Comparison of Five Eulerian Air Pollution Forecasting Systems for the Summer of 1999 Using the German Ozone Monitoring Data

Eulerian state-of-the-art air pollution forecasting systems on the European scale are operated routinely by several countries in Europe. DWD and FUB, both Germany, NERI, Denmark, NILU, Norway, and SMHI, Sweden, operate some of these systems. To apply such modeling systems, e.g. for regulatory purposes according to new EU directives, an evaluation and comparison of the model systems is fundamental in order to assess their reliability. One step in this direction is presented in this study: The model forecasts from all five systems have been compared to measurements of ground level ozone in Germany. The outstanding point in this investigation is the availability of a huge amount of data – from forecasts by the different model systems and from observations. This allows for a thorough interpretation of the findings and assures the significance of the observed features. Data from more than 300 measurement stations for a 5-month period (May–September 1999) of the German monitoring networks have been used in this comparison. Different spatial and temporal statistical parameters were applied in the evaluation. Generally, it was found that the most comprehensive models gave the best results. However, the less comprehensive and computational cheaper models also produced good results. The extensive comparison made it possible to point out weak points in the different models and to describe the individual model behavior for a full summer period in a climatological sense. The comparison also gave valuable information for an assessment of individual measurement stations and complete monitoring networks in terms of the representativeness of the observation data.     

多尺度空气质量模式系统及其验证I. 模式系统介绍与气象要素模拟

主要介绍一个多尺度空气质量模式系统及其在东亚地区的应用. 这套模式系统充分考虑了天气系统与中、小尺度气象过程对污染物的输送、扩散、转化和迁移过程的影响, 区域与城市尺度之间大气污染物的相互影响, 以及污染物在大气中的气相与液相化学过程、非均相化学过程、气溶胶过程和干湿沉积过程对浓度分布的影响, 可用于区域与城市尺度对流层臭氧、大气气溶胶、能见度和其他空气污染物的预报和环境评价. 模拟的气象要素(风向、风速、温度和湿度)与TRACE-P 和ACE-Asia期间三架飞机上获取的观测资料的比较结果表明, 模拟值与其相应的观测值具有非常好的一致性, 它们的相关系数都超过了0.96.     

天津市空气污染数值预报试验中的模式系统

介绍了天津市空气污染数值预报实验中城市空气污染数值预报模式系统的构成及各模块的功能。在天津市空气污染预报实验期间, α中尺度气象模式对区域范围内的逐时风、温、湿及其降水预报取得了较好的效果, 且模式的计算时间短; β中尺度气象模式在边界层内具有较高的垂直分辨率, 模拟结果合理, 其中计算与观测的风向差小于６０ｏ的概率达到６４３％ , 计算与观测的风速比值接近于１; 预报的地面ＳＯ２、ＮＯｘ和ＴＳＰ浓度与同步监测值相比, 城市空气污染预报模式对ＳＯ２和ＮＯｘ的日平均浓度的预报效果较好, 预报准确率达６４０％ , 而对ＴＳＰ的预报则较差     

南宁市空气臭氧污染的TAR模型预报研究

以门限自回归统计方法为基础进行南宁城市臭氧日最大小时浓度TAR模型预报研究.选取日平均气压作为参与预报气象因子、180日作为建模资料长度、开方形式的数据转换后建立模型预报效果更佳.预报结果检验表明预报方案是可行的,并提出了改进建议.     

夏旱期间人工增雨有利天气形势短期预报概模

利用常规资料,通过对有利于人工增雨作业天气型中的两类优势型(T、R型)进行分析发现:在台风型(T)情况下,有利于人工增雨作业的前一天形势是500 hPa副热带高压呈带状分布,脊线位于30°N以北,西脊点在110°E或以西,中低层在120°E附近存在台风倒槽(或存在低压环流),地面处东南气流中;在弱流场型(R)下,其前一天天气形势是中低层有足够水汽输入条件或存在中尺度低值系统,对应地面存在静止锋或为均压场。以此为依据建立了短期(24 h)预报概模。结果表明:预报概模在实际业务应用中效果尚好。     

