NERC fellowship: Improving Estimates of Ocean Primary Production

A mechanistic model which allows ocean colour data to be used to aid understanding of carbon fixation in the global oceans

Ocean colour measures the sunlight that is scattered back out of the oceans When sunlight penetrates into seawater it is absorbed by the water molecules themselves and by plants in the water. The water molecules absorb red light (long wavelength) very strongly but not blue light (short wavelength; see graph) - this is why deep water looks blue. The pigment chlorophyll a in phytoplankton (algae) absorbs blue and red light and reflects green and yellow (which is why the plants look yellowy-green to us - see graph). Some of the sunlight is not absorbed but is scattered; some of this scattered light leaves the ocean and is detected by satellite sensors.

Looking at the ratios of green and blue light detected by the sensor allows chlorophyll concentrations in the surface waters to be estimated The pictures on the left shows a real colour image of the Arabian sea (top left) and an image of the estimated chlorophyll concentration (bottom left). This image is interesting as it shows that chlorophyll can also be used as a tracer to identify eddy structures in the water. Maps of chlorophyll concentrations can be generated globally on a daily basis, giving us a huge amount of information on the state of the ocean ecosystems (below). See SeaWifs web page for more info

Using chlorophyll concentrations to calculate carbon fixation is not trivial! To quantify primary production (carbon fixation) we need to relate the amount of carbon to the amount of chlorophyll present. In each phytoplankton cell the ratio of carbon to chlorophyll (C:chl) changes according to the environment that the algae have experienced (e.g. water temperature and nutrient availability); this is called physiological acclimation. In the oceans there is a lot of light near the surface but this falls off exponentially with depth. However, there is more nutrient available deeper in the water (because it gets quickly used up by the plants in the surface waters). If this situation was static then the algal cells would find ratios of carbon, nitrogen and chlorophyll that allowed them maximum growth for the given light and nutrient levels (known as balanced growth). It is not a static situation however as the water column may stratify during the day and then mix at night (convective overturning). This means that the algal cells are continually changing their internal physiology to adjust to their new environment (known as unbalanced growth - see schematic representation on right).

Modelling all of this...

The idea is to simulate the changes in ocean colour due to turbulent mixing and ecosystem evolution. To simulate ocean mixing I am using GOTM, the general ocean turbulence model. Into this I have coupled the Hadley Centre Ocean Carbon Cycle (HadOCC) which describes the marine ecosystem in terms of flows between four components (inorganic Nitrogen, Phytoplankton, Zooplankton and Detritus). To model physiological changes in phytoplankton a Lagrangian (particle-tracking) submodel is required since it is necessary to know a cell's environment history. Geider et al's 1998 equations are then used to predict the amount of carbon, chlorophyll and nitrogen in each cell. The next step is to predict the absorption and scattering of light by the sea water constituents. Absorption is dominated by the water molecules themselves and chlorophyll, whereas light scattering is more complicated and will be modelled by looking at the size distributons of the particles (possibly dominated by detritus) in the water.