Uncertainty Analysis And Diagnosis Research Of Regional Gridded Modes Inventory

As the key basic data to support regional air pollutant research,air quality model simulation,forecast and early warning,etc.,emission source inventory has played an increasingly prominent role in air pollution prevention and control research in recent years,and has become an important reference for air pollution prevention and control.As the regional air pollution in my country has entered the stage of refined prevention and control,the emission source inventory has also developed rapidly.A relatively complete and practical emission inventory characterization method system has been constructed.Through the processing of time,space and species spectrum,different scales have been established.High-precision gridded emission source inventory to meet the needs of air quality simulation.Nonetheless,due to reasons such as spatial-temporal and species allocation processing,as well as insufficient representation of emission factors and activity level data,my country’s gridded emission source inventory still has large uncertainties,resulting in biased model simulations.Uncertainty analysis is a diagnostic analysis of the sources and impacts of uncertainties that may lead to emission inventory during the compilation process,and is an important means to improve the quality of emission inventory.However,the existing uncertainty analysis methods only consider the impact of parameters such as activity levels and emission factors on the compilation of emission inventories,and quantify the uncertainty of the total emission.However,there is still a lack of corresponding analysis methods and research cases,resulting in a lack of systematic evaluation of the gridded emission source inventory input into the model.In response to this deficiency,this study established a quantitative uncertainty analysis method for gridded emission source inventories by establishing uncertainty quantification methods for time allocation,space allocation and species allocation on the basis of the quantitative uncertainty analysis method framework for emission source inventories established by the team.A framework for deterministic analysis methods.On this basis,taking the 3km grid emission inventory of Guangdong Province in 2017 as a case,typical emission sources were selected,and quantitative uncertainty analysis was carried out on the time distribution,space distribution and species distribution of the grid emission source inventory processing process.The main sources of uncertainty in the gridded process of different emission sources are compared,and the key improvement directions for future gridded emission source inventory processing are proposed.Based on the above scheme,the main conclusions are as follows:1)The uncertainty of time allocation of emission source inventory mainly comes from time representation,regional representation,etc.According to the quantification results,the average uncertainty of time allocation in months for road mobile sources,power plant sources,non-road mobile sources,industrial sources(including solvent sources)and biomass open combustion sources is(-20%,21%),(-35%,38%),(-22%,30%),(-30%,45%)and(-36%,54%).For industrial sources,industries with greater uncertainty in monthly distribution are pesticide production,metal containers,etc.The main reason is that the above-mentioned industries are affected by supply and demand relations,production curves,etc.,and there is a certain degree of difference in monthly production changes in different years.2)Uncertainty in the spatial distribution of emission source inventories mainly comes from temporal representation,representational data representation,etc.In this study,spatial allocation factors were established based on population,land use,Point of Interest(POI),and fire point data,and based on the results of step-by-step Monte Carlo Method(MCM)simulations,the spatial allocation uncertainty of biomass burning sources,non-road mobile sources,and industrial sources was quantitatively analyzed.The results show that the time-representative average coefficients of variation of biomass combustion sources,non-road mobile sources,and industrial sources in a 3km grid are 0.11,0.61,and 0.44,respectively,and the representative uncertainties of characterization data are 0.22,0.17,and 0.15,respectively.Uncertainties in the distribution factors established due to the year and type of characterization data,respectively.3)The source of uncertainty in species allocation in the emission source inventory is mainly the uncertainty caused by the representativeness of the component spectrum.Based on different chemical mechanisms,this study selects typical emission sources,and the results quantitatively analyze the uncertainty of species allocation in the emission source inventory.The results show that the industries with higher uncertainty in species allocation include metal products,automobile manufacturing,and pharmaceutical manufacturing,which are industrial emission sources.Among them,the uncertainty of chemical mechanism of CB05 with fewer species in the same industry is significantly higher than that with more species.Many are based on the chemical mechanism of SAPRC07.4)This study takes the 2017 Guangdong Province emission source inventory as an example.According to the quantitative analysis results of total uncertainty and uncertainty of time,space and species distribution,industrial sources and non-road mobile sources are selected as typical emission sources.Based on MCM Simulation to capture uncertainty in gridded emission source inventories.The 95% confidence interval for uncertainty for industrial sources is(-79%,204%);for non-road mobile sources(-59%,107%).The reason for the difference is that there are differences between the data representing the spatial-temporal distribution factors of different emission sources,which leads to the uncertainty of the grid distribution process.Compared with the total uncertainty results,the gridded emission source inventories established in this study have a wider range of uncertainty due to the consideration of the uncertainty generated by the allocation process.