Displays

Landscapes evolve through surface processes that are often transient in space and time. To understand the underlying geomorphic processes, one must assess how erosion rates vary spatially. This can be done using provenance analysis. Here, we introduce a formal inversion method to derive erosion patterns using detrital zircon age data as fingerprints. Zircons are omnipresent in Earth’s crust and contain information about the time since (re)crystallization in their U/Th-Pb ratio. For each geological unit having undergone a specific tectonic or magmatic history, one can find a unique age-frequency signature. Hence, erosion and sedimentation of grains originating from diverse source areas lead to a mix of the varying age-frequency signatures in sediments found at the outlet of a catchment. Considering that the age signal is not altered during erosion-transportation-deposition events, and given that recent technological advances enable precise dating of large amounts of grains, U/Th-Pb zircon ages provide an appropriate fingerprinting tool. Our inversion approach relies on the least-squares method with a priori information and model covariance to deal with non-uniqueness. We show with synthetic and natural examples that we are able to retrieve erosion rate patterns of a catchment when the age distribution for each geological unit is well known. Furthermore, relying on the nested form of catchments and their subcatchments, we demonstrate that adding samples taken at the outlet of subcatchments improves the estimation of erosion rate patterns. We conclude that the least squares inverse model applied on detrital zircon data has great potential for investigating erosion rates.

How to cite: De Doncker, F., Herman, F., and Fox, M.: Inversion of detrital zircon data to constrain spatially varying erosion rates, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2176, https://doi.org/10.5194/egusphere-egu2020-2176, 2020.

Soil erosion in high mountain ranges plays an important role in redistributing soil organic carbon across landscapes and may influence the global climate on different timescales [1, 2]. Here, we investigate the dynamics of soil organic matter export in the steep mountain belt of the Himalaya by tracing the provenance of soil-derived lipids in riverine sediments from nested catchments with areas ranging from 370 to 57700 km².

Branched glycerol dialkyl glycerol tetraethers (brGDGTs) are a suite of lipids that occur ubiquitously in soils [3, 4]. Their isomer distribution depends on environmental parameters such as the mean annual temperature of the local environment [3]. In this study, we explore the use of brGDGT distributions as a proxy for the altitudinal provenance of soil organic matter in riverine sediments of the Central Himalaya of Nepal. BrGDGT distributions in soils collected along an altitudinal profile, spanning elevations from 200 to 4450 m asl, yield a robust calibration of soil signatures as a function of elevation. This calibration is then used to trace the provenance of soil organic matter exported from their catchments and entrained in suspended sediments of rivers draining the Central Himalaya.

We show that brGDGT compositions of fluvial sediments accurately reflect the mean elevation of the soil-cover in their respective watersheds. The type of land-cover does not seem to have a significant influence on the export of organic matter at a catchment scale. We, therefore, conclude that soil organic matter mobilization in the Himalaya occurs pervasively, and is currently insensitive to anthropogenic perturbations.

[1] Stockmann et al., 2013 – Agriculture, Ecosystems and Environment, 164

[2] France-Lanord & Derry, 1997 – Nature, 390

[3] Weijers et al., 2007 – Geochimica et Cosmochimica Acta, 71

[4] Schouten et al., 2013 – Organic Geochemistry, 54

How to cite: Märki, L., Lupker, M., Gajurel, A., Gies, H., Haghipour, N., Gallen, S., France-Lanord, C., Lavé, J., and Eglinton, T.: Molecular evidence for pervasive riverine export of soil organic matter from the Central Himalaya, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9017, https://doi.org/10.5194/egusphere-egu2020-9017, 2020.

After a wildfire, total or partial removal of vegetal biomass and changes in physicochemical soil properties can lead to an increase in overland flow and sediment yield. Eventual damage must be counteracted urgently identifying erosion hotspots, implementing post-fire management programmes and sampling campaigns to check its effectiveness. Under this context, the sediment source fingerprinting technique is widely used for determining the origin of suspended sediments in catchments and can be useful to evaluate the effectiveness of sediment management programmes. It traditionally relies on the use of physical, biochemical and geochemical properties as tracers. However, measuring these tracers in the laboratory often entails a high economic cost and time consuming. Colour tracers were proven to greatly reduce this cost and measuring time, especially if measurements are done using a common office scanner. Here we propose that colour parameters can be used to investigate SS origin in burned catchments. To this end, we created artificial mixtures of sediment and ash to verify the colour linear additivity and the ash influence on colour parameters. We then used colour parameters for source-ascription of SS samples (n=9) collected during two years after a fire in small Mediterranean catchment (4.8 km²; Mallorca, Spain). Furthermore, reflectance-derived colour parameters were compared with those obtained using a current office scanner. The high correlation between most chromatic indexes (obtained using both methods; p < 0.01) suggested that scanning is a good alternative for measuring soil and sediment colour. A Bayesian tracer mixing model (MixSIAR) was applied to determine the relative contribution of each source. The type of mixing models allowed to appropriately represent natural and sampling uncertainty in tracer data. During the first events, suspended sediment was mainly originated in burned surfaces, whereas its contribution decreased throughout the study period. Tracing results obtained using colour parameters were compared with estimations using ¹³⁷Cs and ²¹⁰Pb_ex, as recognized tracers to discriminate between surface and subsurface sediment sources after wildfires. Estimated source ascriptions with both methods (i.e. reflectance-derived colour parameters and radionuclides) matched in 88% of the samples measured. Colour tracers have been proved to be useful to discriminate between burned and unburned sources, making them suitable for suspended sediment source ascription and monitoring as part of post-fire management strategies.

This work was supported by the research project CGL2017-88200-R “Functional hydrological and sediment connectivity at Mediterranean catchments: global change scenarios –MEDhyCON2” funded by the Spanish Ministry of Science, Innovation and Universities, the Spanish Agency of Research (AEI) and the European Regional Development Funds (ERDF)

How to cite: García-Comendador, J., Martínez-Carreras, N., Fortesa, J., Borràs, A., Calsamiglia, A., and Estrany, J.: Scanner derived colour parameters to determine suspended sediment sources in burned catchments, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-3382, https://doi.org/10.5194/egusphere-egu2020-3382, 2020.

