Satellite Informatics System for Surface Particulate Matter Distribution

Sponsored by
 Home / Overview  

Learn About AQ



People are most concerned about their local environment. To get their attention about Air Quality (AQ), it is important to have specific AQ information down to their neighborhood. Currently, this is not available in HK, or anywhere else.

Based on satellite remote-sensing data, we have developed algorithms to calculate the surface Particular Matter (PM) distribution with resolution down to 1 km. The validity of our algorithms has already been proven with air qulaity monitoring station data from the EPD.

Through this project, we have integrated these algorithms and developed a system capable of deriving wide-area high-resolution surface PM distributions using satellite data. The information content provided by this system is dramatically better than what is available with conventional monitoring technologies.

This system provides a comparative edge for professionals in high-value service sectors like environmental consulting, urban planning and other sectors as HK transits to a knowledge-based society. With this system, these professionals can provide more specific AQ-related assessments and recommendations for their private and public clients.

Current available AQ information

Everyone would agree that having good Air Quality (AQ) is important. However, what most people really care or concern about is their local environment, not the overall situation or average over a large area. To get their attention and support for AQ improvement, it is most useful to provide them with AQ information specifically relevant to their housing estate or business district.

Remotely sensed data for particulate matter monitoring

With fourteen AQ monitoring stations, Hong Kong has already one of the densest AQ monitoring network in the world. Yet, we are still very far from being able to provide a comprehensive description of AQ over HK. For example, many distinct air sheds (e.g., Tseung Kwan O, Sai Kung, Sheung Shui, the southern side of Hong Kong and Lantau and others) are still not adequately covered by the current monitoring network; it is difficult to imagine being able to differentiate the AQ within the same district. The cost for doing so with technologies currently used at the AQ monitoring stations is prohibitive.


Not just for the average people, it would also be beneficial for many sectors, both private and public, if wide-area high density AQ information can be provided. Environmental consulting companies can use it as primary data for analysis or as input to their models to provide specific AQ assessments or recommendations for their client. Planers can use it to choose different design options in their urban or infrastructure development projects. Real estate agents can use it to provide prospective buyers with specific AQ information about the real-estate development projects they are pushing for, while property developers can use it during their site selection process to avoid the embarrassment of building in overly polluted areas. Business related to remedial AQ products like air cleaners or filters can use the information to target specific districts in the deployment of their sales forces. Finally, researchers in both private and public sectors can also use it to conduct more detailed district-based AQ-related studies.

Our progress and approach

During the past two years we have developed a number of algorithm modules capable of using satellite products to derive detailed PM distributions over Hong Kong and the PRD area. The validity of the algorithms has been confirmed with real-time PM measurements from AQ monitoring stations operated by the Hong Kong Environmental Protection Department (EPD).

Our aim is to integrate these algorithms and develop an environmental informatics system capable of deriving high-resolution and wide-area surface PM distribution using remote-sensed data from satellites. Through the web graphical users' interface (GUI), the users can specify spatial and temporal parameters, and the system will extract the datasets for downloading, or generating a distribution map, or both.