【作者】 程兴宏；

【导师】 徐祥德；

【作者基本信息】 中国科学院研究生院 ， 气象学， 2008， 博士

【摘要】 污染源排放清单是空气质量模式的重要输入数据,排放源的不确定性是影响当前空气质量模式研究结果准确性的主要因素之一,亦对污染源的调控产生较大影响。近年来排放源模式、源反演模型等间接源排放估算方法成为定量计算污染源排放量及其时空变化规律的主要研究方向之一,它们可以为空气质量模式提供高时空分辨率、多排放参数的污染源排放清单。采用简单易行、具有较高可信度和应用价值的“Nudging”源同化修正模型反演具有季节变化特征、高分辨率、动态的多尺度排放源清单,成为提高空气质量预报水平的重要途径之一。本文研究的主要目的是采用多尺度源同化反演模型在中国各区域、各季节和不同天气条件下进行模拟试验研究,以探讨空气质量模式“源同化”模型的模拟效果及其理论认识;本文还研究气象场模拟技术的改进、不同空间分辨率的观测资料以及OMI高分辨率柱浓度卫星资料的应用对空气质量预报水平提高的显著性问题;本文着重讨论奥运时段华北不同地区污染源的不同排放控制情景对北京市空气质量的影响以及各种物理、化学过程的贡献率,试图寻找周边污染源的关键控制区,以提供奥运会期间北京地区空气质量保障与调控的科学依据。以下是本文的主要研究工作和结论总结。（1）本文定量分析了“Nudging”污染源同化反演方法在中国不同地区、不同季节、不同天气过程的适用性及其对空气质量预报改进的长期客观效果检验。通过大样本模拟试验,揭示出中国区域排放源的月际和季节变化特征;证实引入源同化反演模型后,CMAQ模式在中国不同地区、不同季节和不同天气条件下预报水平均有较显著的改进。获取了全国47个城市区域具有季节变化特征的动态排放源,其预报效果明显优于2000年David Streets排放源;采用上述源同化反演方法,不仅显著改善了SO₂、NO₂浓度的预报趋势,而且可明显减少预报误差;SO₂排放源强增加导致源影响过程贡献的SO₂浓度增大是源同化方法提高SO₂浓度预报水平的主要物理机制,排放源的改善导致化学反应过程贡献的NO₂浓度增大是源同化方法提高NO₂浓度预报水平的主要化学机制;SO₂、NO₂排放源的改善对PM10、O3逐时预报以及O3最大值预报具有间接影响,可提高PM10、O3逐时及O3最大值浓度预报水平,趋势预报效果较好。（2）比较了两种气象模式（MM5/WRF）提供的气象场对华北地区SO₂、NO₂源同化修正效果及其浓度预报水平的影响差异;采用CMAQ模式中的IPR过程分析模块着重分析了WRF模式中边界层高度、相对湿度模拟的改进对SO₂、NO₂源同化效果的影响。分析结果表明,边界层高度和相对湿度模拟的改进对于SO₂、NO₂浓度趋势模拟效果有一定的影响;相对气象模式模拟效果的改进而言,排放源的改善对于提高SO₂、NO₂浓度的趋势及浓度模拟水平的作用更显著。（3）采用不同空间分辨率的SO₂、NO₂实测资料模拟分析了中国地区SO₂、NO₂不同观测信息密度对SO₂、NO₂源同化反演及空气质量预报的影响,重点分析了奥运时段华北地区SO₂、NO₂浓度加密观测对改善SO₂、NO₂排放源和空气质量预报的重要作用。分析结果表明,采用较高分辨率的实测资料进行SO₂、NO₂源同化修正时,可明显减小SO₂、NO₂浓度的预报误差,华北地区较高分辨率的观测信息对于改善源同化修正效果及奥运时段华北地区SO₂、NO₂浓度的趋势预报效果十分重要,尤其是对SO₂浓度的预报尤为重要。（4）本文采用OMI SO₂、NO₂卫星遥感产品着重讨论北京周边地区SO₂、NO₂排放源的影响效应以及该产品在改善NOX源同化修正效果及提高NO₂浓度预报水平中的初步应用。提出了卫星遥感-地面大气污染观测综合分析及其卫星遥感变分场同化源技术途径。根据卫星遥感与地面观测相结合思路,利用地面实测污染资料变分订正卫星遥感OMI高分辨率柱浓度资料,并采用卫星遥感变分场源同化模型获取了华北地区高分辨率的动态NOX排放源,个例模拟试验研究表明该技术方案具有提高污染物预报水平的可行性。分析结果表明,无论冬季还是夏季,华北地区SO₂、NO₂实测浓度与OMI SO₂、NO₂卫星遥感柱浓度的高低值区分布较一致,SO₂、NO₂柱浓度卫星资料可适用于华北地区卫星遥感-地面观测综合变分分析;经变分订正的OMI SO₂、NO₂柱浓度的空间分布特征可看出,北京的西南、东南地区污染源对北京地区的SO₂、NO₂浓度的影响较大;采用经变分订正、分辨率较高的OMI NO₂卫星遥感资料同化修正排放源时,WRF-CMAQ模式对华北地区冬、夏季NO₂浓度水平预报和趋势预报可取得较显著的改善效果;采用经变分订正的、高分辨率的OMI NO₂卫星遥感资料进行源同化时,可模拟得到与实测浓度分布较一致的、细致的、高分辨率的NO₂浓度信息,弥补了采用地面有限实测资料模拟效果的不足,卫星观测资料对于城市区域尺度的NO₂浓度模拟尤为重要。OMI NO₂高分辨率卫星遥感资料对于奥运期间华北和北京地区的NOX排放源调控以及提高场馆尺度NO₂浓度预报水平有一定的实际应用和参考价值。（5）在夏季奥运时段华北地区同化修正源的改善以及改进后的WRF模式技术的基础上,着重分析了北京区域气候背景、风向频率分布等气象条件以及奥运时段华北不同地区污染源的不同排放控制情景对北京市SO₂、NO₂、PM10、O3浓度的影响和各种物理、化学过程的贡献率。找到了周边污染源关键控制区,为奥运期间北京地区空气质量保障与调控决策提供科学依据。分析结果表明:1)周边污染源对北京地区空气质量影响的主要通道是北京的西南和东南部地区。北京西南部的污染源对北京地区SO₂、NO₂、PM10浓度的影响最大,东南部的污染源影响次之,西北部的污染源对北京地区SO₂和PM10有一定程度的影响。北京西南部对SO₂浓度影响程度最大,PM10次之,NO₂最小。周边污染源对北京地区O3浓度的影响较小;2)北京周边污染源应急调控方案需考虑当时的风场结构特征,选取“上风方”周边污染源为主要调控目标,其中重点调控北京周边城市西南与东南区域,其次为西北区域;另外在弱风或静风条件下需重点调控北京本地污染源。3)控制北京西南地区的SO₂、NO₂污染源排放时,北京及周边地区SO₂、NO₂、PM10三种污染物浓度的减少幅度形成由南向北的带状分布,而且带状分布地区PM10浓度减少比例存在日变化特征。4)夏季多数三级以上污染日北京地区SO₂、NO₂、PM10浓度减少幅度与北京主体“上风方”（西南、东南部）SO₂、NO₂污染源减排比例呈同步变化关系,且北京西南部三种污染物的减排效果明显优于东南部的调控效果,因此北京西南地区的污染源控制对于改善北京地区的空气质量更加重要。更多还原

【Abstract】 Air pollution emission inventory is an important input data of air quality model. The uncertainty of emission inventory is a primary source of error in air quality forecasts and also has a great effect on regulation of air pollution sources. Indirect estimation methods of emission sources such as inverse model of sources is one of important research direction on the quantitative calculation of discharge capacity of pollutant and the law of temporal and spatial variation for emission sources. And the pollution emission inventory of higher resolution and multiple emission parameters is provided to air quality models using the inverse models of sources. Inverse of high resolution and multi-scale emission inventory which has seasonal variation characteristics derived from the simple, reliable, accurate and adaptive nudging scheme is one of important approach to enhance the level of air quality forecasts.The main aim of the paper is 1) to analyze simulation effect of the nudging-based inverse model and theoretical progress by simulated test in different region, season and weather conditions in China; 2) to evaluate the long-term objectivity of improving of the level of air quality forecasts by the above inverse model; 3) to study the