Dataset Information

Sensitivity of fluvial sediment source apportionment to mixing model assumptions: A Bayesian model comparison.

ABSTRACT: Mixing models have become increasingly common tools for apportioning fluvial sediment load to various sediment sources across catchments using a wide variety of Bayesian and frequentist modeling approaches. In this study, we demonstrate how different model setups can impact upon resulting source apportionment estimates in a Bayesian framework via a one-factor-at-a-time (OFAT) sensitivity analysis. We formulate 13 versions of a mixing model, each with different error assumptions and model structural choices, and apply them to sediment geochemistry data from the River Blackwater, Norfolk, UK, to apportion suspended particulate matter (SPM) contributions from three sources (arable topsoils, road verges, and subsurface material) under base flow conditions between August 2012 and August 2013. Whilst all 13 models estimate subsurface sources to be the largest contributor of SPM (median ?76%), comparison of apportionment estimates reveal varying degrees of sensitivity to changing priors, inclusion of covariance terms, incorporation of time-variant distributions, and methods of proportion characterization. We also demonstrate differences in apportionment results between a full and an empirical Bayesian setup, and between a Bayesian and a frequentist optimization approach. This OFAT sensitivity analysis reveals that mixing model structural choices and error assumptions can significantly impact upon sediment source apportionment results, with estimated median contributions in this study varying by up to 21% between model versions. Users of mixing models are therefore strongly advised to carefully consider and justify their choice of model structure prior to conducting sediment source apportionment investigations.An OFAT sensitivity analysis of sediment fingerprinting mixing models is conductedBayesian models display high sensitivity to error assumptions and structural choicesSource apportionment results differ between Bayesian and frequentist approaches.

SUBMITTER: Cooper RJ

PROVIDER: S-EPMC4650832 | biostudies-other | 2014 Nov

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Sensitivity of fluvial sediment source apportionment to mixing model assumptions: A Bayesian model comparison.

Cooper Richard J RJ Krueger Tobias T Hiscock Kevin M KM Rawlins Barry G BG

Water resources research 20141121 11

Mixing models have become increasingly common tools for apportioning fluvial sediment load to various sediment sources across catchments using a wide variety of Bayesian and frequentist modeling approaches. In this study, we demonstrate how different model setups can impact upon resulting source apportionment estimates in a Bayesian framework via a one-factor-at-a-time (OFAT) sensitivity analysis. We formulate 13 versions of a mixing model, each with different error assumptions and model structu ...[more]

PMID: 26612962

Similar Datasets

Project description:Increasing complexity in human-environment interactions at multiple watershed scales presents major challenges to sediment source apportionment data acquisition and analysis. Herein, we present a step-change in the application of Bayesian mixing models: Deconvolutional-MixSIAR (D-MIXSIAR) to underpin sustainable management of soil and sediment. This new mixing model approach allows users to directly account for the 'structural hierarchy' of a river basin in terms of sub-watershed distribution. It works by deconvoluting apportionment data derived for multiple nodes along the stream-river network where sources are stratified by sub-watershed. Source and mixture samples were collected from two watersheds that represented (i) a longitudinal mixed agricultural watershed in the south west of England which had a distinct upper and lower zone related to topography and (ii) a distributed mixed agricultural and forested watershed in the mid-hills of Nepal with two distinct sub-watersheds. In the former, geochemical fingerprints were based upon weathering profiles and anthropogenic soil amendments. In the latter compound-specific stable isotope markers based on soil vegetation cover were applied. Mixing model posterior distributions of proportional sediment source contributions differed when sources were pooled across the watersheds (pooled-MixSIAR) compared to those where source terms were stratified by sub-watershed and the outputs deconvoluted (D-MixSIAR). In the first example, the stratified source data and the deconvolutional approach provided greater distinction between pasture and cultivated topsoil source signatures resulting in a different posterior distribution to non-deconvolutional model (conventional approaches over-estimated the contribution of cultivated land to downstream sediment by 2 to 5 times). In the second example, the deconvolutional model elucidated a large input of sediment delivered from a small tributary resulting in differences in the reported contribution of a discrete mixed forest source. Overall D-MixSIAR model posterior distributions had lower (by ca 25-50%) uncertainty and quicker model run times. In both cases, the structured, deconvoluted output cohered more closely with field observations and local knowledge underpinning the need for closer attention to hierarchy in source and mixture terms in river basin source apportionment. Soil erosion and siltation challenge the energy-food-water-environment nexus. This new tool for source apportionment offers wider application across complex environmental systems affected by natural and human-induced change and the lessons learned are relevant to source apportionment applications in other disciplines.

