Algorithm maps out city's carbon emissions

A team of Chinese scientists is using artificial intelligence to accurately map out carbon emissions in big cities to help urban managers make more evidence-based plans in fighting global warming.

It is part of China's efforts to peak its carbon dioxide emissions by 2030 and achieve carbon neutrality by 2060.

Researchers from the Chinese Academy of Sciences' Aerospace Information Research Institute are collecting data from a refitted carbon-monitoring vehicle that is regularly driven through bustling parts of Shenzhen, Guangdong province. Panoramic cameras to capture street scenes have also been mounted on the vehicle.

The data it collects is transmitted to a laboratory at the institute in Beijing, where a new deep learning algorithm is being trained.

Driven by the team's algorithm, the cameras can discern sources of carbon emissions and carbon sinks, such as moving cars and buildings, and carbon-fixing vegetation. This allows the AI model to produce estimates of their carbon contributions, either positive or negative.

"Take a single site as an example. The AI analysis revealed that small cars contribute the most to emissions, followed by trucks," said Wang Li, a researcher from the institute. "The surrounding roads and buildings also contribute to emissions."

Shenzhen has six carbon monitoring towers, but that's too few to make a detailed emission map, and building more towers would be costly, making alternative and complementary monitoring methods necessary, said Zhang Yonglin, a research assistant on Wang's team.

It occurred to Zhang, who has a computer science background, that AI could lend a hand and he developed the algorithm.

"You have at least 100 elements at one single site, and the objects keep changing," said Zhang. "Only AI can calculate viable results."

So far, 650 kilometers of driving has been completed, obtaining more than 100,000 records of street images and road traffic carbon emissions in Shenzhen.

After training, the AI model can produce a high-definition map accurate to within 100 meters and its accuracy for road traffic carbon emission intensity has exceeded 92 percent, according to Zhang.

The model is projected to provide a useful guide for city managers pondering issues like what causes a certain block to see higher carbon emissions and searching for possible solutions. That is the kind of work that remote-sensing carbon satellites and monitoring towers cannot do.

"The results can remind policymakers to take measures such as increasing efficiency of access in certain road sections, or planting more trees on a block," Zhang said.

"Sometimes, too much greenery is not all beneficial for reducing carbon concentrations when it results in a pipeline effect. That's what the AI map can tell you."

Last month, the Ministry of Ecology and Environment set a goal of building a modernized ecological environment monitoring system by 2035 via stepped-up efforts for digital and intelligent upgrading.

New technologies including AI and the internet of things will be applied to establish an intelligent monitoring system, it said.

Xinhua