Research On The Monitoring And Simulation Of PM_2.5Concentration In Microenvironment Based On Computational Fluid Dynamics And Remote Sensing

With the acceleration of industrialization and urbanization in China,the problem of environmental pollution is becoming more and more serious.As a severe weather phenomenon in China in recent years,haze has posed a serious threat on the human health and social sustainable development.According to the researches on pollutants in China's major cities and megacities,PM_2.5 is the main cause of the haze weather.In order to forecast the haze weather quantitatively,real-time monitoring of PM_2.5 becomes an important work to carry out.At present,nationwide ground envirommental monitoring sites have been basically completed.However,only the data from ground monitoring sites can not reflect the the micro-scale spatial distribution of PM_2.5 well.Remote sensing technology has become one of the effective methods to study the spatial distribution of PM_2.5 because of its ability of reflecting continuous ground spatial information rapidly,comprehensively and objectively.In recent years,many scholars have been working on the research of air quality status,pollution change characteristics and forecasting future development trend,but there are less researches on air quality in urban microenvironment combined with remote sensing technology.In this paper,high resolution images are used to inverse the distribution of PM_2.5 concentration from the perspective of image feature technique,and to analyze the quantitative relationship between atmospheric background fields and the distribution of PM_2.5 concentration in real architectural complexes combined with the CFD numerical simulation technique and optimized neural network method,then to estimate the distribution of PM_2.5 concentration in different background fields in real architectural complexes.The main research contents are as follows:?1?Taking Shanghai as experimentation area,GF-1 images and the measured data of PM_2.5 from ground monitoring sites as main research data,from the perspective of the degradation mechanism of the remote sensing images in haze weather,the paper has analyzed the characteristics of haze in the space,band and frequency domains,targetedly extracted the thickness of haze using dark channel apriori algorithm,optimal cloud detection,homomorphic filtering method and wavelet transform method,established the PM_2.5 concentration inversion model,and compared the four methods.By carrying out the research on inversing PM_2.5 using remote sensing technology exploratorily from the perspective of the degradation characteristics of the haze images,we have found that based on the atmospheric scattering model,the method to extract the thickness of haze is more accurate that obtaining the atmospheric transmittance by using the dark channel apriori algorithm firstly,and refining it through the guided filter secondly.And PM_2.5 inversed using the thickness of haze through the method above is more accurate and more consistent with the haze distribution.?2?Based on the numerical simulation method of computational fluid dynamics?CFD?,using software such as FLUENT,taking two high-rise and low-rise architectural complexes in Xuhui District,Shanghai as the research object,this study has analyzed the relationship between the distribution of PM2.4 concentration and the pattern of architectural complexes,the speed and direction of wind,temperature and other factors.The ratio of height and width,shape and arrangement mode of an architectural complex will have effects on the flow fields in it.The more complicated the arrangement mode of a village is,the more complicated the flow field in it is.The air will accumulation when the spacing between the buildings is too small and when the buildings are tall in the front while short at the behind,which is not conducive to the proliferation of PM_2.5.When the velocity of background wind becomes larger,PM_2.5 high concentration areas will form where the spacing between the buildings are small,but the larger spacing will be conducive the proliferation of PM_2.5.Temperature does not have notable effects on the distribution of the flow fields.However,the higher the temperature is,the more active the gas is,and the more notable PM_2.5 proliferates,only that this effect is relatively small compared to the impact of wind velocity.Flow directions will have remarkable effects on the distribution of the flow field inside architectural complexes.It is conducive to the proliferation of gas when the flow deviates from the orientation of the buildings.?3?Based on the CFD method,this study has constructed the model of flow field distribution under the condition of different atmospheric background fields,analyzed the quantitative relationship between simulated flow fields and real PM_2.5 concentrations combined with the optimized neural network methods,and estimated the distribution of PM_2.5 concentration under the condition of different atmospheric background fields in different architectural complexes.Taking two high-rise and low-rise architectural complexes in Xuhui District,Shanghai as research object,taking the meteorological data on January 21,2015 as atmospheric background field,this study has simulated the wind field in the microenvironment of the architectural complex by CFD technique,simulated the quantitative relationship between flow fields and PM_2.5 concentrations using L-M optimized BP neural network,Bayesian regular optimized BP neural network and genetic optimized BP neural network individually,compared the algorithms from the perspective of spatial distribution,accuracy and stability of thedata,and finally chosen L-M algorithm which had performed best to estimate the distributions of PM_2.5 concentration at different wind velocities.From the whole point of view,compared with in high-rise architectural complexes,in low-rise architectural complexes,the distribution differences of PM_2.5 concentration vary less with the variance of wind velocity.From the point of wind velocity,PM_2.5 concentrations decrease with the increase of the wind velocities,while the distribution differences of concentration increase.The innovations of this study are as follows:?1?Based on the remote sensing images with haze,this study has improved the method of the inversion of remote sensing images.On this basis,image characteristic technique has been used to extract the haze thickness,and then obtained the distribution of PM_2.5,thus compared with the data from the measured values.The results showed that the PM_2.5 obtained by the improved remote sensing image inversion algorithm is extremely accurate.?2?Based on the distribution of the macroscopic haze thickness extracted by remote sensing images,the CFD method has been used to simulate the continuous three-dimensional distribution of the flow fields in different architectural complexes under the condition of real background fields in microenvironment.Combined with optimization algorithms,this study has discussed the quantitative relationship between micro-scale flow fields and the distribution of PM_2.5 concentrations,and combined the study of PM_2.5 on macro-scale and that on micro-scale.?3?Based on the simulations,this study has established the optimization model between micro-scale flow fields and the distribution of PM_2.5 concentrations.On this basis,the meteorological factors have been used to estimate the haze of real architectural complexes,thus to reveal the law of the influence of meteorological factors on the spatial variation of PM_2.5 concentrations in microenvironment.The results of this study can provide a new idea and method for estimating the spatial distribution of PM_2.5 concentrations in microenvironment.