| 05 Mar 2026
Suspended sediment dynamics in an urban, mountain catchment in Nepal
Abstract. Urban mountain catchments are highly vulnerable to erosion and sedimentation due to steep terrain, intense rainfall, and rapid land-use change at the urban fringe. However, event-scale sediment transport remains poorly understood in these regions, particularly in data-scarce areas such as the Himalayas. This study presents the first high-frequency, event-based analysis of suspended sediment transport in the Nakkhu River, a rapidly urbanizing catchment in Kathmandu Valley, Nepal. Using optical backscatter sensors and targeted field sampling during the 2023 monsoon, we analysed how rainfall, antecedent moisture, and human disturbance shaped sediment responses. Sediment transport was highly episodic, with two extreme storms accounting for nearly half of the seasonal suspended sediment load. Analysis of SSC–discharge hysteresis patterns revealed event-specific variability shaped by rainfall intensity, antecedent conditions, and hydrologic connectivity. Peak SSC often lagged peak discharge during low-flow events, suggesting upstream sediment sources. In contrast, high-intensity storms produced rapid sediment delivery, likely from hillslopes, mining zones, and in-channel deposits. Low-frequency (daily) monitoring underestimated sediment loads by approximately 30 % compared to 30-minutes interval data. This study provides the first 30-minute interval event-scale analysis of suspended sediment transport in Kathmandu Valley, revealing how sediment responses vary across the monsoon season in relation to rainfall intensity, discharge dynamics, and antecedent conditions. These insights, including lagged sediment peaks during low-flow events and rapid sediment flushing during intense storms, highlight the value of sub-hourly monitoring for capturing within-event variability and identifying short-lived sediment sources in urban mountain watersheds.
This study analyses the suspended sediment dynamics in a tributary of the Kathmandu valley, which is characterized by monsoonal rainfall patterns and experienced increased human pressure during the last decades. The study is based on a high resolution, sensor-based suspended sediment monitoring program with a focus on the analysis of monsoonal rainfall events. Unfortunately, the monitoring period, which covers six months, is rather short, given the high temporal variability of headwater catchments in monsoonal setting. Despite the short period, the authors were able to monitor a couple of discharge / sediment transport events, which were analyzed in terms of the seasonal(monsoonal) changes, which reflect antecedent rainfall and soil moisture conditions, that effects the runoff efficiency and sediment connectivity in the studied catchments. It provides interesting and valuable information and data in an area, with insufficient knowledge on sediment dynamics and strong human impacts due to mining and other land use activity. In this respect, the manuscript is very valuable for publication in HESS. However, in the current form, the manuscript has some limitations, which need to be addressed before publication (see general and detailed comments below).
Kind regards
Thomas Hoffmann
General comments:
The discussion of the results remains in large parts at a very general level, analyzing the general link between rainfall pattern, hydrology and sediment transport dynamics. To increase the value of this publication, the results should be more strongly discussed in term of the particular situation (physical geography and recent urbanization dynamics) of the study site. For instance, the following questions could guide a more in-depth discussion:
Furthermore, in-depth comparison with other studies on a similar topic is missing. For instance, with respect on event-based characteristics:
Uber, M., Beckers, L.-M., Terweh, S., Helmke, P., Hoffmann, T., 2025. Dynamics of rainfall, discharge, suspended sediment and micropollutant transport in the Moselle River, Central Europe. Environ Sci Europe 37, 204. https://doi.org/10.1186/s12302-025-01243-1
Skålevåg, A., Korup, O., Bronstert, A., 2024. Inferring sediment-discharge event types in an Alpine catchment from sub-daily time series. Hydrol. Earth Syst. Sci. 28, 4771–4796. https://doi.org/10.5194/hess-28-4771-2024
In terms of high-resolution data on sediment dynamics and the inference of suspended sediment loads, the following references should/could be cited:
Moatar, F., Person, G., Meybeck, M., Coynel, A., Etcheber, H., Crouzet, P., 2006. The influence of contrasting suspended particulate matter transport regimes on the bias and precision of flux estimates. Science of The Total Environment 370, 515–531. https://doi.org/10.1016/j.scitotenv.2006.07.029
Slabon, A., Hoffmann, T., 2024. Uncertainties of Annual Suspended Sediment Transport Estimates Driven by Temporal Variability. Water Resources Research 60, e2022WR032628. https://doi.org/10.1029/2022WR032628
The manuscript might profit from restructuring. I suggest to join the results and discussion according the topics covered in this manuscript: i) seasonal changes/ effect of monsoon, ii) event-scale dynamics and importance of high magnitude events, iii) sources and sinks of sediments in relation to urbanization/human pressure. Each of these sections could be structured into results and discussions. This will reduce the redundancy in chapter 3 and 4 and helps to streamline the argumentation.
Detailed comments:
Line 60-63: „during the September 2024 flood-triggered by record-breaking rainfall-which led to landslides“ -> confusing sentence, please rephrase
Line 73: If monsoon dynamics are a central part of the analysis, I suggest to give more details on sediment dynamics driven by monsoon in this area.
Line 79: Start with description of the location and physical geography and continue with LU-change afterwards. If the Nakkhu is a hot spot of urbanization, and if this is a major fact for this study, this should be highlighted in the intro to motivate the study.
Line 83: Indicate that major flow is from S to N
Line 86: Sand mining is typically done within the river itself. You should give more information in the mining activities in the Nakkkhu River catchment, as this seems to be central to understand sediment dynamics. Only in Fig 9. Is becomes obvious that there is hillslope and channel (sand) mining. Pictures in Fig.9 are very helpful to understand the human impact in that catchment and should be shown much earlier. It is also important to locate these areas (Fig.1) and describe the relation to the “synoptic” sampling.
Line 91: Are temperatures given as catchment average, or certain location within the catchment?
