In 1936, the BBC’s first TV weather forecasts featured hand-drawn maps which illustrated the atmosphere’s shifts and sways. Today weather forecasting looks a little different, as climate simulations are constructed by the world’s biggest computers.

In the present day, governments, businesses, and academics use supercomputers to research climate change. Climate research can help governments understand the particular challenges their countries face, and inform how to address them.

Speakers from Supercomputing Asia 2022 as well as Professor Dale Barker, Director of the Centre for Climate Research Singapore (CCRS) shared how their research is tackling the effects of climate change. They discuss how supercomputing has aided climate research and shaping policy.

The technology behind climate research

CCRS observes the atmosphere, land, and oceans for its climate models. This information comes from satellites, aircraft, and ships. Organisations require “millions of calculations per second” to turn these streams of data into climate simulations, Barker highlighted.

The centre uses its own supercomputer and facilities at the National Supercomputing Centre Singapore to perform these calculations, he said. “Climate science is a ‘big data’ challenge”, Barker emphasised.

Supercomputing may be necessary for predicting the worst excesses of climate change and providing possible solutions, agreed Dr Jack Wells, Science Programme Manager, Nvidia at Supercomputing Asia 2022.

Nvidia is building a virtual model of Earth’s climate conditions with supercomputing, he explains.
The digital twin uses deep learning and AI to make climate simulations more accurate and run faster. The frequent weather disasters over the past few years have been far greater than existing models have predicted, shared Dr Wells.

With Earth’s digital twin, agencies can predict and potentially mitigate these disasters. Nvidia’s goal is to make Earth-2 interactive, so that users can explore what-if scenarios, he shares. This would be as simple as any mobile app.

For example, users could use the model to learn about possible wind and cloud activity in a given location. Agencies could use this information to activate the appropriate wind turbine power plants to the best use of wind activity.

The National Supercomputing Centre Singapore is also using digital twin technology to test how effective green technology can be, shared GovInsider. This helps them improve supercomputing cooling systems and reduce energy consumption.

Shaping Singapore’s climate policy

The CCRS is developing climate simulations to help understand the effect of climate change on Singapore and Southeast Asia. This will provide information such as projected temperatures, rainfall levels, and wind levels, Barker highlighted.

This research can help shape government policy, as it outlines the appropriate timescale required for climate change adaptation. For example, its research supported PUB, Singapore’s national water agency, with its coastal protection efforts.

These measures include building mangroves on the coastline to reduce the impact of incoming waves, explained Hazel Khoo, Director, Coastal Protection Department, PUB. Efforts to adapt to climate change will “impact the lives of all who live in Singapore”, Barker said.

Climate simulations are part of Singapore’s Third Climate Change Study, due for completion 2023. Taking into account the possibility of different future scenarios is one feature of this study.

For example, the centre recognises that the future climate will be different depending on how successful countries are in reducing their greenhouse gas emissions, he explained.

Supercomputing Thailand’s sustainability

Thailand provided further lessons for how supercomputing can support climate research.

Supercomputing helped King Mongkut’s University of Technology Thonburi create a model that could examine and forecast upper Indochina’s weather conditions, shared Dr Kasemsan Manomaiphiboon at Supercomputing Asia 2022.

This model has demonstrated how slowing down deforestation could alleviate future climate disasters. They used the model to study two recent extreme climate events in Thailand: 2015’s drought and 2016’s heatwave, both of which were the worst in decades.

They found that if Indochina’s forest cover was as dense as it was in 1961, these events could have been mitigated, Manomaiphiboon explained.

Gone are the days of hand-drawn weather maps. Today, supercomputing promises to help agencies make sense of and prepare for erratic weather behaviour caused by climate change.