Hyperspectral Remote Sensing for Harmful Algal Bloom Detection: Pseudo-nitzschia in the Northeast Pacific

Overview

This study investigates the capacity of hyperspectral remote sensing to distinguish harmful algal blooms caused by Pseudo-nitzschia from benign diatom blooms in the California Current upwelling system. While diatoms are essential contributors to global primary production, carbon sequestation, and fisheries productivity, certain species form harmful algal blooms that threaten marine ecosystems and associated fisheries. Current identification methods rely on expensive and labor-intensive in situ point sampling, limiting the ability to rapidly detect and manage harmful blooms. The research addresses a critical uncertainty regarding whether hyperspectral remote sensing can resolve taxonomic differences within phytoplankton groups at the genus level, particularly for detecting toxic diatom species.

Methods and approach

Four dominant diatom genera from the California Current upwelling system were cultured, including Pseudo-nitzschia, the system's most abundant harmful algae, along with Thalassiosira, Chaetoceros, and Asterionellopsis. Hyperspectral absorption and backscatter measurements were obtained for these cultured taxa and used to model spectral reflectances that would be detectable by satellite and drone-based remote sensing platforms. Vector-based and statistical analyses were employed to quantify differences between the spectral fingerprints of these diatom genera, enabling assessment of discriminability among taxa with similar ecological roles but differing toxicity profiles.

Key Findings

Mean spectral differences of 48% were observed between Thalassiosira, the most dominant diatom in the system, and Pseudo-nitzschia, the most toxic species. Spectral differences of approximately 30% were found between Pseudo-nitzschia and both Chaetoceros and Asterionellopsis, the second and third most abundant diatoms. A unique spectral feature at approximately 560 nanometers was identified as the primary driver enabling successful identification of Pseudo-nitzschia's reflectance fingerprint. This distinct spectral signature demonstrates that Pseudo-nitzschia can be differentiated from non-toxic diatom blooms using hyperspectral remote sensing technology.

Implications

The distinct spectral fingerprint of Pseudo-nitzschia at 560 nanometers establishes the feasibility of using hyperspectral remote sensing platforms to distinguish harmful algal blooms from benign diatom assemblages in the California Current system. This capability addresses significant limitations of current monitoring approaches by enabling scalable, cost-effective detection of toxic blooms across broad spatial scales. The demonstrated ability to resolve taxonomic shifts within a single phytoplankton functional group represents an advancement in remote sensing applications for marine resource management. Implementation of hyperspectral monitoring systems could improve the timeliness and spatial coverage of harmful algal bloom detection, supporting more effective management responses to protect marine ecosystems and fisheries dependent on these resources.