Authors: M. El Jarbi; T. Slawig
Addresses: Department of Computer Science, Algorithmic Optimal Control, CO2–Uptake of the Ocean, Excellence Cluster The Future Ocean, Christian-Albrechts-Platz 4, 24118 Kiel, Germany ' Department of Computer Science, Algorithmic Optimal Control, CO2–Uptake of the Ocean, Excellence Cluster The Future Ocean, Christian-Albrechts-Platz 4, 24118 Kiel, Germany
Abstract: The objective of this paper is to apply an optimisation method for an enhancement of a climate model, namely a one dimensional nonlinear marine ecosystem model of NPZD (N for dissolved inorganic nitrogen, P for phytoplankton, Z for zooplankton and D for detritus) type. The main goal of this work is to obtain periodic parameters and to investigate the impact of the linearisation on the model output. For this purpose, two experiments are carried out. The first experiment deals with linearisation about synthetic data. In the second experiment, we linearise about reference trajectory based on measurement data. In both cases, we use the obtained optimal periodic parameters also in a validation experiment with the original nonlinear version of the model. The quality of the model-data fit of the original nonlinear NPZD model using periodic parameters obtained with synthetic data is poor compared to the data fit quality obtained with measurement data.
Keywords: biogeochemical modelling; NPZD model; dissolved inorganic nitrogen; phytoplankton; zooplankton; detritus; parameter optimisation; open loop control; optimal control; periodic parameters; marine ecosystems; marine ecosystem models; discrete control; climate models; linearisation; nonlinear modelling.
International Journal of Mathematical Modelling and Numerical Optimisation, 2015 Vol.6 No.1, pp.22 - 39
Received: 02 Dec 2013
Accepted: 27 Aug 2014
Published online: 16 Apr 2015 *