impact of improved simulation of meteorological element field, application of observation data with different spatial resolutions and improved retrieval data of tropospheric SO₂ and NO₂ from OMI/Aura satellite（OMI SO₂ and OMI NO₂） with high resolution on correction effect of inverse model and improving of the level of air quality forecasts; 4) to evaluate the impact of different controlling situation of emission sources in different regions in north of China to air quality in Beijing, to compute contribution rate of different physical and chemical processes in the above controlling situation, to find the key controlling region of surrounding emission sources and to provide the scientific basis for service guarantee of air quality forecasts in Beijing and regulation of surrounding emission sources during the 2008 Olympic Games.The following are the main conclusions and results:（1） In this paper, the applicability of multi-scale and corrected emission inventory in different region, season and weather conditions in China using the nudging-based inverse modeling approach and the long-term objectivity of improving of the level of air quality forecasts by using the corrected emission inventory were evaluated quantitatively. The result of long-term simulation test showed that 1) the inverse model of SO₂ and NOX sources using adaptive nudging scheme is suitable for different region, season and weather conditions in China; 2) predictions of SO₂ and NO₂ concentrations in January, April, August, and October using the emission estimations derived from the nudging technique showed remarkable improvements over those based on the original emission data which is from Streets 2000; 3) the accurate relatively and multi-scale emission inventory which has seasonal variation characteristics is obtained by using the nudging-based inverse modeling approach; 4) the error field in the emission data can be reduced effectively through an inverse modeling procedure using observed air pollution levels, and using the improved emissions led to significant improvement in the forecasts of SO₂ and NO₂ concentrations in all seasons. 5) using the improved emissions of SO₂ and NOX indirectly effect hourly forecasts of PM10 and O3 and maximum forecasts of O3 in a day and the forecasting effect of PM10 and O3 is improved through the nudging-based inverse modeling approach.（2） Comparison of the impact of simulation of meteorological element field using MM5 and WRF models to correction effect of inverse model and improving of the level of air quality forecasts was given. Emphasis was placed on the discussion of impact of improved simulation of Planet Boundary Layer （PBL） height and relative humidity using WRF model on correction effect of inverse model of SO₂ and NOX sources. The results indicated that improved simulation of PBL height and relative humidity has certain influence on improvement in the tend prediction of SO₂ and NO₂ concentrations, and can reduce obviously the forecasting error. In comparison with the impact of improved simulation of meteorological element field, correction effect of emission inventory on improvement in the tend prediction of SO₂ and NO₂ concentrations and reduction of forecasting error is more effective.（3） The effect of observation information density of SO₂ and NO₂ concentrations on correction effect of inverse model and improving of the level of air quality forecasts was simulated by using the observed data of SO₂ and NO₂ concentrations with different resolution. The important effect of dense observational data of SO₂ and NO₂ concentrations in north of China on correction effect of inverse model and improving of the level of air quality forecasts was analyzed emphatically. The result showed that the forecasting error of SO₂ and NO₂ concentrations can be reduced effectively through the inverse modeling procedure using dense observational data with higher resolution. The dense observational data with higher resolution in north of China is extremely significant for correction effect of inverse model and improvement in tend prediction of SO₂ and NO₂ concentrations, especially for the forecasts of SO₂ concentration.