Project description:Long-chain saturated fatty acid (LCSFA) isotopic composition in tandem with Bayesian isotope mixing models (BIMM) can provide insight into land use-based sediment sources in catchment systems. Apportioning sediment sources robustly, however, requires careful consideration of how additional factors including topography, surface cover and land use practices interact to influence contributions from individual sources. Prior knowledge can be used in BIMM; however, the full capacity of this functionality has not been thoroughly exploited yet in conjunction with sediment fingerprinting. In response, we propose an approach for applying a state-of-the-art BIMM incorporating a sediment connectivity index (SCI) as an informative prior for sediment source apportionment in a highly hydrodynamic catchment in Nepal. A library of LCSFA carbon isotopic composition was constructed for surface soils collected from mixed forest, upland and lowland terraces in the Kharka micro-catchment. ?13C values of LCSFA of time-integrated suspended bulk (<2 mm) sediment were depleted by 4‰ compared to the fine (<0.063 mm) sediment fraction. Conventional source apportionment for fine sediment samples without the SCI informative prior suggested that 66% of the sediment is derived from forest soils followed by lowland (19%) and upland (15%) terraces. Incorporation of the SCI as an informative prior in BIMM, however, modified the original source apportionment estimates to 90%, 9% and 1% respectively. The lower contributions from agricultural terraces are explained by landscape complexity comprising small levelled terraces that reduce hillslope-to-channel sediment connectivity. This study demonstrates the sensitivity of BIMM posterior distributions to incorporation of an informative prior based on a SCI. Inclusion of SCI linked to land use and management can provide a more physically-grounded approach to estimating sediment source contributions from biogeochemical tracers, and critically one which generates results better reflecting what makes good environmental sense in the context of land management and visual evidence of sediment mobilisation and delivery.

Project description:BackgroundMicrosimulation models are mathematical models that simulate event histories for individual members of a population. They are useful for policy decisions because they simulate a large number of individuals from an idealized population, with features that change over time, and the resulting event histories can be summarized to describe key population-level outcomes. Model calibration is the process of incorporating evidence into the model. Calibrated models can be used to make predictions about population trends in disease outcomes and effectiveness of interventions, but calibration can be challenging and computationally expensive.MethodsThis paper develops a technique for sequentially updating models to take full advantage of earlier calibration results, to ultimately speed up the calibration process. A Bayesian approach to calibration is used because it combines different sources of evidence and enables uncertainty quantification which is appealing for decision-making. We develop this method in order to re-calibrate a microsimulation model for the natural history of colorectal cancer to include new targets that better inform the time from initiation of preclinical cancer to presentation with clinical cancer (sojourn time), because model exploration and validation revealed that more information was needed on sojourn time, and that the predicted percentage of patients with cancers detected via colonoscopy screening was too low.ResultsThe sequential approach to calibration was more efficient than recalibrating the model from scratch. Incorporating new information on the percentage of patients with cancers detected upon screening changed the estimated sojourn time parameters significantly, increasing the estimated mean sojourn time for cancers in the colon and rectum, providing results with more validity.ConclusionsA sequential approach to recalibration can be used to efficiently recalibrate a microsimulation model when new information becomes available that requires the original targets to be supplemented with additional targets.

Dataset Information

Sensitivity of fluvial sediment source apportionment to mixing model assumptions: A Bayesian model comparison.

Publications

Sensitivity of fluvial sediment source apportionment to mixing model assumptions: A Bayesian model comparison.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets