Multivariate Regression Modeling for Coastal Urban Air Quality Estimates

Multivariate regression models for real-time coastal air quality forecasting were suggested from 18 to 27 March 2015, with a total of 15 kinds hourly input data (three-hours-earlier PM and gas meteorological parameters Kangnung (Korea), associated two-days-earlier Beijing (China)). Multiple correlation coefficients between the predicted measured PM10, PM2.5, NO2, SO2, CO O3 concentrations 0.957, 0.906, 0.886, 0.795, 0.864 0.932 before yellow sand event at Kangnung, 0.936, 0.982, 0.866, 0.917, 0.887 0.916 during 0.919, 0.945, 0.902, 0.857, 0.892 after event. As significance levels (p) multi-regression analyses less than 0.001, all very significant. Partial presenting contribution variables 6 output using presented three periods in detail. Scatter plots their distributions values showed quite good accuracy modeling performance current time six high applicability.