Summary of Abstract Submission

Abstract Submission No. IO50-09-0077PresentationPoster


Syam Sankar*1, Stefano Ciavatta2, Luca Polimene2, Lorenzo Marin3, Nandini Menon1, Roberto Pastres3, Annette Samuelsen4, Lasse Pettersson4

1 Nansen Environmental Research Centre India, India
2 Plymouth Marine Laboratory, United Kingdom
3 University of Venice, Italy
4 Nansen Environmental and Remote Sensing Center, Norway


Ecological indicators are functions of environmental variables and are used to describe the status of marine ecosystems. If marine ecosystem models are to be used for the prediction of ecological indicators we need to first evaluate the reliability of the model through a quantitative assessment of the model performance. The objective of the present study is to provide a preliminary assessment of the reliability of European Regional Seas Marine Ecosystem Model (ERSEM) to simulate net primary production, a widely used ecological indicator, in the central Arabian Sea by means of a thorough sensitivity analysis. The model couples ERSEM with the General Ocean Turbulence Model (GOTM). The sensitivity analysis of the model was carried out in two steps, i.e. through a preliminary overall screening analysis that selected a few parameters that underwent a subsequent Monte Carlo analysis for parameter ranking. The screening analysis was carried out using the Morris method, focusing on 207 pelagic parameters of ERSEM, subdivided into 21 groups depending on the physiological/biogeochemical process they describe (i.e. photosynthesis, metabolic losses, grazing etc.). The screening showed that the group characterizing the zooplankton loss terms is the most important (i.e. it got the highest sensitivity index). The group of parameters representing the Q10 of zooplankton was the second most important. The group comprising nutrient parameters and the group representing factors regulating phytoplankton metabolism were ranked 3rd and 4th respectively. The 70 individual parameters contained in these 4 groups were subjected to a Monte Carlo-based global sensitivity analysis, to rank the most relevant parameters for the simulation of annual NPP. The ranking of the parameters is discussed in relation to the processes they represent, and how these are influenced by the environmental forcings in the Arabian Sea.