Till startsida
Sitemap
To content Read more about how we use cookies on gu.se

Modelling Coastal Seas

Chapter in book
Authors Anders Omstedt
Published in Biological Oceanography of the Baltic Sea
Pages 78-79
ISBN 978-94-007-0667-5
Publisher Springer
Place of publication Dordrecht
Publication year 2017
Published at Department of marine sciences
Pages 78-79
Language en
Links www.springer.com/cn/book/9789400706...
Keywords Coastal seas, modelling, Baltic Sea, oceanography
Subject categories Biochemistry and Molecular Biology, Oceanography, Hydrology, Water Resources

Abstract

The use of computational fluid dynamics (CFD) to analyze and predict environmental changes has increased considerably in recent decades. Numerical models are now standard tools in research and in a wide range of practical applications. Intensifying concern about human influence on climatic and environmental conditions has increased the need for multidisciplinary modeling efforts, including the numerical modeling of oceans, lakes, land surfaces, ice, rivers, and the atmosphere. Scientists have traditionally developed specialized models limited to application within their own disciplines. Today, increasing efforts are being made to develop Earth System models that includes major processes that one need to consider when dealing with climate and environmental changes. The models rely in general on conservation laws, including many processes that are not known in detail. These rather unknown processes then need to be parameterized in different ways. For example turbulence which is always present in coastal seas are poorly understood and needs to be parameterized in the models. This is also true for a number of chemical and biological processes. There is no standard parameterization yet available for biological processes such as plankton growth and ecosystem changes, instead most parameterizations are based on some observations available. However, one can regard the new and updated models as a systematic collection of present available knowledge. The models can therefore help us to identify missing understanding and where new research programs needs to be developed. CFD can however not profess adequately without reference to experimental validation. This is also a good reason why models and field experiments or monitor programs need to be strongly linked to each other. Coastal Sea models do not only include codes for solving the conservation laws, they also include initial data (such as salinity and temperature) and forcing data (such as weather, river runoff, atmosphere and land emissions from nutrients and carbon components). There are different kinds of Coastal Sea models available, from simple box models to coupled three dimensional atmosphere-land-ocean models. They are often developed for different applications and use different kind of forcing fields. One class of models is process based models (e.g. Omstedt, 2015) and another class is three-dimensional models (e.g. Meier, Feistal, Piechura et al., 2006) both types which have been used in many Baltic Sea applications. Here we will illustrate the strength of process oriented models from one such study (Gustafsson, Omstedt, Gustafsson, 2015), which was used to analyze CO2 dynamics in the Baltic Sea. In this study the modeling was focused on how air-water CO2 fluxes responds to parameterizations of organic alkalinity, gas transfer, phytoplankton growth, and to changes in river loads. One result clearly demonstrated how air-water CO2 fluxes depend on river load of carbon. The modeling was guided by the new source pCO2 of data available in the Baltic Sea, Figure 1. By analyzing different aspects in the modeling the calculations were coming closer to observation. For example a new parameterization of cyanobacteria improved the result. Another interesting model study is illustrated in Figure 2 where Meier et al (2012) compared the accuracy of Baltic Sea models to produce biogeochemical parameters. This is one of the first studies that develop the concept of adding results from different models together in ensemble calculation. In general the results illustrated that the models were not too far from observations but the best representation of the data were the ensemble mean. The model community therefore now enters a new era as not a single model should be used to give management advice. Instead a number of different models need be used. This is in line with the modeling development within global climate change illustrating that managements actions in the future needs to be closely linked to assessments activities.

Page Manager: Webmaster|Last update: 9/11/2012
Share:

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?