DEOS PhD Dissertation Proposal Defense: Most Israt Jahan Mili

Ocean primary production (PP) by phytoplankton forms the basis of the marine food web and is a critical component of the global carbon cycle. 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. High productivity has also been linked to recurrent hypoxia in this region. Ship-based in situ measurements of PP are limited in temporal and spatial coverage. Bio-optical models of PP enable the extrapolation or estimation of PP over more extended temporal and spatial scales. Such models have been based on in situ measurements of bio-optical properties as well as satellite observations, or combinations thereof.
Categories of models include wavelength-integrated vs wavelength-resolved, depth-integrated vs depth-resolved, chlorophyll-based vs phytoplankton absorption-based and carbon-based models. Wavelength-resolved models are more computationally intensive but provide a more accurate representation of the light field. Absorption-based models have the advantage of more directly characterizing light absorption by phytoplankton in contrast to chlorophyll-based models. Existing global PP models are primarily tuned for open ocean conditions and do not necessarily perform well on regional scales, such as the nGOM.
The optically complex nature of the river-influenced nGOM necessitates the use of a regionally tuned optical algorithm for the estimation of PP. Given this background, the first objective of this work was to describe and develop a regionally tuned, hyperspectral, wavelength-resolved, absorption-based PP algorithm for the nGOM, termed the wavelength-resolved model (WRM). Using in situ observations of photosynthesis-irradiance (P–E) in combination with vertical profiles of diffuse attenuation, we applied the WRM to the estimation of PP in major water mass types in the region. Average water column-integrated primary production (IPP) was 0.408 mol C m^-2 d^-1 and  ranged between 0.094 – 1.574 mol C m^-2 d^-1, with the lowest values observed in the outer and the highest in the mid-shelf region. The performance of the WRM was also evaluated in comparison to a conventional wavelength-integrated model (WIM), which is a chlorophyll-based model using wavelength-integrated photosynthetically active radiation (PAR) as input. Preliminary results revealed that the WRM yielded slightly higher IPP values compared to the conventional PAR-based WIM model, particularly at offshore stations. There was a slight negative bias in the WIM estimates of IPP as compared to the WRM with a mean bias error (MBE) of -0.08 mol C m^-2 d^-1 (0.96 g C m^-2 d^-1) and a p-value of 5.87 × 10^-5 for a two-sided t-test. Furthermore, a modified version of the WRM, termed the wavelength-resolved model estimate (WRME), was developed to facilitate the application of the hyperspectral, absorption-based algorithm to satellite-derived hyperspectral ocean color observations.

The second objective of this study will be the adaptation of the algorithm to hyperspectral satellite ocean color observations such as those from the recently launched Plankton, Aerosol, Cloud, Ocean Ecosystem (PACE) satellite mission and its hyperspectral Ocean Color Instrument (OCI) sensor. Additional future hyperspectral missions, such as the Geosynchronous Littoral Imaging and Monitoring Radiometer (GLIMR), will further enhance this capability. This will enable the observation of PP across broader spatial and temporal scales in the nGOM.  We will also examine the vertical structure of phytoplankton absorption, using relationships between chlorophyll fluorescence and in situ measurements of phytoplankton absorption to assess the influence of vertical structure in IPP estimates.

Finally, the third objective will be to characterize the observed patterns in satellite-derived IPP in relationship to in situ measured biogeochemical (nutrients, pCO₂) and physicochemical properties in the region and more broadly, to major circulation features in the nGOM.

Meeting ID: 950 0971 7853

Passcode: 404584