北半球夏季环流的简化模式季节预报试验

本文在对地气耦合非定常距平模式做了一些改进的基础上，进行了北半球夏季（6～8月）环流的季节预报试验。 对1982～1989年 6～8 月预报结果分析表明:预报效果优于持续性预报；但模式预报存在着明显的系统误差；距平中心向低纬度的漂移严重。用3种不同方法做了消除系统误差的试验，比较了它们的优劣。此外，针对距平模式预报的是月平均距平量这一特点，提出了两种集合预报方案:一种是用滑动的平均量做出多个5月份初始场；另一种是采用不同长度的时段平均，替代5月份初始场；用上述两种方案做了1987年6～8月的集合预报试验，结果表明优于或接近于非集合的单个预报的结果。     

青藏高原夏季臭氧低谷形成的机理-臭氧输送和化学过程

利用三维化学输送模式(OSLO CTM2)模拟青藏高原夏季臭氧低谷。结果表明：在青藏高原夏季臭氧低谷的形成和变化过程中，动力输送过程起着最主要作用，化学过程部分补偿了输送过程引起的臭氧减少。在动力输送过程中，水平输送在5月份是造成臭氧减少的主要原因，可在6月和7月成为使臭氧增加；垂直平流的作用不断增强，在6月和7月成为臭氧减少的主要因素；对流输送的作用在7月份大幅增加，其引起的臭氧减少可以与净的变化相比，其作用也不可忽视。气相的化学过程引起的臭氧增加的量值有时超过了臭氧的净变化的大小，因此它也起着重要作用。     

Impact of Air Transport on Seasonal Variations and Trends of Surface Ozone at Kislovodsk High Mountain Station

The impact of air transport on the surface ozone variations is analyzed at Kislovodsk High Mountain Station for the period 1989–1996 on the basis of 2D back trajectories. It was shown that the contribution of photochemical and dynamical processes is different for the different seasons. In summer months the surface ozone concentration is governed by photochemical ozone production in semi polluted air from the regions of Northern Caspian, Southern Ural and Volga region. Time of the seasonal ozone maximum appearance is defined by joint influence of the processes of photochemical production and destruction in the eastern sectors and advection from Ukraine and Central Europe. The value of the seasonal minimum is determined by the processes of ozone destruction in the air coming from northeastern direction in the stable frontal zone. Distribution of sectors of the air transport changes from year to year and it can partly explain strong negative trend of the surface ozone concentration at the site.     

Budget Calculations for Ozone and Its Precursors: Seasonal and Episodic Features Based on Model Simulations

Results from two air quality models (LOTOS, EURAD) have been used toanalyse the contribution of the different terms in the continuity equationto the budget of ozone, NO_x and PAN. Both models cover largeparts of Europe and describe the processes relevant for troposphericchemistry and dynamics. One of the models is designed to simulate episodesin the order of 1–2 weeks (EURAD), the other is focussing on theseasonal scale (LOTOS). Based on EURAD simulations it is found that theatmospheric boundary layer (ABL) in Central Europe during a summer-smogepisode in 1990 acts as a source of ozone, which is partly exported from theproduction region in Central Europe. About 40% of the ozone producedchemically in the ABL is lost from Central Europe due to net transport(large-scale and turbulent), 40% are deposited within the domain. Vertical mass exchange of ozone is dominated by the prevailing subsidenceand averaged vertical mass fluxes are directed downward. Averaged massfluxes of PAN, which has no stratospheric source, are upward in the upperpart of the ABL. The results from LOTOS are discussed for the same episodeand for a two month period (July/August 1990). The budget calculation showlarger chemical production for the LOTOS model compared to EURAD. Therelative importance of deposition and net transport, however, is in the sameorder. Differences between the two-month calculation and the one weekepisode are only important for Western Europe where the chemical production is enhanced by 30% during the summer-smog episode. The dependence ofthe results on initial and boundary values is discussed for ozone on thebasis of a simple sensitivity study with EURAD where ozone in the FT is setto 10 ppb initially. This leads to a reversal in the direction of averagedozone mass fluxes in the upper part of the ABL.     

Formulation and Evaluation of IMS, an Interactive Three-Dimensional Tropospheric Chemical Transport Model 2. Model Chemistry and Comparison of Modelled CH4, CO, and O3 with Surface Measurements

In part two of this series of papers on the IMS model, we present the chemistry reaction mechanism usedand compare modelled CH₄, CO, and O₃ witha dataset of annual surface measurements. The modelled monthly and 24-hour mean tropospheric OH concentrationsrange between 5–22 × 10⁵ moleculescm^–3, indicating an annualaveraged OH concentration of about 10 × 10⁵ moleculescm^–3. This valueis close to the estimated 9.7 ± 0.6 × 10⁵ moleculescm^–3 calculated fromthe reaction of CH₃CCl₃ with OH radicals.Comparison with CH₄ generally shows good agreementbetween model and measurements, except for the site at Barrow where modelledwetland emission in the summer could be a factor 3 too high.For CO, the pronounced seasonality shown in the measurements is generally reproduced by the model; however, the modelled concentrations are lower thanthe measurements. This discrepancy may due to lower the CO emission,especially from biomass burning,used in the model compared with other studies.For O₃, good agreement between the model and measurements is seenat locations which are away from industrial regions. The maximum discrepancies between modelled results and measurementsat tropical and remote marine sites is about 5–10 ppbv,while the discrepancies canexceed 30 ppbv in the industrial regions.Comparisons in rural areas at European and American continental sites arehighly influenced by the local photochemicalproduction, which is difficult to model with a coarse global CTM.The very large variations of O₃ at these locations vary from about15–25 ppbv in Januaryto 55–65 ppbv in July–August. The observed annual O₃amplitude isabout 40 ppbv compared with about 20 ppbv in the model. An overall comparison of modelled O₃ with measurements shows thatthe O₃seasonal surface cycle is generally governed bythe relative importance of two key mechanisms that drivea springtime ozone maximum and asummertime ozone maximum.     

一维积云模式计算降雹最大尺度

采用一维积云模式计算积云的发展和冰雹的增长，根据当天０７时的探空资料，用模式计算积云的发展，给出云顶高度，云顶温度，云底高度，云中０℃层高度，云内、外温度分布，以及云内上升气流速度和含水量的分布；再由设定计算冰雹的增长，求出雹云中生成的最大冰雹尺度，考虑０℃层以下冰雹的融化，计算出降到地面上的最大冰雹尺度。与实际观测的最大降雹尺度比较，发现计算结果较好，由此得出结论，用此模式预报最在尺度是可行的，     

Ozone in ambient air at a tropical megacity, Delhi: characteristics, trends and cumulative ozone exposure indices

Seven year data of hourly surface ozone concentration is analyzed to study diurnal cycle, trends, excess of ozone levels above threshold value and cumulative ozone exposure indices at a tropical megacity, Delhi. The ozone levels clearly exhibit a diurnal cycle, similar to what has been found in other urban places. A sharp increase in the ozone levels during forenoon and a sharp decrease in the early afternoon can be observed. The average rate of increase in ozone concentration between 09 and 12 h has been observed to be 7.1 ppb h⁻¹. We find that the daily maximum and daytime 8-h (10–17 h) ozone levels are increasing at a rate of about 1.7 (± 0.7) and 1.3 (± 0.56) ppb y⁻¹, respectively. The directives on ozone pollution in ambient air provided by United Nations Economic Commission for Europe and World Health Organization for vegetation (AOT40) and human health protection were used to assess the air quality. The present surface ozone levels in the city are high enough to exceed “Critical Levels” which are considered to be safe for human health, vegetation and forest. The human health threshold was exceeded for up to ~45 days per year. The AOT40 (Accumulated exposure Over a Threshold of 40 ppb) threshold was exceeded significantly during winter (D-J-F) and pre-monsoon (M-A-M) (Rabi crop growing season) season in India. Translating AOT40 exceedances during pre-monsoon into relative yield loss we estimate yield loss of 22.7%, 22.5%, 16.3% and 5.5% for wheat, cotton, soybean and rice, respectively.     

中国的PM2.5和对流层臭氧及污染物排放控制对策的综合分析

本文首先对中国PM2.5和近地面臭氧浓度的观测进行了简要的综述;并利用2010-2013年全球对流层臭氧的卫星观测数据给出了对流层臭氧浓度在全球和中国地区的分布特征,其平均值分别为29.78 DU和33.97 DU。然后,利用一个气溶胶大气化学-全球气候双向耦合模式模拟了中国地区PM2.5的浓度分布和季节变化,其年平均值为0.51×10^-8 kg/m³。在此基础上又分析了5种典型气溶胶对PM2.5总浓度在不同季节的贡献。结合IPCC第五次评估报告（AR5）,讨论了气溶胶和温室气体及其前体物的排放与辐射强迫的联系,以及减排大气臭氧前体物和气溶胶颗粒物质（PM）对气候变化的可能影响。指出减排臭氧前体物对气候的影响还不完全清楚,对短寿命的温室气体和黑碳气溶胶的减排是一种短期（未来50年）的辅助措施;为了保证全球平均温度增长不超过2℃,减少二氧化碳的排放仍是我们需要坚持的长期战略。短期和长期的减排战略对于保护环境和减缓气候变化都是至关重要的。     

