Visual analysis for exploring the relation between air pollution, environmental factors and respiratory diseases

Abstract

Using visualization tool for data analytics makes the complex data more understandable so that some inherent patterns and relationships can be revealed more efficiently and clearly. This paper provides a such paradigm for visual analysis on public health issues. By exploiting the advanced visualization tool based on cloud and artificial intelligence, we collect and analyze several data sets on PM2.5 air pollution, environments and public health in Beijing in order to discover the intrisic relations between those factors. The visual analytics illustrates that the incidences of respiratory diseases increase as the concentration of PM2.5 in the air increases and that a strong correlation exists to temperature. PM 2.5 concentrations are higher in cold winter months than in hot summer months. Meanwhile, an unexpected discovery from the investigation was the pattern found for air concentrations of PM2.5 throughout a 24-hour period. The levels of PM2.5 reach peak at midnight and drop to their lowest level at 2pm in the daytime. Most importantly, there exists a direct positive correlation between the number of emergency room visits for respiratory diseases and high levels of PM2.5 in the atmosphere. These interesting findings can provide direct supports for strategy/decision making in public health and emergency administration.