Phytoplankton optical fingerprint libraries for development of phytoplankton ocean color satellite products.

Phytoplankton respond to physical and hydrographic forcing on time and space scales up to and including those relevant to climate change. Quantifying changes in phytoplankton communities over these scales is essential for predicting ocean food resources, occurrences of harmful algal blooms, and carbon and other elemental cycles, among other predictions. However, one of the best tools for quantifying phytoplankton communities across relevant time and space scales, ocean color sensors, is constrained by its own spectral capabilities and availability of adequately vetted and relevant optical models. To address this later shortcoming, greater than fifty strains of phytoplankton, from a range of taxonomic lineages, geographic locations, and time in culture, alone and in mixtures, were grown to exponential and/or stationary phase for determination of hyperspectral UV-VIS absorption coefficients, multi-angle and multi-spectral backscatter coefficients, volume scattering functions, particle size distributions, pigment content, and fluorescence. The aim of this publication is to share these measurements to expedite their utilization in the development of new optical models for the next generation of ocean color satellites.

Assessing potential of the Geostationary Littoral Imaging and Monitoring Radiometer (GLIMR) for water quality monitoring across the coastal United States.

The Geostationary Littoral Imaging and Monitoring Radiometer (GLIMR) will provide unique high temporal frequency observations of the United States coastal waters to quantify processes that vary on short temporal and spatial scales. The frequency and coverage of observations from geostationary orbit will improve quantification and reduce uncertainty in tracking water quality events such as harmful algal blooms and oil spills. This study looks at the potential for GLIMR to complement existing satellite platforms from its unique geostationary viewpoint for water quality and oil spill monitoring with a focus on temporal and spatial resolution aspects. Water quality measures derived from satellite imagery, such as harmful algal blooms, thick oil, and oil emulsions are observable with glint <0.005 sr-1, while oil films require glint >10-5 sr-1. Daily imaging hours range from 6 to 12 h for water quality measures, and 0 to 6 h for oil film applications throughout the year as defined by sun glint strength. Spatial pixel resolution is 300 m at nadir and median pixel resolution was 391 m across the entire field of regard, with higher spatial resolution across all spectral bands in the Gulf of Mexico than existing satellites, such as MODIS and VIIRS, used for oil spill surveillance reports. The potential for beneficial glint use in oil film detection and quality flagging for other water quality parameters was greatest at lower latitudes and changed location throughout the day from the West and East Coasts of the United States. GLIMR scan times can change from the planned ocean color default of 0.763 s depending on the signal-to-noise ratio application requirement and can match existing and future satellite mission regions of interest to leverage multi-mission observations.

Impact of Growth Phase, Pigment Adaptation, and Climate Change Conditions on the Cellular Pigment and Carbon Content of Fifty-One Phytoplankton Isolates.

Owing to their importance in aquatic ecosystems, the demand for models that estimate phytoplankton biomass and community composition in the global ocean has increased over the last decade. Moreover, the impacts of climate change, including elevated carbon dioxide (CO2 ), increased stratification, and warmer sea surface temperatures, will likely shape phytoplankton community composition in the global ocean. Chemotaxonomic methods are useful for modeling phytoplankton community composition from marker pigments normalized to chlorophyll a (Chl a). However, photosynthetic pigments, particularly Chl a, are sensitive to nutrient and light conditions. Cellular carbon is less sensitive, so using carbon biomass instead may provide an alternative approach. To this end, cellular pigment and carbon concentrations were measured in 51 strains of globally relevant, cultured phytoplankton. Pigment-to-Chl a and pigment-to-carbon ratios were computed for each strain. For 25 strains, measurements were taken during two growth phases. While some differences between growth phases were observed, they did not exceed within-class differences. Multiple strains of Amphidinium carterae, Ditylum brightwellii and Heterosigma akashiwo were measured to determine whether time in culture influenced pigment and carbon composition. No appreciable trends in cellular pigment or carbon content were observed. Lastly, the potential impact of climate change conditions on the pigment ratios was assessed using a multistressor experiment that included increased mean light, temperature, and elevated pCO2 on three species: Thalassiosira oceanica, Ostreococcus lucimarinus, and Synechococcus. The largest differences were observed in the pigment-to-carbon ratios, while the marker pigments largely covaried with Chl a. The implications of these observations to chemotaxonomic applications are discussed.

Satellites for long-term monitoring of inland U.S. lakes: The MERIS time series and application for chlorophyll-a.

Lakes and other surface fresh waterbodies provide drinking water, recreational and economic opportunities, food, and other critical support for humans, aquatic life, and ecosystem health. Lakes are also productive ecosystems that provide habitats and influence global cycles. Chlorophyll concentration provides a common metric of water quality, and is frequently used as a proxy for lake trophic state. Here, we document the generation and distribution of the complete MEdium Resolution Imaging Spectrometer (MERIS; Appendix A provides a complete list of abbreviations) radiometric time series for over 2300 satellite resolvable inland bodies of water across the contiguous United States (CONUS) and more than 5,000 in Alaska. This contribution greatly increases the ease of use of satellite remote sensing data for inland water quality monitoring, as well as highlights new horizons in inland water remote sensing algorithm development. We evaluate the performance of satellite remote sensing Cyanobacteria Index (CI)-based chlorophyll algorithms, the retrievals for which provide surrogate estimates of phytoplankton concentrations in cyanobacteria dominated lakes. Our analysis quantifies the algorithms' abilities to assess lake trophic state across the CONUS. As a case study, we apply a bootstrapping approach to derive a new CI-to-chlorophyll relationship, ChlBS, which performs relatively well with a multiplicative bias of 1.11 (11%) and mean absolute error of 1.60 (60%). While the primary contribution of this work is the distribution of the MERIS radiometric timeseries, we provide this case study as a roadmap for future stakeholders' algorithm development activities, as well as a tool to assess the strengths and weaknesses of applying a single algorithm across CONUS.

Determining the optimal spectral sampling frequency and uncertainty thresholds for hyperspectral remote sensing of ocean color.

Using a modified geostatistical technique, empirical variograms were constructed from the first derivative of several diverse Remote Sensing Reflectance and Phytoplankton Absorbance spectra to describe how data points are correlated with "distance" across the spectra. The maximum rate of information gain is measured as a function of the kurtosis associated with the Gaussian structure of the output, and is determined for discrete segments of spectra obtained from a variety of water types (turbid river filaments, coastal waters, shelf waters, a dense Microcystis bloom, and oligotrophic waters), as well as individual and mixed phytoplankton functional types (PFTs; diatoms, eustigmatophytes, cyanobacteria, coccolithophores). Results show that a continuous spectrum of 5 to 7 nm spectral resolution is optimal to resolve the variability across mixed reflectance and absorbance spectra. In addition, the impact of uncertainty on subsequent derivative analysis is assessed, showing that a 3% Gaussian noise (SNR ~66) addition compromises data quality without smoothing the spectrum, and a 13% noise (SNR ~15) addition compromises data with smoothing.