Reliable and detailed information on the primary sources of suspended sediment (SS) and sediment-associated nutrient and contaminant transfers is needed to target mitigation measures for delivering healthy ecosystems and meeting environmental policy objectives. To this end, the SS source fingerprinting approach is proven an effective tool for assembling reliable information on the sources of SS and SS-associated nutrients and contaminants within a catchment. However, SS source estimates at a high temporal resolution are often lacking due to the high workload and costs involved in collecting and analysing SS and soil samples using conventional means. Given this background, here, we propose the use of submersible spectrophotometers that measure absorbance spectra at 2.5 nm intervals in the 200-750 nm range (UV-VIS) in-situ and at high temporal frequency (i.e. minutes) to fingerprint SS sources. We hypothesise that increasing the measurement frequency will eventually help to better characterise changes in sources over time, whilst also giving further insights on how to improve the classical sediment fingerprinting approach, which is currently based on the use of temporally-lumped data. In this research, we first test our approach under fully controlled conditions in a laboratory experiment. To this end, we use a large cylindrical tank (40-L) equipped with a spectrophotometer as well as a LISST sensor (measuring the effective particle size distribution (PSD)). A mechanical stirring device ensures homogeneous conditions in the system and prevents the settling of soil particles (added in solution). The used soil samples originate from different areas within Luxembourg, whereby a selection was made based on differences in tracer properties and colour. The soils were sieved to three different fractions to take account of PSD control on tracer properties. Using the laboratory experiment, we investigated how suspended particle properties affect the absorbance spectra readings. In particular, we looked at the effects of: (i) increasing concentrations of suspended particles, and; (ii) differences in PSD. We then created artificial mixtures composed of two, three and four soil types mixed in different proportions to investigate if the absorbance readings at different wavelengths (i.e., considered as tracers or fingerprints) can be used to un-mix the known proportions of the SS sources. For this, we used the predictions of MixSIAR, a well-established Bayesian tracer un-mixing model. Our preliminary results indicate the promising use of high resolution absorbance data to un-mix artificial sediment mixtures. Ongoing work is testing the approach at larger scales.

How to cite: Lake, N., Martínez-Carreras, N., Shaw, P., and Collins, A.: Testing the ability of submersible spectrophotometers to trace suspended sediment sources at high-temporal frequency , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20316, https://doi.org/10.5194/egusphere-egu2020-20316, 2020.

Aeolian dust coming from arid regions of inland China supplies mineral resources to Japanese Islands. While annual deposition rate is as small as 5~10 g m^-2 yr^-1, total amount of deposition since the last glacial period can form soil layers with several meters depths. We focused on the mineral fertilization effect of aeolian dust on terrestrial ecosystems in Japan. The deposition and incorporation of dust-derived micaceous mineral into soil through time is our special interest because this mineral has an exceptionally higher ability to retain radiocesium than the others, thereby contributing to reduce soil-to-plant transfer of radiocesium after the Fukushima-Daiichi nuclear power Plant accident. However, little study has focused on aeolian dust as a controlling factor of radiocesium dynamics in soils. In this study, peat-core samples collected from the Kamiyoshi (KMY) basin in Kyoto, Japan was used to obtain the aeolian dust that had been deposited on land during the last glacial period. The KMY core with depth from 9.40 to 8.20 m (corresponding to 75 ka to 68 ka) was horizontally divided into 21 portions. The subdivided samples were treated with H₂O₂ to decompose organic matters and then size-fractionated to obtain <20 μm particles, i.e., typical size range of the aeolian dust. The <20 μm particles was used to determine mineral composition by X-ray diffraction and selective dissolution analysis. Furthermore, fine-quartz obtained after the selective dissolution analysis was used for oxygen isotopic analysis to determine δ¹⁸O value. We found that the <20 μm particles occupy more than 90% of total mineral fraction, composed mainly of quartz and micaceous mineral. The fine-quartz in this fraction showed δ¹⁸O values of 16.8±0.3‰, quite similar to those of loess or aeolian dust previously reported. All these properties strongly indicate the origin of the <20 μm particles is the aeolian dust. Interestingly, the aeolian dust flux between 68 ka and 75 ka ranged widely from 1.9 g m^-2 yr^-1 to 15.7 g m^-2 yr^-1 and highly associated with paleoclimate estimated by the pollen record. Namely, the higher aeolian dust flux was observed at colder and drier paleoclimate as indicated by the higher proportion of pinaceous conifer pollen. We conclude that associative analysis of mineral and pollen record in the peat-core can be a promising method to trace the long term deposition pattern of aeolian dust in terrestrial ecosystems in Japan.             

How to cite: Nakao, A., Tanaka, Y., Takahara, H., Tanaka, R., and Yanai, J.: Peat-core analysis for tracing the paleoclimate-related changes in aeolian dust deposition in Japan, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-22577, https://doi.org/10.5194/egusphere-egu2020-22577, 2020.

Fluvial sediments datasets, similarly as other types of a concentration based data, are typical by their relative nature and therefore they need preprocessing or normalization prior to the main statistical analysis. In the geochemical practice, several normalization methods are used, like a simple normalization of the target element concentration with the concentration of the reference (conservative, lithogenic) one, double normalization or concentration conversion to local enrichment factor. As an alternative to these methods, the approach using the principles of compositional data analysis (CoDA) can be considered.  Instead of the standard statistical analytical methods, like ordinary least squares regression, correlation of principal component analysis (PCA), applied on the raw or the target element normalized concentrations, the CoDA methods consider the relative structure of the whole dataset. CoDA together with the use of robust statistical methods, which are down weighting the influence of the outlying observations, have a potential to provide more accurate results. This property is demonstrated and discussed on the base of dataset from mapping the sediments from the Skalka Reservoir in the Ohře River, Czech Republic, and its tributaries. Mainly the performance of the robust versions of regression, correlation and principal components analysis, respecting the CoDA principles, will be presented and the way to them will be explained. 

How to cite: Fačevicová, K., Grygar, T. M., Hron, K., and Elznicová, J.: The use of methods of compositional data analysis for the separation of geochemical signals in fluvial sediments, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21224, https://doi.org/10.5194/egusphere-egu2020-21224, 2020.

