides, which shows that natural factors contributing to their occurrence were either different or changed over time. Observed shallow landslides are recent processes that all occurred in the past two decades. The analysis of their susceptibility indicates that forest dynamics and the presence of roads play a key role in their regional distribution pattern. Under similar topographic conditions, shallow landslides are more frequent, but of smaller size, in areas where deforestation has occurred since the 1950’s as compared to shallow landslides in forest areas, i.e. in natural environments. We attribute this size reduction to the decrease of regolith cohesion due to forest loss, which allows for a smaller minimum critical area for landsliding. In areas that were already deforested in 1950’s, shallow landslides are less frequent, larger, and occur on less steep slopes. This suggests a combined role between regolith availability and soil management practices that influence erosion and water infiltration. Mining activities increase the odds of landsliding. Mining and road landslides are larger than shallow landslides but smaller than the recent deep-seated instabilities. The susceptibility models calibrated for shallow and deep-seated landslides do not predict them well, highlighting that they are controlled by environmental factors that are not present under natural conditions. Our analysis demonstrates the role of human activities on the occurrence of landslides in the Lake Kivu region. Overall, it highlights the need to consider this context when studying hillslope instability characteristics and distribution patterns in regions under anthropogenic pressure. Our work also highlights the importance of considering the timing of landslides over a multi-decadal period of observation.
Jean-Claude Maki Mateso et al.
Status: open (until 29 Dec 2021)
-
RC1: 'Comment on nhess-2021-336', Anonymous Referee #1, 03 Dec 2021
reply
This study explored the impact of forest cover dynamics, roads and mining activities on the occurrence of landslides in the study area. The results showed that susceptibility patterns and area distributions are different between old and recent deep-seated landslides, and natural factors contributing to their occurrence were either different or changed over time, additionally, the forest dynamics and the presence of roads play a key role in their regional distribution pattern. I enjoyed reviewing your paper and believe it contributes to assess landslide susceptibility/risk for the local government. I have made comments in the hopes that they will be useful to improve the manuscript.
General comments:
- The abstract should be simplified, and it is the embodiment of the core of the article, so you can delete descriptions that are not very important. In addition, I suggest that research methods of article can be added in the abstract.
- In the introduction, you should be added some contents: (i) background information on the hazards of landslides, (ii) the methods of landside susceptibility, and you can analysis the advantages and disadvantages about different methods, (iii) influence factors of landslide should be listed and analyzed based on the previous achievements, especially in the study area or similar area, (iv) you can simplify some contents, such as lines 60 – 75.
- In the section 1.1, you can further analyze the relationship between LULC, population and landslides, because the article results showed that the forest dynamics and the presence of roads play a key role in their regional distribution pattern.
- Authors have chosen 10 predictor variables use for the landslide susceptibility by applying different method, however, the triggering factor may be very difference for the shallow landslide and deep-seated landslide, and the assessment result will be changed, have you ever thought about that? If you considered, and you should be list evaluation factor for different landslide type.
- Fig 7a and 7b presented the shallow landslide susceptibility and old deep-seated landslide susceptibility, author have analyzed the reason of differences, however, the results of fig 7a and 7b were also similar in a certain, you should be further explained.
- The distribution of different landslide was presented in the figure 8, meanwhile, authors should be further analyzed the reason.
- In the section 4.3, authors have said rainfall is the trigger of the shallow landslides that we have identified in this study, and the reason explanation was lacked, however, this part have discussed that anthropogenic factors have an obviously effected on landslide, so you need further analyzed the relationship between shallow landslide and rainfall.
Minor comments:
- Lines 95-100 or 205: you can draw a figure about the change of LULC in the different years.
- Line 110: you can draw a figure about population density or the change of population.
- Lines 155- 160: add the website of different source data.
- Line 175: you can read the relevant references about landslide types, sush as Varnes, 1984; Cruden and Varnes, 1996; Hungr et al., 2014, and it may be better for your research.
- The section 2.2 may be put into section 1.1, you can check it.
- Lines 300-3015: you can simplify.
- The format of Table 3 should be nice.
- Lines 530-545: authors have discussed the difference between Van Den Eeckhaut’s achievements and this study, and this is well. If you can add others’ achievements that is in similar area or nearby the study area, and it may be better.
- I suggest that the previous achievements (similar results or research) should be added, and they can abundant your research in the section 4.1-4.4.
-
RC2: 'Comment on nhess-2021-336', Anonymous Referee #2, 03 Dec 2021
reply
Dear Authors, I have read and carefully evaluatied your manuscript “Natural and human-induced landslides in a tropical mountainous region: the Rift flank west of Lake Kivu (DR Congo)”. I am pleased to report that I found it a relevant, scientifically sound, and well drafted contribution to the journal. It surely deserves publication. However, I have some comments and I recommend to address them to further improve the paper.
Best regards.
---GENERAL COMMENTS---
My main concern is about the structure of the paper. Although it is excellently written, I found it too long and with many repetitions. These shortcomings can maybe be fixed with a reorganization of the paper structure. To be more precise, I found that some concepts are repeated at least twice. The first time in the material and methods section: there, they are outlined with a mid-level of detail, and many questions arise to the reader. Then, the results section repeats everything and add some more details answering most of the answers from the readers. This happens e.g. for landslide, forest, and parts of the analysis. Sometimes, things are repeated once more in the discussion. I think this structure does not help the reader and is not effective. You could try to either reorganize the structure (e.g. moving some preliminary results in the methods section) or shortening the information and comments in the material section to the minimum. In any case, please avoid repetitions and be concise and straight to the point.
The state of the art review could be improved. Basically, the core of your study is a landslide susceptibility mapping (LSM) activity. Therefore, it would be advisable to include a paragraph about LSM. My advice is not to provide a detailed literature review, but you could focus on works that: (i) pertain to the same/nearby areas or areas with similar characteristics; use the same susceptibility model; try deciphering the important role played by LULC dynamics or urbanization. In the literature, the last point is usually accounted for simply by using land cover maps and/or road network as input variables, but you may briefly acknowledge works that tried alternate approaches or specifically addressed this topic, such as:
Luti, T., Segoni, S., Catani, F., Munafò, M., & Casagli, N. (2020). Integration of remotely sensed soil sealing data in landslide susceptibility mapping. Remote Sensing, 12(9), 1486.
Chen, L., Guo, Z., Yin, K., Shrestha, D. P., & Jin, S. (2019). The influence of land use and land cover change on landslide susceptibility: a case study in Zhushan Town, Xuan'en County (Hubei, China). Natural hazards and earth system sciences, 19(10), 2207-2228.
Shu, H., Hürlimann, M., Molowny-Horas, R., González, M., Pinyol, J., Abancó, C., & Ma, J. (2019). Relation between land cover and landslide susceptibility in Val d'Aran, Pyrenees (Spain): Historical aspects, present situation and forward prediction. Science of the total environment, 693, 133557.
Reichenbach, P., Mondini, A. C., & Rossi, M. (2014). The influence of land use change on landslide susceptibility zonation: the Briga catchment test site (Messina, Italy). Environmental management, 54(6), 1372-1384.
To perform the LSM and to assess the variable importance you use logistic regression (LR) and frequency ratio (FR). These methods have a long tradition, but maybe they are a little outdated, as more effective and complex methods are continuously proposed (e.g. in the field of machine learning or deep learning). Don’t you think this is a weakness of your work? I suggest defending the research strategy of using LR and FR on the introduction.
To my understanding, the shape of the area-frequency curves is quite logical. It is normal to have a rollover: it can be interpreted that below that area, the inventory progressively becomes incomplete because smaller landslides are harder to identify (and map), for several reasons. So, I wouldn’t spend so many energies to defend the presence of the rollover in your curves: it is a typical feature, useful to identify the size of the landslides that your model could probably miss.
If I understood correctly, you assess the importance of a variable by running the susceptibility model with only that single variable. I am not very convinced about this approach. The possible interplay among variables is lost. Moreover, a single-variable susceptibility assessment seems of little use. At present, one of the reasons why more sophisticated LSM methods are used is that they also have internal modules that assess the variable importance.
---SPECIFIC REMARKS---
L27 which dynamics? Please, be more specific.
L35 which susceptibility models?
L56-59. It depends also how the human intervention was designed and executed. There is a big difference if you just cut a slope and build a house (or a road), or if the cut is accompanied by some additional works (drainages, concrete walls, …). This should also be highlighted elsewhere in the manuscript when you write about this issue.
Section 1.1 Besides describing the lithology, a short overview of the geological setting could be a nice addendum to this section.
Fig. 1 For the cities, I suggest using a color that better stands out from the colors used for elevation. E.g. black. And you could also add it in the legend. I initially confused cities outside the study area with parts of the study area.
L155-160: From what dates are the images? (This is explained later, but at this point of the manuscript, it is a spontaneous question: see my first general comment).
L174-178. Usually, a landslide is also considered shallow when the ratio depth/width or depth/length is small. I guess this is also your case?
L225. This is not clear to me.
Table 1. the meaning of “reference” in the second column is explained only later. This is confusing.
Table 1. The forest dynamics information is very interesting. In my opinion, it deserves also a figure. Unfortunately, the figure comes only after some pages. This is another example of specific issues comprehended in my first general comment.
Figure 2. The forest cover color hides the information about elevation. Didn’t you already display the elevation in Fig 1? Here, you could just use hillshade and forest cover.
377 “these sources”
Table 4. It seems to me that the bedrock lithology has little influence in determining if a landslide will be shallow or deep seated. Maybe because the lithologies produce similar soils and the actual depth of soils (driven by morphology) is the real control?
478. I like that recent landslides are reasonably well predicted by a model trained with the old ones. This is like a multitemporal validation. It could be worth mentioning it.
493-In an earlier part of the manuscript you mentioned that elevation can be considered a proxy for meteo-climatic characteristics. Why you discard this interpretation here?
593 - Actually, the explanation may be that with this approach you artificially create incompleteness in your inventory. (this interpretation is in accordance with my general comment about frequency-area curves).
604 the influence of vegetation on slope stability is somehow a relevant part of the phenomena you are investigating, but this is never mentioned explicitly. Why didn’t you openly prepare this issue in advance and you don’t mention it explicitly? Forest loss means (I think) reduced root cohesion and reduced evapotranspiration. I would mention it clearly. You could also make reference to some works such as
Masi, E. B., Segoni, S., & Tofani, V. (2021). Root Reinforcement in Slope Stability Models: A Review. Geosciences, 11(5), 212.
Schwarz, M., Preti, F., Giadrossich, F., Lehmann, P., & Or, D. (2010). Quantifying the role of vegetation in slope stability: A case study in Tuscany (Italy). Ecological Engineering, 36(3), 285-291.
Arnone, E., Caracciolo, D., Noto, L. V., Preti, F., & Bras, R. L. (2016). Modeling the hydrological and mechanical effect of roots on shallow landslides. Water Resources Research, 52(11), 8590-8612.
Glade, T. (2003). Landslide occurrence as a response to land use change: a review of evidence from New Zealand. Catena, 51(3-4), 297-314.
608-612. I think there is (also) another explanation: the slope value you are using is an averaged value, while the built environment may be characterized by a locally steeper value. As instance, in a slope cut you could have a small 90° slope, which may not be well captured by the DTM. Even outside artificial environment, a similar situation may be present.
670-675. The stylistic writing of this part is so different from the rest of the paper. Here the sentences are very short and telegraphic. I suggest to better link them.
Jean-Claude Maki Mateso et al.
Jean-Claude Maki Mateso et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 237 | 56 | 3 | 296 | 2 | 2 |
- HTML: 237
- PDF: 56
- XML: 3
- Total: 296
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1