北京秋季地面O3的一维模式模拟研究

用一维光化学模式,基本模拟出静稳天气条件下2001年9月9日北京几个主要大气污染物种:O3、NO、NO2、CO以及SO2的日变化特点,并解释了影响O3及NO、NO2、CO和SO2日变化的控制因子作用。初步分析认为,地面O3对非甲烷碳氢化合物(NMHC)的变化很敏感。NMHC的增加或减少,将会造成O3的明显改变。造成9月9日夜间20:00~23:00一次污染物NO、CO和SO2浓度急剧升高的原因是由于夜间大气层结稳定情况下,大气的垂直扩散减弱,污染物在底层积累造成的。由于实际大气中,影响O3生成和损耗的机制很复杂,同时大气的平流输送是影响O3及其他污染物分布的重要因素,用一维模式虽然能够揭示出影响O3变化的几个因子的作用,但要深入分析O3产生和消耗机制,还需要用三维模式。     

A Comparison of Match Ozonesonde-Derived and 3D Model Ozone Loss Rates in the Arctic Polar Vortex during the Winters of 1994/95 and 1995/96

Ozone loss rates from ozonesonde data reported in the Match experiments of winters 1994/95 and 1995/96 inside the Arctic polar vortex are compared with simulations of the same winters performed using the SLIMCAT 3D chemistry and transport model. For 1994/95 SLIMCAT reproduces the location and timing of the diagnosed ozone destruction, reaching 10 ppbv/sunlit hour in late January as observed. SLIMCAT underestimates the loss rates observed in February and March by 1–3 ppbv/sunlit hour. By the end of March, SLIMCAT ozone exceeds the observations by 25–35%. In January 1995 the ozonesonde-derived loss rates at levels above 525 K are not chemical in origin but due to poor conservation of air parcels. Correcting temperature biases in the model forcing data significantly improved the agreement between the model and observed ozone at the end of winter 1994/95, increasing ozone destruction in SLIMCAT in February and March. The SLIMCAT simulation of winter 1995/96 does not reproduce the maximum ozone loss rates diagnosed by Match of 13 ppbv/sunlit hour. Comparing the data for the two winters reveals that the SLIMCAT photochemistry is least able to reproduce observed losses at low temperatures or when low temperatures coincide with high solar zenith angles (SZA). When cold (T = 192 K), high SZA (90°)matches are excluded from the 1995/96 analysis, agreement between the diagnoses and SLIMCAT is better with ozone loss rates of up to 6 ppbv/sunlit hour. For the rest of the winter SLIMCAT consistently underestimates the Match rates of ozone loss by 1–3 ppbv/sunlit hour. In March 1996 the monthly mean SLIMCAT ozone is 50% greater than observations at 430–540 K. In both winters, ozone destruction rates peaked more rapidly and declined more slowly in the Match observations than in the SLIMCAT simulations. The differences between the observed and modelled cumulative ozone losses demonstrate that the total ozone destruction by the end of the winter is sensitive to errors in the instantaneous ozone loss rates of 1–3 ppbv/sunlit hour.     

Formation of the Summertime Ozone Valley over the Tibetan Plateau: The Asian Summer Monsoon and Air Column Variations

The summertime ozone valley over the Tibetan Plateau is formed by two influences,the Asian summer monsoon(ASM) and air column variations.Total ozone over the Tibetan Plateau in summer was ～33 Dobson units(DU) lower than zonal mean values over the ocean at the same latitudes during the study period 2005-2009.Satellite observations of ozone profiles show that ozone concentrations over the ASM region have lower values in the upper troposphere and lower stratosphere(UTLS) than over the non-ASM region.This is caused by frequent convective transport of low-ozone air from the lower troposphere to the UTLS region combined with trapping by the South Asian High.This offset contributes to a ～20-DU deficit in the ozone column over the ASM region.In addition,along the same latitude,total ozone changes identically with variations of the terrain height,showing a high correlation with terrain heights over the ASM region,which includes both the Tibetan and Iranian plateaus.This is confirmed by the fact that the Tibetan and Iranian plateaus have very similar vertical distributions of ozone in the UTLS,but they have different terrain heights and different total-column ozone levels.These two factors(lower UTLS ozone and higher terrain height) imply 40 DU in the lower-ozone column,but the Tibetan Plateau ozone column is only ～33 DU lower than that over the non-ASM region.This fact suggests that the lower troposphere has higher ozone concentrations over the ASM region than elsewhere at the same latitude,contributing ～7 DU of total ozone,which is consistent with ozonesonde and satellite observations.     