（4） The impact of surrounding emission sources to the level of SO₂ and NO₂ concentrations in Beijing by using the improved retrieval data of OMI SO₂ and OMI NO₂ with high resolution was evaluated. And application of OMI SO₂ and OMI NO₂ data to correction effect of inverse model and improving of the level of air quality forecasts was discussed emphatically. The results revealed that 1) the spatial distribution characteristics of OMI SO₂ and OMI NO₂ column densities is in accordance with that of observational data of SO₂ and NO₂ concentrations in winter or summer. And OMI SO₂ and OMI NO₂ column densities are applicable to variational processing by means of observational data of SO₂ and NO₂ concentrations on the ground in north of China; 2) from the spatial distribution characteristics of OMI SO₂ and OMI NO₂ column densities which corrected by variational method, the conclusion was obtained that surrounding emission sources in southwest and southeast area have a great effect on the level of SO₂ and NO₂ concentrations in Beijing; 3) the tend prediction and the level of NO₂ concentration are improved obviously using revised emission inventory of NO₂ source by OMI NO₂ data with high resolution, which corrected by variational method; 4) it has a certain reference and actual application value in improvement of emission inventory of NO₂ source in north of China and Beijing and improving of the forecast of NO₂ concentration around Olympic gymnasiums.（5） The impact of different controlling situation of emission sources in different regions in north of China to air quality in Beijing and contribution rate of different physical and chemical processes in the above controlling situation were evaluated and computed, based on improved simulation of meteorological element field using WRF model and relatively accurate emission of SO₂ and NOX sources with high resolution, which are corrected by using the nudging-based inverse modeling approach. The results of control and sensitivity tests showed that 1) the main transport channels of impact of surrounding emission sources to air quality in Beijing are southwest and southeast area of Beijing. Among two main affected zone, southwest area has the most significant influence, the next is southeast area. And emission sources in northwest region have an influence on the levels of SO₂ and PM10 concentrations in Beijing to some extent. Among three main pollutants, emission sources in southwest area have the most significant influence on the level of SO₂ concentration in Beijing, the next is PM10 concentration, the least is NO₂ concentration. And the impact of surrounding emission sources to O3 concentration in Beijing is relatively small; 2) the regulation schemes of SO₂ and NOX emission sources around Beijing need consider characteristics of wind field structure at that time. The main regulation object is upwind emission sources. The key regulation area is southwest and southeast region around Beijing, the next is northwest region. As breeze and static wind blowing, the local emission sources are the key of regulation; 3) when emission of SO₂ and NOX sources in southwest region is controlled, decreasing amplitude of SO₂, NO₂ and PM10 concentrations in Beijing and adjacent area from south to north form the zonal distribution characteristics. And the decreasing ratio of PM10 concentration in the zonal distribution area has diurnal change. 4) as the grade of air pollution in summer in Beijing is third-level or above, the decreasing ratio of SO₂ and PM10 concentrations changes simultaneously with the decreasing ratio of emission of SO₂ and NOX sources in main upwind area of Beijing （southwest and southeast regions）. 5) the regulation effect of surrounding emission sources in southwest area is superior to that in southeast area. Regulation of emission sources in southwest area of Beijing is very important for improving air quality in Beijing during the 2008 Olympic Games.更多还原