The Research Of Data Visualization And Prediction Models Of Urban Air Quality

There are closely linked between ambient air quality and people’s production and life. With the acceleration development of industrial and urbanization, the sustained growth of energy and traffic scale has led to the air quality (such as respirable particulate matter (PM10, and PM2.5), sulfur dioxide (SO2), nitrogen dioxide (NO2), nitric oxide (NO), nitrogen oxides (NOx), carbon monoxide (CO), ozone (O3) etc. as major pollutants) from bad to worse. Air pollution is gradually becoming a major concern problem.For decades, a city Environmental Protection Monitoring Center has accumulated a large number of air quality monitoring data. These data may not completely describe ambient air quality of the city accurately, but they can provide important foundation of analysis and judgment. With the continuous improvement of ambient air quality monitoring system, monitoring data is growing more rapidly, and traditional data processing methods and means cannot be right to carry out effective analysis and processing of high-dimensional air quality monitoring data. Hence we need a new technical support to achieve analysis and utilization of the available data urgently. It is an important research topic of how to find valuable information from massive data, analysis of the contact of the complex linkages between pollutants and air quality change in the future to make a correct prediction and evaluation and achieve effective control of the region’s ambient air quality ultimately. This work possess important theoretical significance and practical value, can also provide decision-making reference for the environmental management sector, in order to better safeguard the social production and living environment of urban and rural areas.We have used parallel coordinate method as the main tool to study the concentration time distribution of the major air pollutants and interrelated features of the visual analysis in the ambient air quality monitoring data; different pollutant concentration data were analyzed at different times of change law, discussed the characteristics of the relationship between different pollutants through the visual analysis method based on clustering; discussed how to use2D and3D parallel coordinates in the limited screen space to display more complex data, to make it easier pollutants in the complexity of the spatial and temporal distribution of information.The statistical theory and methods are used in the paper, through the preprocess of the ambient air quality monitoring data (PM10, SO2, NO2, NO, NOx, CO, O3and other monitoring data) in the last decade of the city, we calculate the air pollution index (API) data through national standards and the moving average of the different moving average cycle of API, analysis and forecast of API using of statistical moving average method. We have obtained good average predictions of the decades as a whole of the city through consolidating the longer and the shorter moving average cycle of API moving average and at the same time we have got better average predictions of the different periods in one year. All these research will provide important basic data support to the management and control of urban air quality.