Olyaei, Mohammadali2025-03-212025-03-212024-12https://hdl.handle.net/11299/270604University of Minnesota Ph.D. dissertation. December 2024. Major: Civil Engineering. Advisor: ardeshir ebtehaj. 1 computer file (PDF); x, 93 pages.Plastic debris pollution transported by river systems to lakes and oceans has emerged as a significant environmental concern, exerting profound impacts on ecosystems, food webs, and human health. Remote sensing presents a cost-effective approach to bolster interception and removal efforts, particularly in remote regions. However, unlike marine environments, the optical properties of plastic debris in fresh waters remain poorly understood. This study aims to advance knowledge in better understanding of the reflectance signatures of floating plastic, from visible to short-infrared wavelengths. The research has two main sections. First, we applied statistical and machine learning algorithms to detect the occurrence of plastic debris using the available public multispectral and hyperspectral data in the marine environment. Key wavebands containing spectral signatures of floating plastic debris are determined using various band selection algorithms. Second, we provide an open-access hyperspectral reflectance database of floating weathered, and virgin plastic debris found in river systems. Utilizing natural waters from the Mississippi River, the database was assembled using a remote sensing data acquisition system deployed over a hydraulic flume operating under subcritical flow conditions and varying suspended sediment concentrations. The database is archived in Network Common Data Form (NetCDF) and contains (a) reflectance spectra of diverse sizes and types of floating plastic debris within a field of view (FOV); (b) corresponding RGB images and binary labels delineating the spatial distribution and fractional abundance of debris within the FOV; (c) details on flow conditions and sediment concentrations; and (d) characteristics of debris materials. This dataset offers valuable insights for better understanding and pinpointing key spectral signatures indicative of floating plastic debris in freshwater ecosystems.enband selectionhyperspectral remote sensingopticalPlastic debrisThesis or Dissertation