Predicting the Future of the Ocean with Supercomputers
Yasuhiro Yamanaka , Professor
Faculty of Environmental Earth Science/Graduate School of Environmental Science
High school : Tokyo Metropolitan Igusa High School
Academic background : Doctorate at Tokyo University
- Research areas
- Marine Science
- Research keywords
- global warming, numerical simulation, anthropogenic, CO2, marine ecology
How do you predict the future of the oceans?
We divide the earth up into a grid pattern, and express the flow, water temperature and other features of the sea in each grid square using numbers. It is then possible to calculate these conditions from now into the future, using formulas based on the principles of physics. Furthermore, we can also use biological and chemical principles to calculate the supply of minerals from deep sea areas as a result of currents and up/down mixing, to understand the ecosystem of marine plankton, nutrients etc., as well as calculate photosynthesis according to light and water temperature conditions, and understand the food chains among plankton. Furthermore, we have become able to calculate the marine resources such as migrating sauries and small fish such as sardines who feed on plankton, as well as past changes in climate conditions (paleoclimate and paleocean) (Fig. 1).
Predicting the future of the ocean has become possible thanks to developments made between 1990 and 2009 in global systems science, which crosses disciplines such as physics, biology and chemistry, and advances in supercomputers. In 1989 when I graduated university, discussions of global warming and the internet were only just beginning. Predicting the future of the oceans, even if done according to the rules, requires us to look at everything from phenomenon occurring over decades at a global level, such as global warming, to phenomenon occurring in an instant within cells, such as plankton photosynthesis, and involves simplifying many basic processes. As such, you are not guaranteed to get the right results if you perform the calculations too simply. You have to compare it to knowledge gained from observation, and it is important to check your results, but this allows you to discover things that could not be learned by observation. It is important to have a good, broad understanding of the necessary physics, biology and chemistry, and to continue communicating with people doing research into the basic processes in each of these subjects.
Fig. 1: Concept diagram from climate change to variation in marine resources. We integrate these and recreate them by computer.
How will the ocean change with global warming?
A rise in ocean temperatures brings the spring bloom, when plankton are generated in large quantities, forward, reducing the amount of photosynthesis that takes place year-round. Furthermore, fish begin to swim north to find seas of the temperature they prefer, and we know that changes are occurring to the habitat distribution of fish. Furthermore, with a decline in the availability of plankton they use for food, marine resources are also reducing. Future predictions for saury migration calculated in 2005 showed that the development of fry and young fish would slow, and that the speed at which they swim would also slow down, with the result that they would be swept into the Kuroshio extension current and their summer habitat distribution would move significantly east, reducing the size of the adult fish by a substantial amount (Fig. 2). Subsequently, more accurate water temperature and ocean current distributions were introduced, the algorithm for saury migration has been improved, and we are improving monitoring of spawning locations and periods. As a result, the significant changes shown in Fig. 2 have disappeared, but individual basic processes have become more reliable. Furthermore, the ocean absorbs around 1/3 of the carbon dioxide emitted as a result of human activity, and this is lowering the pH of the sea (ocean acidification). This may have a serious impact on marine ecosystem. At present, we are working on research to assess this impact.
What are you aiming to do next?
The model ecosystems for plankton etc. are fairly basic compared to the real thing, so we need to bring them closer to reality. In the end, I would like to create a virtual world in which anything can be understood using a computer, so that we can implement many different simulations based on “what if…” scenarios, and understand the global system better. This will enable us to look after the only earth we have.
Fig. 2 Migration routes for saury, obtained in 2005, at present and at the end of this century. Bigger boats may be needed in order to catch them.