Evaluating the usefulness of spatially-distributed soil erosion and sediment delivery models is inherently difficult. Complications stem from the uncertainty in models and measurements of system responses, as well as from the scarcity of commensurable spatial data for model testing. Here, we present an approach for evaluating distributed soil erosion and sediment delivery models, which incorporates sediment source fingerprinting into model testing within a stochastic framework. We applied the Generalized Likelihood Uncertainty Estimation (GLUE) methodology to the Sediment Delivery Distributed (SEDD) model for the Mortes River catchment (~6600 km²) in Southeast Brazil. Sediment concentration measurements were used to estimate long-term sediment loads with a sediment rating curve. Regression uncertainty was propagated with posterior simulations of model coefficients. A Monte Carlo simulation was used to generate SEDD model realizations, which were compared against limits of acceptability of model errors derived from the uncertainty in the curve-estimated sediment loads. The models usefulness for identifying the sediment sources in the catchment was assessed by evaluating behavioral model realizations against sediment fingerprinting source apportionments. Accordingly, we developed a hierarchical tributary sampling design, in which sink sediments were sampled from multiple nodes in the main river channel. The relative contributions of the main sub-catchments in the basin were subsequently estimated by solving the fingerprinting un-mixing model with a Monte Carlo simulation. Results indicated that gauging station measurements of sediment loads were fairly uncertain (average annual specific sediment yields = 0.47 – 11.95 ton ha^-1 yr^-1). This led to 23.4 % of SEDD model realizations being considered behavioral system representations. Spatially-distributed estimates of sediment delivery to water courses were also highly uncertain, as grid-based absolute errors of SEDD results were hundredfold the median of the predictions. A comparison of SEDD outputs and fingerprinting source apportionments revealed an overall agreement between modeled contributions from individual sub-catchments to sediment loads, although some large discrepancies were found in a specific tributary. From a falsificationist perspective, the SEDD model could not be rejected, as many model realizations were behavioral. The partial agreement between fingerprinting and SEDD results provide some conditional corroboration of the models capability to identify the sources of sediments in the catchment, at least with some degree of spatial aggregation. However, the uncertainty in the grid-based outputs might dispute the models usefulness for actually quantifying sediment dynamics under the testing conditions. For management purposes, both SEDD and fingerprinting results indicated that most of the sediments reaching the hydroelectric power plant reservoir located at the outlet of the Mortes River originated from mid and upper catchment tributaries. The convergence of model results therefore evince that reducing reservoir sedimentation rates requires widespread soil conservation efforts throughout the catchment, instead of local/proximal interventions. Ultimately, we have shown how sediment source fingerprinting can be incorporated into the evaluation of spatially-distributed soil erosion and sediment delivery models while considering the uncertainty in both models and observational data.

How to cite: Velloso Gomes Batista, P., Laceby, J. P., Davies, J., Soares de Carvalho, T., Tassinari, D., Naves Silva, M. L., and Quinton, J.: An uncertainty-based framework for incorporating sediment source fingerprinting into spatially-distributed soil erosion model testing, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20211, https://doi.org/10.5194/egusphere-egu2020-20211, 2020.

Compound-specific stable isotopes (CSSI) technique based on the measurement of δ¹³C signatures of organic biomarker compounds such as fatty acids (FAs) has been used since the end of the 2000s to reinforce the knowledge about sediment production and budget in various ecosystems.
The watershed of Petzenkirchen, located 100 km west from Vienna (Austria), was selected to establish the origin of the sediment produced at its outlet using δ¹³C-FAs analysis. The climatic conditions of the area are temperate with continental influences with a mean temperature of 9.5˚ C and yearly precipitation of 823 mm. The dominant soil types are Cambisols and Planosols. Based on existing land-use records, most of the agricultural fields are dominated by a rotation of winter wheat followed by maize cultivation.
Considering the specific geomorphology, the flow of the runoff, the significant interaction of roads, the distance and connection of the potential sources to the outlet, the contributing area of the site has been set to around 50 ha and 7 sources most likely to contribute to the sediment at the outlet were investigated. Using the mixing polygons approach, the δ¹³C of saturated long chain FAs (i.e. C24:0 and C26:0) allowed the best discrimination for establishing the contribution of sources to the sediment collected at the exit of the watershed (i.e. the mixture). The relative contribution to the soil mixture of the different source soils identified has been determined using the Stable Isotope Mixing Models in R (SIMMR) and the specific organic carbon content of each source. The simulated results derived with SIMMR highlights that more than half of the sediment reaching the outlet of the watershed originates from stream bank. 
This Austrian study confirms that the information gained with δ¹³C-FAs analysis could provide unique support for allowing effective agroecosystems management.

How to cite: Mabit, L., Mbaye, M., Toloza, A., Gibbs, M., Swales, A., and Strauss, P.: Use of soil organic biomarkers for tracing the origin of eroded sediment: case study in Petzenkirchen (Austria) , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2014, https://doi.org/10.5194/egusphere-egu2020-2014, 2020.

Environmental DNA (eDNA) is a complex mixture of genetic material extracted from environmental samples like soil, water or sediment in order to obtain reliable information on the past and current biological communities. In recent years, the eDNA technique was successfully applied to sediment accumulated in lakes for providing information on past land use and land cover changes in their drainage areas. Recently, the potential of eDNA for providing detailed information on the plant species found in sediment sources was investigated. These research highlight the powerful potential of this method for improving our ability to detect the vegetal communities causing erosion and sediment delivery. Nevertheless, some fundamental questions remain like for example the DNA memory effect of soils. How long the plant signature can persist in soils? Are we recording the last species cultivated or a mixture of past plants in agricultural areas? These issues are of prime importance for examining the potential of eDNA as a new sediment tracer.

To answer these questions, two contrasted sites located in intensively cultivated environments in France were studied. In the first site, soil samples were collected (n=30) in plots for which the crop rotation history was well documented since 1975. In particular, crops cultivated only once during the rotation were used as potential chronological markers. The impact of agricultural practices on eDNA preservation was also investigated comparing soil signatures under conventional and conservation farming. In the second site, samples were collected (n=40) to compare the abundance of currently observed taxa versus detected taxa in cropland, grassland, woodland and river channel banks.

The results showed that the last cultivated crop was detected in 100% of the samples as the most abundant taxa under conventional farming and 75% under no-tillage. The last cultivated species was the most abundant in 80% of the studied plots. Interestingly, grapevine was detected in 46% of the cultivated plots of the second site, although this plant is no longer cultivated in this catchment. In addition, a large variety of weeds were detected in both sites in addition to the cultivated species. eDNA results provided by the current research illustrate the potential of this method for identifying the recent (<7 years) land cover history of soils which may allow to improve our understanding of sediment mobilization and transfer processes over short timescales.

How to cite: Foucher, A., Evrard, O., Ficetola, G. F., Gielly, L., Poulain, J., Giguet-Covex, C., Laceby, J. P., Salvador-Blanes, S., Cerdan, O., and Poulenard, J.: Potential of environmental DNA for tracing land-use based sediment sources, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8358, https://doi.org/10.5194/egusphere-egu2020-8358, 2020.

So far, it does not exist a set of tracers which fulfill all the characteristics for being an ideal sediment tracer such as, environmentally friendly, inexpensive or easily analysed (Zhang et al., 2001). For this reason, and in order to address some of the unsolved issues of water erosion processes, more research enquiring into the development of these soil and sediment tracers is needed.

