Geochemical Fingerprinting - Application of Geochemical Tracers to Fluvial Sediment - page 10

Chapter 2
Geochemical Fingerprinting

Abstract Use of geochemical fingerprinting methods to determine sediment
provenance has progressively increased since the late 1990s, and is now considered
by many investigators as the method of choice to quantify sediment source contribu-
tions at the catchment scale. Application of geochemical fingerprinting largely rests
on four factors: (1) the inability of other techniques (e.g., sediment load monitoring,
photogrammetric methods, and mathematical modeling approaches) to effectively
determine sediment provenance at the required spatial scales, (2) improvements in
analytical methods that allow for the analysis of large numbers of samples for a
wide range of elements, (3) the modification of the utilized statistical methods (e.g.,
inverse/unmixing models) to more effectively account for uncertainty in the modeled
results, and (4) the ability to apply the methods to historic sedimentary deposits retro-
spectively to determine changes in sediment provenance at a site through time. In this
chapter, we focus on the application of geochemical fingerprinting to contemporary
river sediments as well as alluvial deposits that are less than about 150 years old. Our
intent is not simply to summarize the voluminous and growing body of literature on
the subject, but to document the strengths, weaknesses, and uncertainty inherent in
the approach.

· · ·

Keywords Geochemical fingerprinting Sediment provenance Unmixingmodels

Model uncertainty

2.1 Introduction

In order to mitigate the impacts of sediment and sediment-associated contaminants
on aquatic ecosystems, one must first determine from where the sediment is derived.
Once identified, the predominate sediment sources can be targeted using the often
limited financial resources available. While conceptually simple, identifying sedi-
ment sources is not as easy as you might think. For example, the use of site specific
monitoring of sediment loads to determine the source of sediments to a water body
has proven to be a costly, labor intensive, long-term process with a spatial resolution
limited by the number of monitoring sites that can be effectively maintained for sig-
nificant periods of time. An alternative approach is to identify upland areas that are
being eroded and then quantify the rate at which sediment is being removed. Such