Crowdsourced Data Fusion for a Liveable Future

From left: Prof. Edward Ng, Prof. Leung Yee, Prof. Ren Chao, and Dr. Xu Yong of the Institute of Future Cities at CUHK

High-density city living is a global trend. According to the United Nations, more than half of the global population are urban dwellers, and the figure will reach two-thirds by 2050. Large-scale urbanization will result in soaring energy demand and denser buildings, followed by problems like urban heat island and smog. To build a sustainable future, data acquisition on urban forms and functions is urgently needed for solutions and adaptation strategies. Researchers at CUHK and Xi’an Jiaotong University have devised an innovative data acquisition and fusion method—proved effective for integrating crowdsourced data—to produce 2D or even 3D urban maps. With more encompassing and accurate data, the government and academia are enabled to devise comprehensive and long-term urban planning and development policies.

More Sophisticated 3D Data

The research team is composed of Prof. Leung Yee, director of the Institute of Future Cities at CUHK, and the Institute’s researchers Prof. Ren Chao and Dr. Xu Yong, together with Dr. Ma Fan and Prof. Meng Deyu of Xi’an Jiaotong University. Dr. Xu said, ‘The new mapping technology, which combines data fusion and image processing, enables integration with satellite images from sources like National Aeronautics and Space Administration, European Space Agency and Google Earth Engine to produce 2D or even 3D digital urban maps.’ Professor Ren said, ‘The accurate information can translate the cities’ land surface and urban morphological data into meteorological, climate change, energy assessment and other forecasting models to tackle environmental issues including urban heat island and air pollution. Unlike the traditional labour-intensive method, the efficient and accurate new urban morphological data acquisition will offer a tremendous help to the planning of smart cities, public health policies and climate change measures.’

The Institute’s associate director Prof. Edward Ng added, ‘2D images offer data for weather forecast and air pollution, but 3D data cover more details: building height, usage, window size, energy consumption, population density, ventilation, etc. Compared to the 80% accuracy rate of the 2D maps, the 70% accuracy rate of the new technology is a satisfactory one. To build a more sophisticated urban database, we intend to apply the 3D technology to other places, especially the developing countries.’

In the Kowloon 3D map produced by CUHK, the height of buildings is indicated by colours. The buildings in red are the tall ones

To help build liveable high-density cities, the team has been providing data support and urban climatic analysis for urban development projects locally and internationally, namely, ‘Hong Kong 2030+’ and ‘Guangdong, Hong Kong and Macau Dawan District’. Their data have also been deployed by European Union’s Group on Earth Observation and World Meteorological Organization. The team will support the Digital Belt and Road initiative and conduct its feasibility study, which aims to efficiently and comprehensively digitalize the urban forms of 20 countries and 80 cities along the Belt and Road. In the meantime, the team will also participate in the Healthy Cities Analysis Project for Latin American Countries, supported by the World Bank and United Nations, to provide training in urban database and climate-sensitive urban planning. ‘As not many are familiar with the Belt and Road cities and remote regions, the database will help them understand the places and foster sustainable development in the areas,’ said Professor Ren.

The research team took part in the Data Fusion Contest 2017, and stood out from over 800 worldwide competitors as one of the top four awarded teams. The contest is co-organized by the Institute of Electrical and Electronic Engineers and the Geoscience and Remote Sensing Society with an aim to promote crowdsourced data fusion and city morphology analysis. The theme this year is to obtain land use classification in different urban environments—also known as ‘local climate zones (LCZs) mapping’. The accuracy and applicability of the classification methods that the joint team had proposed impressed the international jury panel. The team will present its work and receive a trophy at the International Geoscience and Remote Sensing Symposium to be held in July in the US.

WUDAPT for Ideal Urban Planning

The award-winning method is part of the free and open-sourced World Urban Database and Access Portal Tools (WUDAPT) initiative which, through simple and uniformly classified LCZs mapping, presents the large-scale urban morphological data such as architectural morphology and land use. The initiative has already mapped 120 cities worldwide, including 57 Chinese cities and some Belt and Road cities. Led by Professor Ren, CUHK is the Asian representative of the initiative, overseeing the mapping of cities’ LCZs on the continent and the expansion of the database’s application to provide users with information on climate and weather models, energy balance research, as well as urban planning and implementation.

‘This is the first time a complete database has been established for the study and promotion of sustainable development. It answers the calls for better planning of housing supply, balanced development, smart cities and cleaner environment in both Hong Kong and mainland China,’ said Prof. Edward Ng. Several on going international collaborations are exploring various applications of WUDAPT data on sustainable urban development in China, namely, the Eco-city's performance evaluation and planning for the Beijing-Tianjin-Hebei region, study on the climate change and natural hazards in the Yangzi River Delta region, and Urban Heat Island and Extreme Heat events in the Pearl River Delta region.

The Pearl River Delta region’s local climate zone maps unveil the urban development over two decades

Effort in Making Cities Liveable

With the closing of the ‘Hong Kong 2030+’ public engagement exercise in April, the issues of better liveability and a more balanced spatial development have been very much in the public consciousness. As a solid data foundation is essential to the planning of a liveable city in future, CUHK will further its effort in enhancing the crowdsourced data fusion technology to enrich the urban database locally and globally. A new chapter of sustainable development will be unveiled.

This article was originally published in No. 499, Newsletter.