Study On The Correlation Between Land Surface Temperature And Air Pollution Basic On MODIS Data In Wuhan

Air pollution is an important environmental issues which affect urban development, in recent years, the haze event happened frequently which brings trouble to the human production and living, so it is important for people to get the quality of atmospheric environment timely and effective. Currently, the way we get air quality is from the air monitoring stations which set up in the different regions of the city.By processing the data of theses air monitoring stations, we get the city’s air quality data. Current monitoring method can only get air quality of several monitoring stations, instead of the entire city’s distribution.Satellite remote sensing monitoring showed a rapid and macro, economy, objectivity characteristic in the study of air pollution, which cause the attention of scholars, So this article try to use the MODIS satellite remote sensing data to calculate the air quality. The main results are as follows:(1)The relationship between the surface temperature and the main atmospheric pollutants, such as SO2, NO2, PM10, CO, O3 and API showed a significant correlation in the 0.01 level, the correlation coefficient are respectively 0.486,0.727,0.524,0.652, 0.834 and 0.834.(2) To build model of relationship between the surface temperature and API from the Angle of annual and seasonal. We selected the multivariate regression model for the yearly model, relational model index for spring, multivariate regression model as summer, autumn and winter relationship model. Model for the whole year and autumn model fitting effect is good, their determination coefficient R2 are 0.44 and 0.56 respectively, the fitting effect for the model of spring, summer and winter are all poorer, their determination coefficient R2 are 0.11,0.04 and 0.05.(3) Analysis the yearly model and all season model in the step of 1,5,10,25,50. In the step of 1, the fitting precision and prediction precision are both poor, so we can’t get the air quality by the surface temperature directly. With the higher step, two precision are both improved. In the step of 50, the fitting precision and prediction precision of each model is the best, the fitting precision and prediction precision of the yearly model were 100% and 94.4% respectively; Spring model is 100% in both precision; in summer model, the fitting precision was 93.65% and the prediction accuracy is 83.33%; Autumn’s fitting precision and prediction precision are both 88.89%; Winter’s fitting precision was 90.48%, and the prediction accuracy is 100%. The step of 50 is also the classification standard of API, so the model can satisfy the demand for the prediction of API level.(4) Using the yearly model and the winter model calculated the distribution of the air quality of Wuhan on February 7,2016, as the same of the actual release, the API level is secondary grade.