UC Berkeley

Cyanotoxins, the group of toxic chemicals produced by the blue-green algae or cyanobacteria that can proliferate in fresh and salt-water, cause a range of harmful health effects including skin rashes, flu-like symptoms, nausea, diarrhea, tingling and nerve damage, liver damage, tumors, and death ([1, 2]). Although cyanobacteria are one of the oldest organisms on the planet, anthropogenic development (e.g. dams, water diversions, nutrient rich runoff from intensive agriculture, decreased impervious area from urbanization, etc.) has caused many watersheds to lose substantial water volume, suffer tremendous inputs of nutrients and other organic and inorganic pollutants, increase in temperature and overall become more suitable for the proliferation of harmful blooms of cyanobacteria. Cyanobacteria blooms are now increasingly prevalent in freshwaters as eutrophication becomes ever more common with human stressors, and climate change is likely to only further exacerbate this problem. Public health professionals are dependant upon early and dependable information on the presence, concentration, and location of cyanobacteria blooms in order to inform the public and reduce potential exposure. This research project evaluated the efficacy of remote sensing data to provide this kind of surveillance and early detection for characterizing the presence of toxic algae in freshwater systems. It explored the relevance of specific remote sensing techniques to freshwater cyanobacteria bloom identification. The various remote sensing platforms available for this kind of research vary in cost, swath coverage, spatial scale and spectral resolution. For this study, three different remote imagery platforms were compared in terms of their ability to identify surface blooms and to distinguish gradients in cell density or bloom intensity. This exploration of the application of remote sensing used a hyperspectral airborne sensor with high spatial resolution (SpecTIR), a multispectral satellite image also with high spatial resolution (IKONOS), and a lower spatial resolution multispectral satellite image (Landsat). Water sampling data (algal pigment concentrations, turbidity, transparency, and temperature) from known blue-green algae blooms dominated by microcystis aeruginosa on the Iron Gate and Copco Reservoirs on the Klamath River were used to evaluate and classify these different images.

This research successfully used both satellite and airborne remotely sensed data to visualize where the medium or high density sections of the bloom were located, to quantify the intensity of the bloom in terms of area impacted, and to compare the intensity of the bloom at different dates in time. Remote sensing can provide a synoptic overview of the entire system, making it possible to truly assess relative bloom intensity. Furthermore the results indicate that when given the choice, the investment in higher spectral resolution should be chosen over higher spatial resolution as the former appears to provide more benefits in cyanobacteria detection.