Challenges for mapping cyanotoxin patterns from remote sensing of cyanobacteria

Area of Science:

• Environmental Science
• Remote Sensing
• Water Quality Monitoring

Background:

• Satellite remote sensing of cyanobacterial toxins, particularly microcystins (MC), faces challenges due to the inability to directly detect toxins.
• Reliance on surrogate pigments like chlorophyll-a (Chl-a) and phycocyanin (PC) is complicated by variable relationships with MC and non-standardized measurement methods.
• Existing remote sensing algorithms for pigment detection (analytic, semi-analytic, derivative) have limitations in sensitivity, robustness, and standardization.

Purpose of the Study:

• To address challenges in quantifying cyanobacterial toxins using satellite imagery.
• To propose a dual-model strategy combining in situ pigment measurements with remote sensing algorithms for MC estimation.
• To evaluate the suitability of Chl-a and PC as surrogates for MC and discuss algorithm choices for pigment detection.

Main Methods:

• A dual-model approach: one model estimates MC concentration using in situ Chl-a or PC as surrogates, and a second model uses remote sensing algorithms to estimate pigment concentration.
• Comparison of Chl-a and PC as surrogates, recommending PC for mixed blooms and Chl-a for cyanobacteria-dominated waters.
• Review of remote sensing algorithms (analytic, semi-analytic, derivative) for pigment detection, highlighting the common use and potential of derivative algorithms.

Main Results:

• Phycocyanin (PC) is a better surrogate than chlorophyll-a (Chl-a) in mixed algal blooms, while Chl-a is preferable in cyanobacteria-dominated waters.
• Derivative algorithms are more robust for satellite-based pigment estimation than analytic or semi-analytic methods, despite needing standardization.
• The relationship between MC and surrogate pigments exhibits significant intra- and inter-annual variability, necessitating careful consideration.

Conclusions:

• A dual-model strategy utilizing surrogate pigments and remote sensing algorithms can improve the quantification of cyanobacterial toxins from space.
• Standardization of PC laboratory methods and derivative remote sensing algorithms is crucial for reliable and reproducible MC estimations.
• Addressing the inherent variability in the MC-pigment relationship is essential for developing accurate predictive models for cyanotoxin levels.