Iron oxide-based tracers has been already tested in several water erosion trials with satisfactory results (e.g. Guzmán et al., 2010, 2013, 2015). In 2015, three cascade plots with a different iron oxide (magnetite, hematite and goethite) each were set up in order to evaluate soil redistribution after the rainy season (Obereder et al., 2016). While these authors presented the total iron content of sediments after clorhydric acid extraction, the present study will show only the free iron content of soil and sediments using a different extraction method (CBC, citrate-bicarbonate-ditionite), as this method is more adequate in high iron content soils, as is our case.

The results depict the suitability of the CBD method extracting the three tracers with an average recovery rate of 0.7. The analysis of the iron content of soil and sediment samples indicates a relatively low movement of soil although showing significant statistical differences with background and mixture values. These results are in line with the ones detected by the magnetic susceptibility measurements. Further textural and visible spectrum analysis of the samples will allow to determine the possible selectivity factor and to discriminate qualitative and quantitatively hematite and goethite tracers, respectively.

References:

Guzmán et al. 2010. Catena, 82(2), 126-133.

Guzmán et al. 2013. Soil Science Society of America Journal, 77(2), 350-361.

Guzmán et al. 2015. Journal of hydrology, 524, 227-242.

Obereder et al. 2016. Geophysical Research Abstracts Vol. 18, EGU2016-2455-1. EGU General Assembly 2016.

Zhang et al. 2001. Soil Science Society of America Journal, 65(5), 1508-1515.

How to cite: Guzmán, G., Ramos, A., Barrón, V., and Gómez, J. A.: Optimizing the potential of iron oxide-based tracers, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4949, https://doi.org/10.5194/egusphere-egu2020-4949, 2020.

Quantifying the main sources delivering harmful sediment loads to river systems is required to improve our knowledge of soil erosion processes. Among these potential sources, quantifying the contributions of surface (e.g. cultivated topsoil) and subsurface (e.g. channel bank, gully, landslide) material to sediment transiting river systems is of particular interest. Radiocesium (¹³⁷Cs) that was emitted during the atmospheric bomb tests that took mainly place in the 1960s and nuclear accidents provides an effective tracer to distinguish between topsoil material exposed to the fallout and subsoil sheltered from this fallout. A global synthesis of research articles (n=123) that used radiocesium to fingerprint sediment sources indicated that the largest number of publications (~55% of the total) were found in the United Kingdom, Australia and the United States. On the contrary, very few studies (~9% of the total) were published for catchments located in Africa or South America. Given the low proportion of fallout recorded in regions located between 0-20°N and 0-20°S, the potential of this technique for quantifying sediment source contributions may be limited in this part of the world. A similar conclusion may be drawn for applying this method in agricultural areas exposed to several soil erosion during the last several decades, such as Chinese Loess Plateau and South Africa. Overall, 94% of studies incorporating ¹³⁷Cs as a potential tracer included this property in mixing models. In the future, given the continuous decay of the initial radiocesium fallout that peaked in the 1960s, the access to ultra-low background gamma-ray spectrometry facilities will be increasingly necessary to measure this important sediment tracing property. In addition, more research should be devoted to develop surrogate tracers providing discrimination between surface and subsurface material. Based on this extensive study review, researchers are also recommended to systematically include basic catchment information, details on the soil/sediment sampling design and access to raw data to facilitate the dissemination of this information among the communities of scientists and catchment managers.

How to cite: Evrard, O., Chaboche, P.-A., Ramon, R., Foucher, A., and Laceby, J. P.: A synthesis of worldwide sediment source tracing research including fallout radiocesium (Cs-137) , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8312, https://doi.org/10.5194/egusphere-egu2020-8312, 2020.

Land degradation and fine sediment supply induced by soil erosion processes are exacerbated in intensively cultivated catchments of South America, leading to deleterious consequences for ecosystems and to the disturbance of global biogeochemical cycles. In order to evaluate the sustainability of agricultural practices, a fundamental prerequisite is to quantify soil erosion rates and compare them with the tolerable soil losses. Although its use has been debated in the literature, the fallout radionuclide (137Cs) is one of the few techniques available to reconstruct soil redistribution rates during the Great Acceleration period. Accordingly, it is increasingly used to quantify soil redistribution rates in catchments across the world. Assessment of erosion and deposition rates is commonly based on the comparison of (137Cs) inventories in undisturbed soil profiles (i.e. where neither erosion nor deposition has occurred) and those measured at locations affected by soil redistribution in the landscape. However, in intensive agricultural landscapes, appropriate reference sites can be difficult to find. The objective of this study is to determine and map the initial (137Cs) fallout at the scale of South America, based on measurements made on reference soil sites and spatialized co-variables. This new map will be useful for the scientific community and public authorities to evaluate the sustainability of farming practices, especially in the eastern part of South America where intensive agricultural practices dominate. Furthermore, the methodology presented here could be applied to other regions located in the Southern Hemisphere in order to avoid the systematic sampling of reference sites for conducting soil erosion studies using the (137Cs) technique.

How to cite: Chaboche, P.-A., Evrard, O., and Lefèvre, I.: Spatial distribution of (137)Cs in reference soil sites of South America to reconstruct soil erosion in intensive agricultural landscapes during the Anthropocene, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8286, https://doi.org/10.5194/egusphere-egu2020-8286, 2020.

Skid trail and heavy machinery for forest practice becoming more common for effective forest practices, but these causes soil disturbance in the forest, leading to a sudden increase in the amount of suspended sediment during and post thinning. The discharged sediment can flow into the river and may cause downstream water pollution. To evaluate the effect of thinning on sediment production, sediment fingerprinting techniques can be an effective tool for proper forest practices. In Tochigi prefecture in Japan, in addition to the Cs-137 by global fallout and Pb-210ex, additional FRN, the Fukushima-derived Cs-137 and Cs-134 on March 2011 are available, but few studies are available for combining use of Fukushima-derived radiocesium and fallout Pb-210ex. Therefore, the objective of this study is to determine the transport of the fine sediment in the forest pre- and post- thinning with using fingerprinting techniques.

