Abstract
The northern Gulf of Mexico (nGOM) is a region of high biological productivity largely influenced by the nutrient input from the Mississippi and Atchafalaya rivers. Prior studies utilizing bio‐optical algorithms to estimate primary production (PP) in this region have generally employed wavelength‐integrated models (WIM) involving attenuation of photosynthetically active radiation integrated over the 400–700 nm spectral range. Here, we describe a hyperspectral, wavelength‐resolved, absorption‐based PP model (WRM). This WRM is more computationally intensive than wavelength‐integrated approaches but provides more accurate representation of the subsurface light field. Additionally, the WRM more directly characterizes light absorption by phytoplankton in contrast to chlorophyll‐based models. To implement the algorithm, bio‐optical observations and photosynthesis‐irradiance (P‐E) measurements were acquired during five GulfCarbon cruises (January, April, July, November 2009, and March 2010). Hyperspectral downwelling irradiance was determined using a HyperPro optical profiler along with phytoplankton spectral absorption from filterpad absorption measurements. Results from the WRM produced consistently higher values of PP at depth than a conventional WIM, particularly in offshore waters. Water column‐integrated primary production (IPP) ranged between 0.09–1.66 mol C m −2 d −1 (1.08–19.92 g C m −2 d −1 ). This range is comparable to or higher than historical observations, and future work would benefit from further independent validation. A modified version of the WRM, the wavelength‐resolved model estimation, was also described, which utilizes P‐E parameters derived from regional empirical relations. This novel hyperspectral PP algorithm can be applied to hyperspectral ocean color satellite observations and is adaptable for use in other regions. Ocean primary production (PP) is a key component of the global carbon cycle and forms the basis of the marine food web. Here we describe a regionally tuned, wavelength‐resolved, absorption‐based PP algorithm for the northern Gulf of Mexico, which includes representation of how light is utilized by phytoplankton across the spectrum. The model was implemented using photosynthesis‐irradiance and bio‐optical observations collected during multiple cruises. Estimated values of PP using this approach were compared to that of a conventional wavelength‐integrated model. The hyperspectral absorption‐based model consistently estimated higher PP values at depth, particularly in offshore stations, compared to the conventional wavelength‐integrated model. We also describe a modified version of the absorption‐based wavelength‐resolved hyperspectral algorithm that enables estimation of PP solely from ocean color satellite observations, making it well suited for the hyperspectral observations from the platforms such as the Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) and the upcoming Geosynchronous Littoral Imaging and Monitoring Radiometer (GLIMR). This regionally tuned hyperspectral algorithm can be used for other regions and for open ocean applications where information about photosynthetic parameters is available or can be estimated and other model inputs properly scaled. The absorption‐based wavelength‐resolved model yielded higher primary production (PP) at depth than a conventional wavelength‐integrated model, particularly at mid‐ and outer shelf stations This novel hyperspectral primary productivity algorithm will be applicable to new emerging hyperspectral ocean color satellite observations The hyperspectral PP model can be adapted for other regions where information about photosynthetic parameters is available