the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Jan 2024
River suspended-sand flux computation with uncertainty estimation, using water samples and high-resolution ADCP measurements
Abstract. Measuring suspended-sand fluxes in rivers remains a scientific challenge due to their high spatial and temporal variability. To capture the vertical and lateral gradients of concentration in the cross section, measurements with point samples are performed. However, the uncertainty related to these measurements is rarely evaluated, as few studies of the major sources of error exist. Therefore, the aim of this study is to develop a method determining the cross sectional sand flux and estimating its uncertainty. This SDC (for Sand Discharge Computing) method combines suspended-sand concentrations from point samples with ADCP (Acoustic Doppler Current Profiler) high-resolution depth and velocity measurements. The MAP (for Multitransect Averaged Profile) method allows to obtain an average of several ADCP transects on a regular grid, including the unmeasured areas. The suspended-sand concentrations are integrated vertically by fitting a theoretical exponential suspended-sand profile to the data using Bayesian modelling. The lateral integration is based on the water depth as a proxy for the local bed shear stress to evaluate the bed concentration and sediment diffusion along the river cross-section to evaluate the bed concentration and sediment diffusion along the river cross-section. The estimation of uncertainty combines ISO standards and semi-empirical methods with a Bayesian approach to estimate the uncertainty due to the vertical integration. The new method is applied to data collected in four rivers under various hydro-sedimentary conditions: the Colorado, Rhône, Isère and Amazon Rivers, with computed flux uncertainties ranging between 18 and 32 %. The relative difference between the suspended-sand flux in 21 cases calculated with the proposed SDC method compared to the ISO 4363 method ranges between -16 and +3 %. This method, which comes with a flexible, open-source code, is the first proposing an applicable uncertainty estimation, that could be adapted to other flux computation methods.
- Preprint
(27112 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2348', Anonymous Referee #1, 10 Mar 2024
The authors introduce a method to determine the sand flux in a river cross-section based on water samples and ADCP measurements, with a systematic uncertainty estimate. The team of authors have a long history in the topic, and build on much of their previous work. I am very positive about the fact that the authors aim to publish their work in peer-reviewed literature, because monitoring methods like this often remain described in grey literature. At the same time, my review below is very critical. Frankly, I believe there is too much uncertainty in the uncertainty estimation, and personally I would not adopt this approach. Despite this, once again, I appreciate the effort and I am convinced the approach deserves to be published after revision.
Major points
- The velocity field is treated completely independent from the suspended sand concentration (SSC) field, while the latter depends on the former. Standard SSC profiles such as the Rouse profile proceed from crude assumptions such as a steady state, and neglect of streamwise and lateral SSC gradients. These assumptions are never truly met, and it would be better to use the measured velocity profiles to derive a theoretical SSC profile, or to use the acoustic backscatter profiles. The authors acknowledge in their discussion that use of ADCP backscatter would be a promising way forward. They argue that the use of ADCP backscatter for SSC monitoring is also prone to uncertainty. This is true for calibration methods with the ultimate aim to translate ADCP backscatter profiles to SSC without direct SSC samples in the profile, but not so much if the aim is to extrapolate. My point is that if one has an ADCP backscatter profile and three samples at different depths across that same profile, and SSC is to be extrapolated over the complete vertical, then there really is no need to rely on a theoretical profile based on assumptions of a steady flow and no horizontal gradients.
- In their uncertainty analysis, the authors assume that errors in Q and C are independent, which I think is unlikely. The rivers subject to monitoring feature transient secondary flow cells, or coherent flow structures, causing that deviations from the dominant flow and SSC fields go hand in hand. Could the authors prove from their data that indeed errors in Q and C are mutually independent?
- The previous comment addresses just one out of a long list of assumptions that had to be made in the uncertainty analysis. How can the quality of the uncertainty analysis be evaluated? In the end, I would be more comfortable with an approach in which an extremely large dataset is used to investigate the loss of accuracy when fewer field samples are available. The authors claim that their method reduces the time and effort required for sampling, compared to ISO standards. This comes at a cost though, and this cost remains unclear to me. Wouldn’t it make more sense to take a dataset that fully complies with the ISO standards, and then show how the new method yields similar results with less data?
- The authors fully focus on sand concentrations, whereas suspended sediment concentration estimates from samples typically include the fine sediment part. The authors tend to neglect the complexity of fine sediment dynamics. On line 25 they write “fine suspended sediments (…) are relatively homogeneous throughout the cross section”. On line 425: “However, as only sand concentrations are considered here, the percentage of sand was set equal to 100%.” That seems crude. The spatiotemporal variation in fine sediment concentrations needs a more nuanced treatment and deserves a place in the uncertainty analysis.