The study area is Mt. Karasawa, located 180 km southwest of the Fukushima nuclear power plant in Tochigi prefecture. The fallout inventory of Cs-137 and Cs-134 is 8 kBq/m²(Kato et al, 2012). The study site has two catchments which are called K2(17ha) and K3(9ha) respectively and the observation period was from August 2010 to August 2019. In K2, strip thinning was performed with heavy machines from June to October 2011 while randomly thinning without heavy machines was applied for K3 from January to March 2013.Soil samples were collected from the slope surface, skid trail and stream bed, which are the possible sources of suspended sediment. The suspended sediment concentration was measured based on the data of the turbidity censor installed in the stream. The particle size distribution and radionuclide concentration of sediment collected from SS sampler and soil samples are also measured. Hysteresis analysis based on suspended sediment concentration and flow rate and fingerprinting using Fukushima-derived Cs-137, Cs-134, and Pb-210ex was applied to determine the contribution of the slope surface layer and streambed to suspended sediment. By using the difference in the depth distribution of Cs-134 and Cs-137, the production source depth of suspended sediment was estimated.

In the K2 catchment where strip thinning was performed with heavy machinery, suspended sediment concentration during high flow period was rapidly increased up to 2833 mg/L during thinning period and then decreased down to 503 mg/L. On the other hand, in the controlled catchment(K3), no increase in suspended sediment concentration was observed during the same period. By using End- Members Mixing analysis, we found that the contribution of suspended sediment from hillslope increased (from 22% to 50%) more than the stream bed (30%) in the thinning period. Since 2014, the trend has reversed and in 2019, the contribution from the streambed is dominant (50%), and the contribution from slope is decreasing (28%).

How to cite: Kinoshita, M., Onda, Y., Nam, S., Kato, H., Gomi, T., Chen-wei, C., and Taniguchi, K.: Evaluating of the effect of thinning on suspended sediment runoff in a cypress and cedar plantation forest using Fukushima-derived Cs-137, Cs-134 and Pb-210ex, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13119, https://doi.org/10.5194/egusphere-egu2020-13119, 2020.

Assam-Arakan Basin comprises Cenozoic sedimentary successions, located in northeastern India is juxtaposed to both the Himalaya and Indo-Burman Ranges (IBR). The Upper Miocene-Pliocene (Tipam sandstone) and the overlying younger Upper Pliocene-Pleistocene units (Dupi-Tila/Namsang/Dihing) of this foreland basin are fluvial successions. Heavy mineral as detritus provenance indicator has been used as one of the multiproxy records on the fluvial sequences of Assam-Arakan Basin to unravel the drainage system that deposited the same in this basin. Previous workers have advocated that the paleo-Brahmaputra river had initially flowed east of Shillong Plateau before being deflected northwesterly taking the present-day course parallel to the Plateau. However, unequivocal evidence of paleo-Brahmaputra remains enigmatic. The study demonstrates the provenance for the fluvial sedimentary units of the above foreland basin using petrography and heavy mineral distributions. X-ray Diffraction (XRD) and Electron Probe Micro Analyzer (EPMA) analyses were employed to correctly identify the heavy mineral species and support the semi-quantitative analysis of heavy minerals in the basin. The outcome of the study provides new insights towards the paleo-drainage evolution of the river course accountable for the fluvial sedimentation in the Assam-Arakan Basin. Clast petrography and heavy mineral observations indicate the probable source from Lohit- Dibang valley. Initial analysis of detrital zircon U-Pb ages from studied samples reveals major age peaks at around 500 Ma and 1025 Ma with young ages between 16 Ma and ~140 Ma. These samples do not provide ages < 10 Ma, signifying the sediments not derived from Namche Barwa massif, eroded by the Tsangpo-Siang-Brahmaputra river system. It is in contrast to similar sediments in the Siwaliks of NE Himalaya. The data supports our observation that the paleo-Brahmaputra seems not the cause for these deposits, at least during the Pleistocene. If Paleo-Brahmaputra got diverted during this period, it requires scanning the detritus from Tipam units and additional samples from Dupi-Tila/Namsang/Dihing units across the entire Assam-Arakan range to infer source and drainage system for these deposits. We tentatively propose that the Tipam and the younger Dupi-Tila/Namsang/Dihing units in the Assam-Arakan Basin were deposited by drainage flowing from Dibang-Tezu valley, that was initially linked to the Irrawaddy river system. The uplift along Naga thrust caused drainage migration, eventually meeting the present-day Brahmaputra course.

Keywords: Heavy mineral; Detrital zircon U-Pb ages; Paleo-Brahmaputra; Assam-Arakan Basin

How to cite: Rai, P., Borgohain, B., and Mathew, G.: Multiproxy records on the fluvial successions of Assam-Arakan Basin: Implication for paleo-drainage evolution, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-2358, https://doi.org/10.5194/egusphere-egu2020-2358, 2020.

Temporal and spatial sediment dynamics in an East-African Rift Lake (Lake Manyara, Tanzania), and its river inputs, have been evaluated via a combination of sediment tracing and radioactive dating. Changes in sedimentation rate were assessed using radioactive dating of sediment cores in combination with geochemical profile analysis of allogenic and autogenic elements. Geochemical fingerprinting of riverine and lake sediment was integrated within a Bayesian mixing model framework, including spatial factors, to establish which tributary sources were the main contributors to recent lake sedimentation. The novel application of Bayesian source attribution on sediment cores and subsequent integration with sedimentation data permitted the coupling of changes in the rate of lake sedimentation with variations in sediment delivery from the tributaries. These complimentary evidence bases demonstrated that Lake Manyara has experienced an overall upward trajectory in sedimentation rates over the last 120 years with distinct maxima in the 1960s and in 2010. Sedimentation rates are largely a result of a complex interaction between increased upstream sediment delivery following changes in land cover and natural rainfall fluctuations. Modelling results identified two specific tributaries as responsible for elevated sedimentation rates, contributing 58% and 38% of the recently deposited lake sediment. However, the effects of sedimentation were shown to be spatially distinct given the domination of different tributaries in various areas of Lake Manyara. The application of source-tracing techniques constrained sedimentation problems in Lake Manyara to specific tributary sources and established a link between upstream land degradation and downstream ecosystem health. This novel application provides a solid foundation for targeted land and water management strategies to safeguard water security and environmental health in Lake Manyara and has potential application to fill knowledge gaps on sediment dynamics in other East-African Rift Lakes.

How to cite: Wynants, M., Millward, G., Patrick, A., Taylor, A., Munishi, L., Mtei, K., Brendonck, L., Gilvear, D., Boeckx, P., Ndakidemi, P., and Blake, W.: Determining tributary sources of increased sedimentation in East-African Rift Lakes using a combination of sediment tracing and radioactive dating. , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10686, https://doi.org/10.5194/egusphere-egu2020-10686, 2020.

