Keywords: numerical simulation, multiple linear regression, MLR, Lanzhou, air pollution, air quality management, urban air quality, meteorological conditions, temperature stratification, environmental pollution, China, urban planning
Numerical model–based relationship between meteorological conditions and air quality and its implication for urban air quality management
Relationships between winter air pollutant concentrations and ten meteorological parameters were investigated based on observed pollutant concentrations and high–resolution meteorological data from weather research and forecast model, and multiple linear regression models were developed for estimating NO2 and PM10 concentrations over urban Lanzhou, North–western China. Results indicated that pollutant concentration correlated better with boundary layer height and potential temperature lapse rate than with other meteorological parameters. There is a lag time of 3–12 hrs for meteorology conditions to have an effect on pollutant concentrations in urban and 12–17 hrs in rural areas, with longer lag time for wind than for temperature stratification. The overall performance of the multiple linear regression models is basically comparable to the widely used comprehensive air quality models and artificial neural network models. The method used here provides another way to estimate possible spatial distributions of pollutant concentrations with less computing time and input data, and could be applied to other cities for urban planning or air quality management purposes.