Inversion Of Atmospheric Aerosol Optical Thickness And Its Application To Haze Detection

With the rapid development of industrialization and urbanization, the content ofparticulate matter in the atmosphere is increasing. Air pollution is quite serious in themajority areas of China, especially the haze weather occurred frequently innationwide. The haze weather is a major meteorological disaster which can causegreat unfavorable effects on transportation, industrial and agricultural production andhuman health. This paper take Hangzhou district as an example to research theaerosol optical thickness retrieval and haze detection method based on data fromMODIS, AERONET observation station, the National Meteorological Bureau and theground meteorological data supplied by Hangzhou meteorological station. In thispaper, the specific research work is written as follows:1. Analyzes the basic principle of remote sensing inversion of atmosphericaerosol optical thickness, establishes the aerosol optical thickness inversion modeland put forward a method to determinate the aerosol types based on6S atmospherictransmission equation. After data preprocessing, this custom aerosol types and thedefault city aerosol types in6S equation are used respectively to invert theatmospheric aerosols optical thickness in Hangzhou district, and use the sunphotometer observation as a basis for the inversion results proceed to verify theaccuracy. The results show that the relative errors of the inversion values and theobserved values are within20%, the relative error by using custom aerosol type ofinversion value is3%lower than that using the default aerosol types inversion values.The inversion results show that the aerosols optical thickness value is slightly largerthan that considering the hygroscopicity characteristics of aerosol and climate inHangzhou is humid, it is consistent with the weather conditions of weather stations inHangzhou.2. The basic principles and grading criteria of atmospheric haze detected areresearched, define the two key indicators of haze detection are visibility and relativehumidity. The estimation model of visibility is established by using the AOT data andthe expression of model is amended according to the result of calculation. It can beconcluded by testing the calculation accuracy of the model that removing ofindividual points, the maximum relative error of estimation value and real value of visibility is no more than12%. The remote sensing estimation model of relativehumidity is set up by using MOD05atmospheric data and MOD07temperature dataas the data source and put forward two weighted methods of atmospheric precipitationvalues, the model test results show that the minimum relative error between real valueand the estimate value is2.1%. Then detect haze days in Hangzhou area in2012based on the visibility and relative humidity, use year-round fog data in2012providedby Hangzhou weather stations as the standard data to test the accuracy of inspection.The results show that that cleaning day and fog detection accuracy is higher than hazedays; for mild, slight haze days the detection accuracy is higher than that of moderateand severe haze days; and the detection accuracy in spring and summer is higher thanin autumn and winter. Finally the possible impact factors and detection error in thehaze detection are analyzed.