Line by line comments
66 “However, acoustic inversion techniques require many physical samples for calibration, and are affected by acoustic modelling issues (Vergne et al., 2023).” See major point 1.
78-81 please better explain
85 “the large amount of additional samples required for the uncertainty analysis is not realistic to apply” Not in Europe, but this is different in China. Please acknowledge this.
90 Please better motivate the approach of Colby (1964), explaining when it is applicable.
95 you forgot h_l
139-140 How does the QRevInt optimized extrapolation law compare with the approach by Vermeulen et al. (2014)?
Equation 8: where does the 6 in the equation come from? I would appreciate a derivation (just for me to check).
210 How are those uncertainty numbers verified? There is large uncertainty in these uncertainty estimates.
275 Ta facilitate -> To facilitate
340 “In the best case, these measurements should be conducted on every sampling campaign, however, in reality, this is not possible.” It is costly, but possible, and worth doing.
341 “The sampling campaign with additional samples for the uncertainty estimation would take very long, so that the variation in river discharge would become too great.” One could collect samples at a higher frequency.
385 Does the behavior becomes less Rousean with flow strength, or size of the river?
465 Please differentiate between inter- and extrapolation, and calibration to measure stand alone.
480 This is what Vermeulen et al. (2014) do.
495 “differences from the results from the ISO method are up to 15%” In fig. 10 I see differences exceeding 35%. I assume you claim your method is an improvement. How can you quantify/prove this?
506 Compared to the ISO method, this reduces considerably the required sampling time and effort. Yes, but at the cost of a large number of assumptions.
515 Kästner et al. (2018) offers an in-depth analysis of lateral velocity profiles, which may help to put this in a broader context.
Kästner, K., Hoitink, A. J. F., Torfs, P. J. J. F., Vermeulen, B., Ningsih, N. S., & Pramulya, M. (2018). Prerequisites for accurate monitoring of river discharge based on fixed‐location velocity measurements. Water Resources Research, 54(2), 1058-1076.
Vermeulen, B., Sassi, M. G., & Hoitink, A. J. F. (2014). Improved flow velocity estimates from moving‐boat ADCP measurements. Water Resources Research, 50(5), 4186-4196.
Citation: https://doi.org/10.5194/egusphere-2023-2348-RC1 -
RC2: 'Comment on egusphere-2023-2348', Anonymous Referee #2, 15 Apr 2024
This paper proposed a toolbox and a method to use high-resolution ADCP data and point-measured sand concentration data to estimate the suspended sand flux at a river cross-section. To achieve this, the paper proposes a method (SDC method) that employs the MAP method to interpolate the velocity field at the cross-section, then uses the BaM! method to estimate the vertical concentration profiles, and applies a physics-based approach for lateral interpolation. This work also includes uncertainty and error analysis of the SDC method. The toolbox is open-source and has been published online. The data used in this paper are also available for public access. This work performed a thorough analysis and quantified the error propagation in the sand flux measurement process; many of these errors are commonly neglected in other studies, which demonstrates the value and novelty of this work. How to use high-resolution measured data like ADCP data is always a question in scientific communities. The toolbox appears to provide a good solution to such an issue compare to the ISO method. This work is of good quality and worthy of publication. The main issue I encounter while reading this paper is the confusing definitions of different terms. This work focuses on the sediment transport theory proposed by Camenen and Larson (2008). There exists a variety of other classic sediment transport theories, which might have the potential to change the results of these analyses. Please find the detailed comments in the PDF attached.
Data sets
Suspended sediment measurements in the Isère River at Grenoble Campus Jessica Laible, Benoît Camenen, Jérôme Le Coz, Guillaume Dramais, François Lauters, and Gilles Pierrefeu https://doi.org/10.57745/YTCYSX
Data set of solid gaugings in several rivers Jessica Laible, Guillaume Dramais, Benoît Camenen, Jérôme Le Coz, David J. Topping, William Santini, and Gilles Pierrefeu https://doi.org/10.57745/NLFT7Q
Model code and software
Analysis solid gauging Jessica Laible, Blaise Calmel, and Guillaume Dramais https://gitlab.irstea.fr/jessica.laible/analysis-solid-gauging
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 233 | 55 | 14 | 302 | 6 | 6 |
- HTML: 233
- PDF: 55
- XML: 14
- Total: 302
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1