Characterization of Hyperspectral Inherent and Apparent Optical Properties of Harmful Algal Blooms in Lake Erie

Michael Sayers1,2, Karl Bosse2, Steven A Ruberg3, Gary Fahnenstiel4, Robert A Shuchman4 and Reid Sawtell5, (1)Michigan Technological University, Michigan Tech Research Institute, Houghton, MI, United States, (2)Michigan Tech Research Institute, Ann Arbor, MI, United States, (3)NOAA, Great Lakes Environmental Research Laboratory, Ann Arbor, MI, United States, (4)Michigan Tech Research Inst, Ann Arbor, MI, United States, (5)Michigan Tech Research Inst, Ann Arbor, United States
Ocean color remote sensing has been used to estimate and even predict harmful algal bloom (HABs) occurrences in both freshwater and marine environments. In almost all cases, the remote sensing products used to estimate HABs are based on phytoplankton biomass measurements (i.e., chlorophyll-a concentration) or spectral features in the remote sensing reflectance that are indicators of biomass from multi-spectral algorithms using from 2 to 4 bands. This is a limitation imposed by the available band sets in ocean color sensors over the past few decades, although the addition of channels in the red and near infrared (NIR) (e.g., MERIS) has shown to improve detection of high biomass conditions often associated with freshwater cyanobacteria HABs (CHABs). While these satellite products have proved useful, they are not without issues. For example, all products based on phytoplankton biomass and biomass indices have the potential to provide false positives as they measure total phytoplankton abundance and are not specific for CHAB species.

Hyperspectral observation is well suited to monitoring CHABs in freshwaters, as there is more information available to differentiate phytoplankton groups and even large species assemblages based on unique spectral features. This is particularly true for freshwater cyanobacteria, which have distinctive optical properties that impact different parts of the spectrum, and have a capability to alter their vertical position through buoyancy control, which strongly impacts the observed remote sensing reflectance.

The purpose of this study is to characterize the natural inter- and intra-annual variability of the inherent and apparent optical properties (IOP/AOP) over the HAB growth cycle in Lake Erie to support the development of hyperspectral bio-optical retrieval algorithms. Weekly IOP/AOP measurements were made at standardized locations in Lake Erie from May-October for the 2015-2019 period. Additionally, high temporal frequency (1-5 minute sampling interval) remote sensing reflectance measurements were made from a fixed location hyperspectral radiometer system from July-October in 2018 and 2019. Results suggest algal assemblages differ significantly within and between years indicating the need for dynamic hyperspectral bio-optical algorithm parametrization.