Grid Facilitates Ozone Predictions
Remember learning about ozone? It was easy to relate to that distinctive smell that signaled the onset of precipitation. Even the chemical formula had a simple elegance, O3.
We learned that a ring of these triatomic molecules created a protective buffer way up in the earth’s atmosphere, and that this layer filters out up to 99 percent of the sun’s harmful UV radiation, which would otherwise cause DNA damage in humans and animals.
Ozone has a bit of a dual nature, however: a life-sustaining substance at higher altitudes becomes an air pollutant when it occurs at ground-level. And at higher concentrations, it can cause cause serious health problems.
Scientists had already established that ozone concentrations, both in the atmosphere and on the earth’s surface, are linked to meteorological conditions, like temperature and prevailing winds, so perhaps long-term climate patterns would have a role to play as well.
Eleni Katragkou, a climate scientist at Aristotle University of Thessalonikihe (AUTh) in Greece, decided to test this hypothesis. She wanted “to predict ozone behaviour in a changing climate, in order to be able assess the impacts on air quality, human health, agricultural production and ecosystems.”
“People with lung diseases, children, older adults, and people who are active outdoors may be particularly sensitive to ozone,” Katragkou explained. “[Ozone] also affects sensitive vegetation and can damage crop production and ecosystems.”
Using the grid computing resources of the AUTh computing centre (an EGI site), Katragkou’s team performed a series of regional climate-air quality simulations for two future decades (2041–2050 and 2091–2100) and one control decade (1991-2000) to study the impact of climate change on surface ozone in Europe. The simulations relied on an established climate model, called scenario A1B, developed by the Intergovernmental Panel on Climate Change (IPCC).
The conclusions, published in the Journal of Geophysical Research, indicate that levels of ground level ozone are set to increase near the end of the century, with the highest concentrations expected for south-west Europe.
The grid resources were crucial for the accuracy of the model, and allowed the simulations to be done in a reasonable time frame. Performed on a single desktop computer, the same job would have taken 40 years to complete.
“The usual bottleneck for performing those types of simulations at a finer resolution is the huge demands on CPU time. This makes me think that grid computing may facilitate very much our future work in this direction,” Katragkou stated.
Full story at European Grid Infrastructure Case Study