GSA Connects 2021 in Portland, Oregon

Paper No. 84-9
Presentation Time: 9:00 AM-1:00 PM


GORMAN, Sarah G.1, EDWARDS, Jack1, KNUTTE, Alex1, GROSSHANS, Cole1, KLINKHAMMER, Jenna1, ZABROSKI, Jordan1, BARRETT, Braxton1 and KREKELER, Mark2, (1)Geology & Environmental Earth Science, Miami University, 250 s Patterson, Oxford, OH 45056, (2)Department of Geology and Environmental Earth Science, Miami University - Hamilton, Hamilton, OH 45011

Lake Michigan and its coastline is a critical resource for northwestern Indiana. One threat to this resource is plastic pollution. The global impact that plastic pollution has stems from the fact that it takes thousands of years for plastic to decay. Plastic eventually breaks down into smaller and smaller pieces, making it harder to identify and easier for wildlife to ingest. Monitoring plastic pollution poses many problems including scale, identification of aged plastics, and capturing variation over time. Hyperspectral remote sensing is an environmental monitoring approach where each pixel in a captured aerial or satellite image contains a reflective spectra of the material in the scene at that position and thus can enable identification. Libraries of spectra must be carefully built to enable successful hyperspectral identification. Our group assembled a collection of reflective spectra on numerous plastic items such as blue plastic bag, a red Solo cup, a Coleman blue cooler, plastic lids, and a styrofoam cooler. For comparison we also assembled spectral data of Ottawa sand and Lake Michigan Dune sand, along with appropriate meta data. All data was collected using an ASD FieldSpec 4 High Resolution spectroradiometer with a spectral range of 350 to 2500 nm and a contact probe attachment. Numerous features in the near infrared and shortwave infrared are useful for identifying plastics observed. Styrofoam cooler spectra have the highest levels of absolute reflectance and are very distinct while the red solo cups had the lowest reflectance and a re far less distinct. Sands investigated have expected features that are consistent with mineralogy. Results of our study indicate that hyperspectral imaging may be useful for identifying numerous plastic types in southern Lake Michigan and potentially may be useful to observe changes over time in the amounts and locations of plastic in near shore environments.