Investigating the importance of recreational roads as a sediment source in a mountainous catchment using a fingerprinting procedure with different multivariate statistical techniques and a Bayesian un-mixing model

A - Papers appearing in refereed journals

Nosrati, K. and Collins, A. L. 2019. Investigating the importance of recreational roads as a sediment source in a mountainous catchment using a fingerprinting procedure with different multivariate statistical techniques and a Bayesian un-mixing model. Journal of Hydrology. 569 (February), pp. 506-518.

AuthorsNosrati, K. and Collins, A. L.
Abstract

Road construction associated with land development generally increases erosion and sediment yields. Construction of unpaved roads has the potential to alter hydro-sedimentological behavior and catchment sediment source dynamics and, to date, this has largely been investigated in forested environments. The objective of this study, therefore, was to assess the relative importance of unpaved recreational roads as a sediment source alongside hillslope surface soils and stream channel banks in a non-forested mountainous catchment in northern Tehran, Iran, using a fingerprinting procedure. Eleven geochemical tracers were measured on 27 samples collected to characterise the sediment sources and five suspended sediment samples collected at the study catchment outlet. The statistical analysis employed to select three different composite fingerprints for discriminating the sediment sources comprised: (1) the Kruskal–Wallis H test (KW-H), (2) a combination of KW-H and discriminant function analysis (DFA), and (3) a combination of KW-H and principal components & classification analysis (PCCA). A Bayesian un-mixing model was used to ascribe sediment source contributions using the three composite fingerprints. Using the KW-H composite signature, the respective relative contributions (with uncertainty ranges) from recreational roads, hillslope surface soils and channel banks were estimated as 64.5% (57.7–73.1), 1.1% (0.1–4.9), and 33.9% 24.9–41.0), compared to 55.3% (45.5–68.5), 1.9% (0.1–7.9)and 42.1% (27.8–52.4) using a composite signature selected using a combination of KW-H and DFA, or 82.0% (69.7–93.8), 8.2% (0.7–22.7) and 7.3% (0.7–21.0) using a fingerprint selected using KW-H and PCCA. The root mean square
difference between the apportionment results using the fingerprints identified on the basis of the three different
statistical approaches ranged from 5.5% to 25.7%, highlighting the sensitivity of source estimates to the tracers used. Regardless, the different composite signatures all suggested that unpaved recreational roads were the dominant source of the suspended sediment samples, underscoring the need for mitigation measures targeting these anthropogenic features of the catchment system, including closure to permit re-vegetation, surface ripping and/or mulching to improve infiltration or gravel re-surfacing to reduce exposure of bare surfaces to sediment
mobilisation.

KeywordsGeochemical tracers; Modified MixSIR Bayesian model; Sediment source tracing; Statistical techniques
Year of Publication2019
JournalJournal of Hydrology
Journal citation569 (February), pp. 506-518
Digital Object Identifier (DOI)doi:10.1016/j.jhydrol.2018.12.019
Open accessPublished as ‘gold’ (paid) open access
Funder project or codeS2N - Soil to Nutrition - Work package 3 (WP3) - Sustainable intensification - optimisation at multiple scales
FunderBiotechnology and Biological Sciences Research Council
Publisher's versionNosrati and Collins 2019.pdf
Output statusPublished
Publication dates
Online22 Dec 2018
Copyright licenseCC BY
PublisherElsevier Science Bv
ISSN0022-1694

Permalink - https://repository.rothamsted.ac.uk/item/8w87z/investigating-the-importance-of-recreational-roads-as-a-sediment-source-in-a-mountainous-catchment-using-a-fingerprinting-procedure-with-different-multivariate-statistical-techniques-and-a-bayesian

Total views: 29

Total downloads: 14

Views this month: 2

Downloads this month: 2