Fig. 1: Show elevations in 1b and slope in 1c or used shaded relief to represent hillslope gradients in the catchment. Clearly indicate the location of the monitoring station used in this study in Fig 1c.
Line 105: Indicate the time period for which Q and OBS is measured. The authors write that they measure SSC three time a week, but in Fig 2a, only 14 SSC measurements are shown. Furthermore, sediment station in Fig 1c show those station sampled during the synoptic campaign. However, where is the station located that is used in this study to analyse the events during the six months.
Line 106: 30min intervals of tipping bucket means that you count the number of tips within 30min?
Line 110: How well is the Q~h relationship described by the measurements? This is crucial to evaluate the quality of the results and should be discussed.
Line 116: Specify the fabricate and type of the filter and of the OBS sensor. Indicate wavelength of the light and scattering angle of the sensor. Is the sensor calibrated using formazine standards?
Fig 2: in (a) only 14 measurements are shown in contrast to (b) where you sow much more measurements. Please indicate how much sample are taken overall. In (b) the shaded region represents the 95% confidence interval but many data points (more than 50%) are outside of the interval. I assume that the approach to estimate the CI is not reliable.
Line 153: May to Oct. 2023 -> should be state at beginning of section 2.2.
Line 155: Database of 9 events! Do you need a DB for such a low number of events?
Line 174-184: Testing the effect of the temporal resolution of the sampling intervals on the calculated sediment loads is strongly dependent on the subsampling of the SSC values/measurements. It is not clear from the description of the paragraph how this was done. The authors use for sub-daily sampling the SSC derived from OBS and for larger intervals the SSC measurements. This may introduce bias between both datasets. Furthermore, the authors apply the sediment rating curve approach, however, if high-res data are available not rating curve is needed. A similar study is done by Slabon and Hoffmann. Here subsampling was done using random selection from the high-res OBS data. A similar approach should be used here.
Slabon, A., Hoffmann, T., 2024. Uncertainties of Annual Suspended Sediment Transport Estimates Driven by Temporal Variability. Water Resources Research 60, e2022WR032628. https://doi.org/10.1029/2022WR032628
Line 216: please indicate the location of the site in Fig. 1
Line 228: Use “:” instead of “-“ -> “…were analysed: 1 in pre-monsoon and 8 in monsoon…”
Line 233: indicate number of events in Fig 3 (EX)
Line 252: “consistent evolution of hysteresis” -> what do you expect how this evolved? This should be explained (for instance in the intro).
Fig. 4: Events could be labeled in the Figure, as the number of events is rather low.
Line 290: From Fig A1 it is hard to see for which event downstream rainfall contributions were higher/lower.
Line 297: How did the authors calculate the runoff coefficient? This requires some inter- /extrapolation of the rainfall data from the two met stations to the catchment. Please describe how this was done.
Tab.1: I suggest to add the runoff coefficient of the events to the table.
Fig. 6: 80% of suspended sediment transport in less than 10% of the time, this is comparable to other rivers. See for instance Slabon and Hoffmann (2024, WRR)
Line 338: You refer the first time to the location of the mining site here in this text. The locations of the mines and the sand mining area should be explained in the intro of the study site and be explained in context of the synoptic sampling along the river network (line 216ff).
Line 348: The area of sediment remobilization downstream of the confluence is impacted by disturbances of in-channel sand mining. Please clearly indicate this link here!
Fig. 8: The authors should locate the mining activities in these graphs!
Fig. 9d: I don’t see massive sediment accumulation in this Figure. Deposited material in front of the bridge (right side) seems to be artificial dumping!
Line 367: Add runoff efficiency to Tab. 1
Line 369: remove “or in-stream sediment deposits” -> these can be rapidly mobilized, leading to clock-wise hysteresis, if no threshold effects are involved.
Line 404: “…early upstream-driven peaks and later downstream-driven peaks…” -> is this a general pattern? If yes, could this be detected in many events connected with a typical hysteresis pattern and is this pattern connected to certain precipitation events / weather pattern?
Line 409-410: This is a very general statement, could this be made more specific with respect to the study site? What are general sources and where are these located? Which part of the catchment is more prone to erosion and how is this effected to the human impact in the area?
Line 419-421: Again, very general -> be more specific to the study area. What are main driving factors affecting future dynamics in the study site?
Line 423: The authors highlight the important of intense storm events for the sediment transfer and export from the studied catchment. A more detailed analysis of return intervals of relevant flood magnitudes should be conducted. How often do sediment transport events occur during a year?
Line 438: Does the 420 t/km2 refer to the six month or the annual load (i.e. 420 t/km2/a)? Please clarify. Comparison of the load from the six months with other study site requires consideration of the very strong interannual variability of these systems, depending on rain conditions of each year. Was the monitoring period a rather dry, normal or wet period?
Line 476-484: Again, this is very general, and it would be nice to more strongly work with the results from the study. Where are the major mining areas? Is there location related to increased SSC? Etc…
Line 487: Awkward sentence (haphazard landad????)
Line 490ff: Finally, the authors give more details on the questions I raised above. I am wondering, if these info’s can be given earlier in the MS.
Line 496-498: While the grain size decreases the overall SSC strongly increases down to the outlet, suggesting remobilization of fines in the same reach. This is in contrast to the “…diminished transport capacity…” stated by the authors in relation to the fine sediments (line 501-502). Please rephrase!
Line 511: “rapidly urbanizing mountain catchment” -> can you give numbers of this development. This would be helpful to extend the discussion toward expected changes of sediment transport dynamics.
Line 516: “limited sediment availability in downstream reaches” -> this again is in contrast with the strong increase of SSC from the confluence to the outlet (Fig.8).
Line 527: sediment hazards -> please specify the most important hazards in the study area