Fingerprinting the spatial sources of fine-grained sediment deposited in the bed of the Mehran River, southern Iran.

Accurate information on the sources of suspended sediment in riverine systems is essential to target mitigation. Accordingly, we applied a generalized likelihood uncertainty estimation (GLUE) framework for quantifying contributions from three sub-basin spatial sediment sources in the Mehran River catchment draining into the Persian Gulf, Hormozgan province, southern Iran. A total of 28 sediment samples were collected from the three sub-basin sources and six from the overall outlet. 43 geochemical elements (e.g., major, trace and rare earth elements) were measured in the samples. Four different combinations of statistical tests comprising: (1) traditional range test (TRT), Kruskal-Wallis (KW) H-test and stepwise discriminant function analysis (DFA) (TRT + KW + DFA); (2) traditional range test using mean values (RTM) and two additional tests (RTM + KW + DFA); (3) TRT + KW + PCA (principle component analysis), and; 4) RTM + KW + PCA, were used to the spatial sediment source discrimination. Tracer bi-plots were used as an additional step to assess the tracers selected in the different final composite signatures for source discrimination. The predictions of spatial source contributions generated by GLUE were assessed using statistical tests and virtual sample mixtures. On this basis, TRT + KW + DFA and RTM + KW + DFA yielded the best source discrimination and the tracers in these composite signatures were shown by the biplots to be broadly conservative during transportation from source to sink. Using these final two composite signatures, the estimated mean contributions for the western, central and eastern sub-basins, respectively, ranged between 10-60% (overall mean contribution 36%), 0.3-16% (overall mean contribution 6%) and 38-77% (overall mean contribution 58%). In comparison, the final tracers selected using TRT + KW + PCA generated respective corresponding contributions of 1-42% (overall mean 20%), 0.5-30% (overall mean 12%) and 55-84% (overall mean 68%) compared with 17-69% (overall mean 41%), 0.2-12% (overall mean 5%) and 29-76% (overall mean 54%) using the final tracers selected by RTM + KW + PCA. Based on the mean absolute fit (MAF; ≥ 95% for all target sediment samples) and goodness-of-fit (GOF; ≥ 99% for all samples), GLUE with the final tracers selected using TRT + KW + PCA performed slightly better than GLUE with the final signatures selected by the three other combinations of statistical tests. Based on the virtual mixture tests, however, predictions provided by GLUE with the final tracers selected using TRT + KW + DFA and RTM + KW + DFA (mean MAE = 11% and mean RMSE = 13%) performed marginally better than GLUE with RTM + KW + PCA (mean MAE = 14% and mean RMSE = 16%) and GLUE with TRT + KW + PCA (mean MAE = 17% and mean RMSE = 19%). The estimated source proportions can help watershed engineers plan the targeting of conservation programmes for soil and water resources.

Comparison of Bayesian, k-Nearest Neighbor and Gaussian process regression methods for quantifying uncertainty of suspended sediment concentration prediction.

Suspended sediment transport in river system is a complex process influenced by many factors that their interactions lead to nonlinear and high scatter of concentration-discharge relationships. This makes the model prediction subject to high uncertainty and providing one value as the model prediction is somehow useless and cannot provide adequate information about the model accuracy and associated uncertainty. Current study compares the efficiency of Bayesian (i.e. Bayesian segmented linear regression (BSLR) and Bayesian linear model (BLR)), Gaussian Process Regression (GPR) and k-Nearest Neighbor (k-NN) in quantifying uncertainty of the suspended sediment concentration prediction in three watersheds namely Arazkoseh, Oghan and Jajrood located in Iran. Three input combinations including, contemporary discharge, slow and quick flow components and contemporary, one and two antecedent days discharge, were used. The BSLR model was able to identify threshold value, furthermore, pre-threshold and post-threshold slopes of BSLR model indicated that for Arazkoseh watershed channel and for Oghan and Jajrood watersheds, upland area are dominate sediment sources. In all three studied cases, given prediction interval width and the percent of enclosed observed data by prediction interval, k-NN model provided more reliable prediction interval. Moreover, separation stream flow into slow and quick flow components lead to improved performance of GPR and k-NN models in the studied watersheds, and the best results for Arazkoseh and Oghan watersheds were obtained when slow and quick flow components were used as the model input.

Mapping the spatial sources of atmospheric dust using GLUE and Monte Carlo simulation.