Open-cast mining has strongly increased soil erosion and the subsequent downstream transfer of sediments in river systems. New Caledonia, a French island located in the south-west Pacific Ocean and currently the world's 6th largest nickel producer, is confronted in particular to unprecedented sediment pollution of river systems: hyper-sedimentation. A significant fraction of this sediment is likely originating from tributaries draining nickel mining sites. Nevertheless, the contribution of this sediment source has not been quantified and this estimation is required to guide the implementation of efficient management measurements to mitigate fine sediment supply to New Caledonian river systems and lagoons. To this end, a pilot sediment tracing study has been conducted in one of the first areas exploited for nickel mining, the 397-km² Thio River catchment.

Sediment deposits were collected after two major floods (~10 yr return period): the tropical depression of February 25, 2015 and Cyclone Cook on April 10, 2017. Sources (n=25) were sampled on (i) tributaries draining mines, and (ii) tributaries draining ‘natural’ areas affected by landslides occurring frequently in the region, and sediment (n=19) on (iii) the main stem of the Thio River. In addition, (iv) a 1.60 m long sediment core was collected in the Thio river deltaic floodplain in April 2016. Six sediment tracing techniques were tested based on the following properties: fallout radionuclides, geogenic radionuclides, elemental geochemistry, colorimetric parameters and reflectance spectra.

Several of these methods were identified as relevant to the New Caledonian context and allowed to estimate the contributions of both mining and non-mining sources according to their variations both in space and time. In particular, the sedimentary contributions of mining sources were estimated between 65-68 % for the 2015 flood and 83-88 % for the 2017 flood. The impact of the spatial variability of precipitation was highlighted to explain the variations in the spatial contributions of the sources. The temporal variations in the contributions of the sources deduced from the analysis of the sediment core were interpreted at the light of the mining history in the Thio River catchment (pre-mechanization, mechanization, post-mechanization of mining activity). The contributions of mining sources were again dominant with an average contribution along the sedimentary profile of 74 ± 13 %.

In the future, similar studies should be carried out in other catchments draining mines in New Caledonia and potentially across similar South Pacific and other tropical islands.

How to cite: Sellier, V., Evrard, O., Navratil, O., Laceby, J. P., Allenbach, M., and Lefèvre, I.: Comparing the discrimination power of contrasted sediment tracing techniques to quantify the impact of nickel mining on river and lagoon siltation in New Caledonia, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8401, https://doi.org/10.5194/egusphere-egu2020-8401, 2020.

Erosion processes are accelerated by the presence of unpaved roads in catchments with shallow soils and steep slopes, favoring overland flow and sediment connectivity between hillslopes and the river network. Soil erosion modeling studies conducted at the catchment scale focus primarily on the hydrological behavior of cultivated hillslopes. Few studies address unpaved roads and suitable practices to limit their impact on hydro-sedimentary transfers in a catchment system. This study simulates soil conservation measures on unpaved roads and hillslopes and their effect on the hydrological and erosive dynamics in a small order catchment. The rainfall events were monitored at Lajeado Ferreira’s creek, in Arvorezinha, Southern Brazil (1.2 km²). The catchment is characterized by shallow soils, steep slopes, intense agricultural activity and sediment yields (SY) of around 150 t km^-2 y^-1. Unpaved roads cover about 3% of the catchment surface area and supply 36% of the annual average SY. The Limburg Soil Erosion Model (LISEM) was used to simulate the roads’ impact on soil erosion. Eight rainfall events, monitored from 2014 to 2017, were calibrated. Rainfall ranged from 9 to 97 mm, total runoff volume (Q) varied from 1462 to 60765 m³ and SY from 0.6 to 81 tons. These events represent different precipitation and land use conditions, so that the road’s effects on the hydrological dynamics of this catchment may be investigated. Then, modeling scenarios consisting of three levels of intervention to mitigate sediment supply were tested.  The low intervention (LI) level was based on cost-efficient practices, applied to the road only through the installation of rockfill and energy-deflecting small reservoirs. The medium intervention (MI) included the sowing of grass for gutter protection on the road area and filter strips were installed near the drainage channels. For the high intervention scenario (HI), additional grass strips were installed on hillslopes. Their impact was evaluated by comparing the hydrosedimentological variables Q (m³), peak flow (L s^-1) and SY (tons), modelled at the catchment’s outlet. Rockfill and energy-deflecting small reservoirs were not enough for reducing peak flow (Qp), on average. Indeed, the main proportion of overland flow originates from other landscape components, such as hillslopes. Under the MI and HI scenarios, Qp decreased by 2 and 46%, respectively. The LI and MI scenarios led to an average Q reduction of 12%, compared to 53% under scenario HI. For one event, HI promoted a reduction of 92% of calculated Q, representing 15,693 m³. HI also showed the most positive effects on limiting SY. It becomes evident that hillslope interventions are necessary, as they allow increasing infiltration, reducing both runoff volume and stream power when the flow reaches the roads. For rainfall events of higher magnitude, it was observed that HI was responsible for reducing Qp between 9 and 25%, while during smaller events, this reduction reached 61 to 93%. This indicates the importance of managing roads in order to reduce runoff energy and concentration, but also to take measures on hillslopes to limit overland flow and erosion inputs, as well as to delay peak Q.

How to cite: Dambroz, A., Minella, J., Silva, C., Schlesner, A., Bernardi, F., Barros, C., and Evrard, O.: Monitoring and modeling the impact of soil conservation scenarios to limit overland flow and sediment supply from unpaved roads in a small catchment of Southern Brazil, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10936, https://doi.org/10.5194/egusphere-egu2020-10936, 2020.

