Microsoft, IBM help on pollution

Air pollution in China is a constant threat, and two US companies - Microsoft and IBM - are using their technology expertise to provide air-quality forecasting.

Yu Zheng, a researcher at Microsoft, told China Daily that technology companies "can leverage their computing infrastructures, data management, analytics tools and knowledge in data science to help forecast air pollution".

Zheng said that Urban Computing, a Microsoft research theme that "aims to tackle urban challenges by using big data in cities" can "create solutions that improve the urban environment, human life quality and city operation systems".

"While other companies have a two- to three-day forecasting capability, IBM has leveraged cognitive computing technologies to develop a 10-day pollution trend forecast which is already available to its clients," wrote Jin Dong, associate director of IBM's research division.

"Cognitive computing systems ingest, analyze and understand this data, identifying valuable correlations and providing actionable insight to those fighting air pollution," he wrote. "With machine learning, the systems self-configure and constantly improve creating unprecedented levels of accuracy."

Microsoft has signed with China's Ministry of Environmental Protection, and the environmental protection bureaus in Fujian province and Chengdu, Sichuan province. The company also has created a website called Urban Air and a smartphone app with a 48-hour air pollution forecast for cities across China.

The first part of Microsoft's plans includes the ability to decipher "real-time and fine-grained air quality" of an arbitrary location by using data at monitoring stations, as well as meteorological, traffic, human mobility and road network data.

This step in the process to determine the root cause of the pollution includes "studying the correlations between vehicular emission and air quality; identifying the spatio-temporal causality between air pollutants of different cities; and suggesting new locations for additional pollution monitoring stations," according to Microsoft's Urban Air webpage.

IBM's China Research lab launched its "Green Horizons" initiative in 2014.

"Using scenario modeling, [IBM] came up with a way to create hypothetical 'what if' scenarios - enabling city officials to try out the effectiveness of different action plans," IBM's website says.

IBM's first client was the Beijing Environmental Protection Bureau, and the company also has signed deals with Baoding and Zhangjiakou, Hebei province, which will serve as one of the host cities for the 2022 Winter Olympics with Beijing.

"IBM has launched a multidisciplinary initiative to support China in delivering on its ambitious energy and environmental goals," said Brad Gammons, general manager of IBM's Global Energy & Utilities Industry.

"The 10-year project sets out to leap beyond current global practices in three critical areas: air quality management, renewable energy forecasting and energy optimization for industry."

In December, Beijing officials declared two "red alerts" as a warning that heavy pollution was expected for several days across the capital.

Five other cities soon followed(Tianjin and four cities in Hebei province: Baoding, Handan, Langfang and Xingtai).

Ming Xu, an assistant professor at the University of Michigan,said that he believes people are well-educated on the metrics and indicators of air pollution issues in China, like the PM2.5 air quality index, which measures particulate matter with a diameter less than 2.5 microns that can be harmful to one's health.

"But people need to know also how these numbers are produced (e.g., from direct measuring or estimation/forecasting)," Xu said. "If it is estimation and/or forecasting, scientists should do a better job to explain the method and make sure people understand its limitations."

"The Chinese government went to great lengths to reduce pollution for the 2008 Summer Olympics in order to protect the health of athletes and spectators alike," Dong said.

"This time they will be able to target specific activities - with the maximum effect, but with much less impact on economic activity and the daily lives of citizens," Dong added.

Nonetheless, "forecasting smog is different from forecasting air quality," Zheng said.

"[Smog] is a kind of weather condition, whereas [air quality] is the concentration of air pollutants. Air quality is impacted by multiple complex factors, such as weather conditions (foggy, smoggy, rainy), traffic conditions, pollution emission from factories, and the dispersion condition of a location," Zheng said.

jackfreifelder@chinadailyusa.com

(China Daily USA 01/19/2016 page2)