Improving the Algal Bloom Prediction in the North Sea by Ensemble Kalman Filtering in the GEM/BLOOM Model

The ecological state of the North Sea surface water can be indicated by ocean variables such as the Chlorophyll-a (Chlfa) concentration. Chlfa is the principal photosynthetic pigment and is common to all phytoplankton and can therefore be used as a measure of phytoplankton biomass. The GEM/BLOOM model developed at Deltares is a generic ecological model that simulates transport of substances in a water system along with various ecological processes. This model is able to estimate the Chlfa concentration. Models are always prone to errors due to assumptions made in the development and the use of numerical approximations. Such errors can be reduced through the use of data assimilation and thus can significantly improve the forecast. The ensemble Kalman filter (EnKF) is a generic data assimilation method which is suited for highly nonlinear models with a large scale. This filter is validated by the use of twin experiments on the GEM/BLOOM model. It successfully improves the prediction of Chlfa,but however shows filter divergence in some grid points. The performance is further improved by the use of the Ensemble Square Root Filter (ESRF) with a localized analysis. Finally, application of this filter to assimilating daily MERIS remote sensing images is explored and shows to be promising, but requires more tuning before it can operate.Mathematical PhysicsApplied MathematicsElectrical Engineering, Mathematics and Computer Scienc