LANDSAT-derived 28-year history of phytoplankton blooms in Western Lake Erie shows changes in peak bloom timing and challenges due to phytoplankton species variability

Jeff C Ho1,2, Anna M Michalak2,3 and Richard P Stumpf4, (1)Stanford University, Stanford, CA, United States, (2)Carnegie Institution for Science, Department of Global Ecology, Stanford, CA, United States, (3)Stanford Earth Sciences, Environmental Earth System Science, Stanford, CA, United States, (4)NOAA Natl Ocean Service, Silver Spring, MD, United States
Abstract:
Information on historical phytoplankton blooms can yield insight that can be used to tackle present-day challenges in inland waters. Thus far, attempts to use historical information to understand long-term changes relevant for present-day management have been stymied by a limited historical record in many freshwater systems. For Lake Erie, we augment the existing record based on MERIS imagery that begins in 2002 by gleaning information on the spatial extent and timing of blooms for 1984-2001 from LANDSAT 5 imagery. Despite limitations stemming from LANDSAT’s long revisit period and regular cloudiness obscuring scenes, we demonstrate that remotely-sensed historical data add relevant context for addressing current blooms by documenting the presence, magnitude, and timing of past blooms. Comparing the remotely-sensed history to the sparse in situ data taken during this period, we find that new information can be gleaned about historical bloom presence, magnitude, and seasonality despite LANDSAT’s limitations. We also find that the remotely-sensed record shows systematic differences in the remotely sensed signal as a function of the colony size of the dominant phytoplankton species present, providing a note of caution for using remote sensing to monitor water quality in systems with multiple phytoplankton regime shifts.