No-till is considered an efficient soil and water conservation practice when accompanied by crop species diversification and overland flow control. However, the last twenty years in south Brazil, farmers under no-till abandoned crop diversification and others soil conservation practice. As a consequence, severe erosion process have been observed caused by overland flow. The overland flow has been intensified due to a combination of problems such as soil compaction and poor surface residues. Understanding the formation and propagation of overland flow when the no-till is submitted to different agronomic conditions and practices mechanics as terraces is crucial for the improvement of conservationist agriculture, since many associated problems have caused environmental and economic damage. Unfortunately, there is currently a lack of information about the surface runoff processes occurring under no-till, which is essential to define the best water management practices. The objective of this study was to determine parameters related to surface runoff under different conditions of no-till, specifically regarding the amount of biomass, mechanical scarification and presence of terraces. The study was performed at the hillslope scale in southern Brazil characterized by a deep, clay and weathered soil under no-till. Two zero order catchments (2.4ha) and four hillslopes runoff plots (0.5ha) were monitored during natural rainfall from 2014 to 2018. These catchments and runoff plots are paired and have different soil management to test their respective effects on infiltration and overland flow. In the catchments the presence of terraces was evaluated and in the runoff plot the amount of biomass and mechanical scarification was evaluated. Precipitation was measured with pluviographs and overland flow was carried out automatically on H flumes with sensor pressures. A set of 63 rainfall-runoff events were monitored at the outlet of zero order catchments (2,4ha) and 27 events were observed in the along the runoff plots (0.5ha). From these events, we derived information reflecting the influence of mechanical scarification, of terraces and of biomass. The variables derived for each management mode were: Runoff coefficient (RC), Peak water discharge (PWD), Curve Number (CN), Sediment Yield (SY) and apparent infiltration (AI). In catchments, the main results demonstrate a, influence of terraces on overland flow control, where the reduction reached 50% for RC, 12% for CN, and 60% for PWD, and 70% for SSC. Furthermore, infiltration increased by 11%. In runoff plots, the influence of increasing biomass with scarification the reduction was 16% for RC, 5% for CN, 6.3% for PWD and 6% for SSC. In runoff plots, the influence of increasing biomass without scarification the reduction was 27% for RC, 5.4% for CN, 13% for PWD and 81% for SSC. The results demonstrated that the presence of terraces mitigated the overland flow; however, despite the positive effect of amount of biomass and mechanical scarification, its performance is of less impact. This set of original results will provide the data to quantify the impact of different management conditions and will assist the local managers in the planning of conservationist practices adapted to the conditions observed in Southern Brazil.

How to cite: Schneider, F. J. A., Minella, J. P. G., Londero, A. L., Deuschle, D., Dambroz, A. P. B., Merten, G. H., and Evrard, O.: Overland flow characterization under no-tillage in Southern Brazil, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11993, https://doi.org/10.5194/egusphere-egu2020-11993, 2020.

The conversion of the natural grasslands of the Pampa biome (Southern Brazil) into cropland may lead to an increase in soil erosion rates and sediment delivery to the rivers. Grasslands represent a significant sink of carbon, and according to the literature, 59% of the soil organic carbon (SOC) is lost when pastures are converted into cropland. It makes soils even more vulnerable to water and land degradation. This study aims to evaluate the impact of land use change on the river sediment composition by calculating the sediment contribution of each potential sediment source using organic matter composition, ultra-violet and visible (UV-VIS) spectra derived parameters and fallout radionuclide activities, as potential tracers in a sediment fingerprinting approach. The study site (Ibirapuitã river basin – 5,942 km²) is located in the Pampa biome, Southern Brazil, were sandy and shallow soils predominate, occupied mainly by native grasslands that are gradually being converted to cropfields, especially soybean. Potential sediment sources were sampled, which include croplands (n=36), native grasslands (n=31), unpaved roads (n=31) and subsurface sources (channel banks (n=18) and gullies (n=16)). Samples were taken from the soil surface layer of croplands and grasslands, as well as from the top layer of exposed sites of gullies, channel banks and unpaved roads. Samples were oven dried (50 °C), gently disaggregated and dry sieved to 63 mm to avoid particle size effects prior to further analysis. Suspended sediment samples were collected using time integrated samplers deployed in the bottom of the river, and during rainfall runoff events at the outlet of the catchment. Organic matter parameters (total organic carbon - TOC, total nitrogen - TN, δ¹³C and δ¹⁵N) were measured using a continuous flow isotope ratio mass spectrometry (EA-IRMS). Diffuse reflectance spectra in the UV-VIS wavelengths was measured using a Cary 5000 UV-VIS-NIR spectrophotometer, and 33 parameters were derived from the spectra. Fallout radionuclide (¹³⁷Cs and ²¹⁰Pb_xs) activities were measured by gamma spectrometry using low-background high-purity germanium detectors. Tracers were selected following a three step procedure, including: (i) a conservative range test, (ii) a Kruskal–Wallis H-test, and (iii) a linear discriminant function analysis. The selected tracers were introduced into a mass balance mixing model to estimate the source contributions to in-stream sediment by minimizing the sum of square residuals. TOC and TN show significant differences between cropland and native grassland, while the isotopes δ¹³C and δ¹⁵N, presented a lower discrimination potential. TOC and UV-VIS derived parameters did not present a good discriminant potential when they were tested in isolation, although they increased the source discrimination when combined with organic matter parameters. Fallout radionuclides have a good discriminant potential between surface and subsurface sources, but also between native grasslands and croplands. Croplands are the main sediment source in the Ibirapuitã river catchment (36%), followed by the native grasslands (33%). However, the area occupied by croplands is approximately eight times smaller, demonstrating that erosion processes have been intensified by the conversion of native grasslands into croplands and/or croplands are better connected to the river network.

How to cite: Ramon, R., Evrard, O., Tiecher, T., Huon, S., Bernardi, F., Batista, A. A. M., Tiecher, T. L., Vaury, V., and Minella, J.: Soil degradation due to the conversion of native grassland into cropland in the Pampa biome – (Southern Brazil) and impact on suspended sediment supply to the rivers, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12057, https://doi.org/10.5194/egusphere-egu2020-12057, 2020.

The evolution of landscapes with steep slopes and subject to tropical or alpine climates is mainly controlled by several mechanisms of geomorphic transport such as soil formation, river dynamics, slope stability and mass wasting. The time scale over which the climate influence acts on these mechanisms ranges from seasonal to decennial time span.

On the seasonal time scale, for accessible locations and when manpower is available, direct observations and field survey are the most useful and standard approaches. While very limited studies have been focused on the decennial and century scale due to observational constrains. Here, we present a reproducible workflow based on historical aerial images (up to ~70yrs time span) that includes sensor internal calibration and external orientation, dense matching and elevation reconstruction over two areas of interest that represent pristine examples for tropical and alpine environments: The Rempart Canyon in Reunion Island, and the Arveyron river in the French Alps share a limited accessibility (in time and space) that can be overcome only from archive remote-sensing observations. 

We reach unprecedented resolution: the aero-triangulation falls at sub-metric scale based on ground truth, which is comparable to the initial images spatial sampling. This provides elevation time series with a better resolution to most recent satellite images such as Pleiades. In the case of the Rempart Canyon, we identified and quantified the results of 2 landslides that occurred in 1965 and 2001, and characterized the landslides dynamics. As for the river sediment transdport we emphasize similar tracking for the alpine case, which is controlled by glacial dynamics.  In both cases, we emphasize the strong effect of climatic forcing (precipitation and temperature) over multi-decennial to century time-scales.