Atmospheric dust has many negative impacts within different ecosystems and it is therefore beneficial to assemble reliable evidence on the key sources of the dust problem. In this study, for first time, two different source modelling approaches comprising generalized likelihood uncertainty estimation (GLUE) and Monte Carlo simulation were applied to map spatial source contributions to atmospheric dust samples collected in Ahvaz, Khuzestan province, Iran. A total of 264 surficial soil samples were collected from five potential spatial dust sources. Additionally, nine dust samples were collected in February 2015. The performance of both GLUE and Monte Carlo simulation for quantifying uncertainty associated with the source contributions predicted using an un-mixing model were assessed and compared using mean absolute fit (MAF) and goodness-of-fit (GOF) estimators as well as 14 virtual sediment mixtures (VSM). Finally, the erodible fraction (EF) of topsoils and HYSPLIT model were used as further tests for validating the results of the GLUE and Monte Caro simulation. Based on both uncertainty modelling approaches, the loamy sand soil texture was recognized as the main spatial source of the target dust samples. Silty clay soils were estimated to be the least important spatial source of the target dust samples using the two modelling approaches. Both GLUE and Monte Carlo simulation returned MAF and GOF estimates >80%, with Monte Carlo performing slightly better. Based on the virtual mixture tests, the RMSE and MAE of the Monte Carlo simulation (<13.5% and <11%, respectively) was better than for GLUE (<20% and <16.3%, respectively). Spatial source maps generated using both GLUE and Monte Carlo simulation were consistent with the EF map generated using multiple regression (MR) and with routes dust transportation detected by HYSPLIT. Therefore, we recommend that future research into to the sources of atmospheric dust pollution integrates modelling approaches, VSM, EF and HYSPLIT model to quantify and map dust provenance reliably.

Correction to: Monte Carlo fingerprinting of the terrestrial sources of different particle size fractions of coastal sediment deposits using geochemical tracers: some lessons for the user community.

The original publication of this paper contains a mistake. The correct University name of the 3rd affiliation is shown in this paper.

Monte Carlo fingerprinting of the terrestrial sources of different particle size fractions of coastal sediment deposits using geochemical tracers: some lessons for the user community.

A sediment source fingerprinting method, including a Monte Carlo simulation framework, was used to quantify the contributions of terrestrial sources of fine- (< 63 µm) and coarse-grained (63-500 µm) sediments sampled from three categories of coastal sediment deposits in the Jagin catchment, south-east of Jask, Hormozgan province, southern Iran: coastal dunes (CD), terrestrial sand dunes or onshore sediments (TSD), and marine or offshore sediments (MD). Forty-nine geochemical properties were measured in the two size fractions and a three-stage statistical process consisting of a conservation test, the Kruskal-Wallis H test, and stepwise discriminant function analysis (DFA) was applied to select final composite fingerprints for terrestrial source discrimination. Based on the statistical tests, four final fingerprints comprising Be, Ni, K and Cu and seven final fingerprints consisting Cu, Th, Be, Al, La, Mg and Fe were selected for discriminating terrestrial sources of the coastal fine- and coarse-grained sediments, respectively. Two geological spatial sources, including Quaternary (clay flat, high and low level fans and valley terraces) and Palaeocene age deposits, were identified as the main terrestrial sources of the fine-grained sediment sampled from the coastal deposits. A geological spatial source consisting of sandstone with siltstone, mudstone and minor conglomerate (Palaeocene age deposits) was identified as the main terrestrial source for coarse-grained sediment sampled from the coastal deposits.

Fingerprinting sources of reservoir sediment via two modelling approaches.

Reliable quantitative information about sediment sources is a key requirement for river catchment management, especially in settings with high sediment loads. This study explores the potential for using source fingerprinting techniques to establish the relative contribution of three sub-basins to the sediment deposited in a reservoir impounded by an earth dam located at the outlet of the Lavar watershed, in Hormozgan Province, southern Iran. The three sub-basins feeding the reservoir are characterized by complex topography and underlying geology. The source material and target sediment samples were analyzed for 53 potential geochemical tracers, including trace elements and rare earth elements (REEs) and their ratios. Stepwise discriminant function analysis (DFA) was applied to select optimum composite fingerprints from those fingerprint properties passing the range test and we compared two different modelling procedures to estimate the relative contribution of the three sub-basins to the sediment deposited in the reservoir. The first involves a Bayesian mixing model within a Markov Chain Monte Carlo framework (BM) and, the second, an un-mixing model within a Monte Carlo simulation framework (UM). The latter model permits the use of ratio properties, which represents a novel aspect of our study. Particular attention was directed to the uncertainty associated with the source contribution estimates provided by the two models. A goodness of fit estimator was employed to evaluate the results of the UM. Both modelling procedures demonstrated that the southern sub-basin was the main source of the majority of samples we collected from the reservoir. The BM model indicated that the central sub-basin was the dominant source of two samples (S6 and S8). Overall, the results provided by the BM model for the source of seven sediment samples (S1, S2, S3, S4, S5, S7 and S9) are compatible with those provided by the UM model and the central sub-basin was recognized as the most important source supplying sediment in the study area. Both approaches offer potential for using geochemical fingerprinting to quantify spatial sediment source contributions and the uncertainty associated with those estimates.