Seth J. K. Mason | Geology Reports

USGS2021

Distribution, abundance and spatial variability of microplastic pollution on the surface of Lake Superior

In 2014, 94 paired neuston net samples (0.5 mm mesh) were collected from the surface waters of Lake Superior. These samples comprise the most comprehensive surface water survey for microplastics of any of the Great Lakes to date, and the first to employ double net trawls. Microplastic abundance estimates showed wide variability, ranging between 4000 to more than 100,000 particles/km 2 with most locations having abundances between 20,000 to 50,000 particles/km 2 . The average abundance in Lake Superior was ~30,000 particles/km 2 which was similar to previous estimates within this Laurentian Great Lake and suggests a total count of more than 2.4 billion (1.7 to 3.3 billion, 95% confidence interval) particles across the lake’s surface. Distributions of plastic particles, characterized by size fraction and type, differed between nearshore and offshore samples, and between samples collected in the eastern versus western portion of the lake. Most of the particles found were fibers (67%), and most (62%) were contained in the smallest classified size fraction (0.50–1 mm). The most common type of polymer found was polyethylene (51%), followed by polypropylene (19%). This is consistent with global plastics production and results obtained from other studies. No statistically significant difference was detected between the paired net samples, indicating that single net sampling should produce a representative estimate of microplastic particle abundance and distribution within a body of water.

Lake Superior↗

USGS2017

A software tool to assess uncertainty in transient-storage model parameters using Monte Carlo simulations

Researchers and practitioners alike often need to understand and characterize how water and solutes move through a stream in terms of the relative importance of in-stream and near-stream storage and transport processes. In-channel and subsurface storage processes are highly variable in space and time and difficult to measure. Storage estimates are commonly obtained using transient-storage models (TSMs) of the experimentally obtained solute-tracer test data. The TSM equations represent key transport and storage processes with a suite of numerical parameters. Parameter values are estimated via inverse modeling, in which parameter values are iteratively changed until model simulations closely match observed solute-tracer data. Several investigators have shown that TSM parameter estimates can be highly uncertain. When this is the case, parameter values cannot be used reliably to interpret stream-reach functioning. However, authors of most TSM studies do not evaluate or report parameter certainty. Here, we present a software tool linked to the One-dimensional Transport with Inflow and Storage (OTIS) model that enables researchers to conduct uncertainty analyses via Monte-Carlo parameter sampling and to visualize uncertainty and sensitivity results. We demonstrate application of our tool to 2 case studies and compare our results to output obtained from more traditional implementation of the OTIS model. We conclude by suggesting best practices for transient-storage modeling and recommend that future applications of TSMs include assessments of parameter certainty to support comparisons and more reliable interpretations of transport processes.

Freshwater Science↗