How to cite: Lucas, A. and Gayer, E.: River sediment dynamics in high relief landscape from shape from motion, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8145, https://doi.org/10.5194/egusphere-egu2020-8145, 2020.

The compositional data analysis (CoDA), unbiased interpretation of geochemical composition of sediments and soils, must correctly treat several major challenges, well-known to environmental geochemists but still improperly handled. Among them, dilution by autochthonous components, e.g., biogenic carbonates or organic matter, and grain size effects are the most relevant. These effects cannot be eliminated by sample pre-treatment, e.g. by sieving or chemical extraction of diluting components, but they can be handled by implementation of interelement relationships and particle size distribution functions. The challenges of CoDA are principally twofold: geochemical/mineralogical and mathematical/statistical. Geochemical/mineralogical challenge is that complete deciphering of sediment composition would need knowledge of mineral composition (and stoichiometry of individual minerals and their content of major and trace elements) in each grain size fraction. This information can be achieved by analysis of finely divided grain-size fractions of studied sediments, which is enormously demanding, in particular in the silt and clay size fractions; that approach can, however, be found in scientific papers. Mathematical/statistical challenge consists in need to respect nature of compositional data (relative nature, i.e. scale dependence, data closure – content of each component impacts all other components), polymodal data distributions, including the cases when “outliers” (in terms of Gaussian distribution) are a regular part of compositional datasets. Compositional data are best treated using log-ratio methodology and robust algorithms (not based on the least squares fitting methods), which are not familiar to geoscientists.

Most traditional geochemical approaches to CoDA are based on empirical knowledge, models, and assumptions which are hardly proven, e.g. a tracer conservativeness or its grain size invariance, which are not easy tested independently. Most novel mathematical/statistical tools are too abstract and computations too complicate for common geochemists. The bottleneck here is to convert geochemical tasks to formal mathematical/statistical terms and develop novel tools, having chance to become routinely used in future.

We studied composition of 483 sediment samples from floodplain and reservoir impacted by historical pollution from chemical industry in Martktredwitz, Germany. We will demonstrate mathematically/statistically correct routes to (1) distinguishing anthropogenic portion of risk elements in sediments of variable grain size and (2) characterisation of grain size control of sediment composition. Task (1) is best achieved by robust regression with log-ratios of concentrations, which still needs certain a priori geochemical expertise. Task ad (2) is best achieved by the use of a functional analysis of particle size distributions (densities) based on Bayes spaces. To support our recommendations, insufficiency of PCA to solve task (1) will be demonstrated.

How to cite: Matys Grygar, T., Hron, K., Babek, O., Facevicova, K., Talska, R., Hosek, M., Elznicova, J., and Alvarez Vazquez, M. A.: Grain size control of sediment composition variability still not resolved, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8082, https://doi.org/10.5194/egusphere-egu2020-8082, 2020.

Sediment fingerprinting quantifies proportional contributions of catchment sediment sources to downstream sediment mixtures and deposits. While the sediment fingerprinting technique is well established it is still challenged by the coarse spatial resolution of sediment source discrimination which hinders understanding of catchment sediment dynamics and limits its usefulness for land management where finer resolution spatial information is required.

Here, two conventional sediment fingerprinting datasets from New Zealand are used to show how spatial modelling of source loads can improve 1) the usefulness of sediment fingerprinting approaches and 2) visualization of results for end-users by combining source apportionment with terrain information and sediment budget modelling. The two case studies involve unmixing sediment source contributions to 1) overbank deposits within the Oroua River catchment, where six sediment sources (Mountain Range, Mudstone, Hill Subsurface, Hill Surface, Unconsolidated Sandstone, and Channel Bank) across two size fractions (<63 µm and 125–300 µm) were characterized using bulk geochemical and radionuclide tracers and 2) suspended sediment in the Manawatu catchment, where a similar six sources were characterized using bulk geochemical tracers to determine changes in source contributions to hourly suspended sediment samples during a 53hr storm event.

The proportional source contributions for each case study are applied to modelled sediment loads and spatially distributed using terrain information for rock type, slope and channel network extent to produce specific sediment yield maps for both catchments and spatial visualization of sediment generation during the Manawatu catchment storm event. Integrating sediment fingerprinting results with spatial datasets and modelling can enhance interpretation of source apportionment results and improve the utility of this information for end-users focused on the spatial targeting of erosion sources for mitigation to reduce downstream sediment impacts.   

How to cite: Vale, S. and Smith, H.: Improving spatial representation of erosion sources using Sediment Fingerprinting with terrain data and sediment budget modelling, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11959, https://doi.org/10.5194/egusphere-egu2020-11959, 2020.

Within a catchment context, statistical models are widely used to predict the load of pollutants (i.e. fine sediments, chemicals compounds) from potential sources around it, into a main channel (mixture). MixSIAR is a Bayesian mixing model framework that has been used in many environmental studies. As with other models, it presents some assumptions that might be assessed before its use. In this study, a set of artificial mixtures (from real sources) were created using four different catchment sediment sources (Channel Bank; Cultivated land; Pasture and Road Material). The material collected from each source was sieved (<63um) then analysed via WD-XRF for elemental composition. The data collected from this analysis was used to test and assess the main model parameters within an experimental context. A simple range test was performed to initially select tracers that were potentially good predictors. In the end, the model was structured with 43 tracers (elements) using the mean and standard deviation among 10 replicates. Furthermore, it was run using 10^6 iterations (length of the chain) and two different error structures to be compared (residual vs multiplicative error). The results demonstrated the accuracy of the MixSIAR approach to get the real composition in different mixture combinations using a large number of tracers, although in some mixtures a statistically different value was observed where the source term with highest internal variability was present in larger proportion (frequently when %CB >10%). The most precise and reliable results based on convergence were those using the “Residual error” structure, where the value of each mixture was closer to the real and model convergence was achieved more easily. On the other hand, “Multiplicative error” structure led to longer model run times (due to its complexity) and in most cases the model did not converge as for the “Residual error” structure when using the full set of tracers. To mitigate this problem, a posterior tracer selection based on diagnostic information was devised which made it  possible to increase dramatically the convergence of the predicted parameters without a significant difference in the result. Although the “Residual error” structure showed to be the most convenient for further analysis, the technique applied for “Multiplicative error” structure can be used as a potential solution to achieve model convergence while reducing model runtime.

How to cite: Ovando-Fuentealba, L., Taylor, A., Clason, C., Bravo-Linares, C., and Blake, W.: Comparing different error structures in MixSIAR analysis using artificial mixtures from real sediment sources, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-17206, https://doi.org/10.5194/egusphere-egu2020-17206, 2020.