Simulations for small spaces
Measuring only 27 kilometers from north to south, Singapore is one of the smallest nations in Asia. Even if scientists were to try simulating its weather on some of the best global climate models, it would just show up as a blip in one of the grid cells. Thankfully, researchers from the Climate and Water Research cluster at the local Tropical Marine Science Institute have been able to downscale global and regional climate models to capture the whims of the weather in the city-state.
If you thought calculating a country’s climate would be way less intensive than computing the whole world’s, you’re in for a surprise. To process the terabytes of data required for detailed and accurate weather prediction in such a small geographical area—and to do so quickly enough to issue four forecasts a day—the scientists have to rely on supercomputers.
In particular, they run their islandwide weather simulations at a high resolution of 400 meters on the Köppen system at the National Supercomputing Center (NSCC) Singapore. With 1.2 petabytes of storage and 160 teraFLOPS of power, the system is specifically designed for climate and environment research.
Beyond short-term weather forecasts, the researchers also run simulations to investigate the effects of Singapore’s weather and climate on its urban climate, as well as the seasonal haze phenomenon. They’ve even extended their simulations to the Southeast Asia region, modeling climate change at 8–10 kilometer resolutions on multi-decade time scales. Singapore may be small, but its supercomputing prowess is mighty.
This article was first published in the print version of Supercomputing Asia, January 2021.
Click here to subscribe to Asian Scientist Magazine in print.
———
Copyright: Asian Scientist Magazine; Illustrations: Oi Keat Lam.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.
