Pollutant Release And Simultion Of PM_2.5 Diffusion In Biomass Straw Open Burning

Straw open burning has become an important source of global gaseous pollutants and particulates and will bring great harm to the atmospheric environment and human health.Based on the data of crop（rice,wheat,corn and cotton） yield in Tianjin,ratio of straw to grain,and different pollutant emission factors（EFs）,an accurate pollutant emission inventory was established.We also used the CALPUFF model to simulate and analyze the PM_2.5 emission and diffusion process of straw open burning in Tianjin.And the BP artificial neural network was used to predicte and verify the PM_2.5 concentration by combining with the actual observation data of straw open burning.Based on the theoretical calculation,the pollutants emission inventory from straw open burning from 1996 to 2014 was established in Tianjin.The annual pollutant emissions were 39.39 Gg with an annual growth rate of 9.92%.Among the four crop straw,the contribution rate of corn stalks was the highest,followed by wheat straw,and the contribution rate of rice straw was the lowest.Also the Monte-Carlo method was used to simulate the pollutants emission inventory uncertainty,and the overall uncertainty was-44.56%⁸0.35%.The migration and diffusion process of PM_2.5 from wheat straw and corn stalk open burning in Tianjin was respectively simulated by CALPUFF system.The high PM_2.5concentration region is concentrated in the fire point and high fire point dense area.The PM_2.5 concentration also will gradually decrease under the downward wind direction.And the concentration is also proportional to the number of fire points,the more fire points and the higher pollutants concentration.Based on the observation of the meteorological data and PM_2.5 concentration during the burning process,it was found that the time of the PM_2.5 peak concentration decreases with the increase of the observation distance.The two process during the straw open burning respectively were fire burning and smoldering burning,resulting in two peaks of PM_2.5 concentration at each observation point.Based on the BP artificial neural network,the PM_2.5 diffusion concentration prediction model of straw open burning was established.And the comparison between the observed data and the predicted data showed that the average absolute error rate is 1.77%.It illustrated that the BP artificial neural network model could be used to predict the PM_2.5 diffusion concentration of straw open burning at microscopic scale.