Nowadays the study of the marine ecosystem is becoming quite important especially in the context of marine ecosystem management and climate change.
The impacts of climate change and its related phenomena on marine systems present a complex challenge when taken in the context of other pressures faced by the ocean.
In situ monitoring, remote sensing satellite and mathematical models are all good methods for estimating biogeochemical patterns in shelf seas, but each of them are affected of errors.
In situ measurements can provide accurate and precise estimates of a wide range of marine ecosystem variables, but they are discrete in time and space, covering a small portion of the whole ecosystem.
Satellite observations provide frequent estimates of variables to a wider scale, but they are able to detect a narrow range of variables such as chlorophyll and particularly in coastal areas satellites measurement are influenced by land inputs such as organic matter and suspended solids. Sophisticated and often coupled modelling approaches offer a reliable way of including all of the parameters necessary to begin to understand how the intricate web of marine life interacts, and how it might be affected by natural and anthropogenic changes.
Complex hydrodynamic and ecosystem models can be brought together to predict the marine ecosystem dynamics.
Numerical models are an approximations of the full dynamic and thermodynamic equations, consequently their solutions are inaccurate due to uncertainty in the model structure, parameterization and errors in the input data. One way to correct for errors in the model solutions is to apply a procedure known as data assimilation (DA), which blends the model approximations with observations of the real ocean in a least-square error sense. This blending takes account of model errors as well as measurement errors in the observations.
The main goal of my current work at the PML is to evaluate the capacity of the coupled biogeochemical hydrodynamic model GOTM-ERSEM to predict the seasonal evolution of biogeochemical variable and fluxes through the assimilation of in situ and remote sensed data.
For this purpose I am applying data assimilation technique called Ensemble Kalman filter (EnKF), to the assimilation of MODIS chlorophyll data over a whole year 2009 in the station L4 of the Western English Channel.
I am also trying to compare quantitatively the output of the reference model run and of the assimilation scheme with five time series of in situ data (chlorophyll, nitrate, ammonia, phosphate, silicate), collected at the long-term monitoring station L4 for the year 2009.
More on this research when the final results are ready ….