Long Range Forecasting Indian Summer Monsoon Rainfallby a Hybrid Principal Component Neural Network Model

Summary  The existing methods based on statistical techniques for long range forecasts of Indian monsoon rainfall have shown reasonably accurate performance, for last 11 years. Because of the limitation of such statistical techniques, new techniques may have to be tried to obtain better results. In this paper, we discuss the results of an artificial neural network model by combining two different neural networks, one explaining assumed deterministic dynamics within the time series of Indian monsoon rainfall (Model I) and other using eight regional and global predictors (Model II). The model I has been developed by using the data of past 50 years (1901–50) and the data for recent period (1951–97) has been used for verification. The model II has been developed by using the 30 year (1958–87) data and the verification of this model has been carried out using the independent data of 10 year period (1988–97). In model II, instead of using eight parameters directly as inputs, we have carried out Principal Component Analysis (PCA) of the eight parameters with 30 years of data, 1958–87, and the first five principal components are included as input parameters. By combining model I and model II, a hybrid principal component neural network model (Model III) has been developed by using 30 year (1958–87) data as training period and recent 10 year period (1988–97) as verification period. Performance of the hybrid model (Model III) has been found the best among all three models developed. Rootmean square error (RMSE) of this hybrid model during the independent period (1988–97) is 4.93% as against 6.83%of the operational forecasts of the India Meteorological Department (IMD) using the 16 parameter Power Regression model. As this hybrid model is showing good results, it is now used by the IMD for experimental long-range forecasts of summer monsoon rainfall over India as a whole. Received August 20, 1998/Revised April 20, 1999     

The Australian Air Quality Forecasting System: Exploring First Steps Towards Determining The Limits of Predictability For Short-Term Ozone Forecasting

Physical parameterisations of turbulent transfer processes in the atmospheric boundary layer, such as the stability parameterisations developed by Joost Businger, and recent advances in computing capabilities, have been important factors leading to the emergence of operational, numerical air quality forecasting systems. The present paper investigates the performance of the Australian Air Quality Forecasting System (AAQFS) in forecasting the peak 1 h ozone for the current or next day. These 24/36 h forecasts are generated for the Sydney and Melbourne regions and issued twice daily. Quantitative evidence is presented of the potential for the AAQFS to provide accurate numerical air quality forecasts. A second goal is to provide an initial benchmark for investigating the limits of predictability for air quality in the Sydney and Melbourne regions by looking at the dependence of the forecasts on the domain spatial scale (while maintaining the same model grid resolution), the starting time and length of the forecast (0000 UTC starts are 36-h forecasts and 1200 UTC starts are 24-h forecasts), and the sophistication of the photochemical mechanism (simple chemistry, Generic Reaction Set (GRS) and complex chemistry, Carbon Bond IV (CBIV)). The probability of detection by the forecast model is much better than persistence, showing considerable skill. The normalised bias, in general, decreases going from regional scale to sub-regional scale and becomes negative at the station scale. In Melbourne the gross error increases as the domain spatial scale decreases, but in Sydney there is a dip in the error at the sub-regional scale due to a sampling artifact. Better results are obtained at the smaller domain scales for 1200 UTC forecasts in Sydney. These are attributed to the shorter forecast period and secondarily to greater model spin-up effects at 0000 UTC. In Melbourne the results are ambiguous. Similar conclusions are derived from scatter plots of forecasts versus observations. Dividing the scatter plots into four sections by plotting vertical and horizontal lines (at 60 ppb) forms contingency tables for categorical forecasting. These plots show the increase in missed forecasts due to underprediction and the decrease in the number of extreme events detected as the spatial scale decreases. A comparison of the highly condensed GRS photochemical mechanism with the comprehensive CBIV mechanism indicates that, in general, GRS performs well for predicting ozone in urban situations provided that the background concentrations are appropriately specified. The potential to improve the forecasts at the smaller spatial scales, particularly for extreme events at high ozone concentrations, may require moving to a more complex mechanism as computer resources become available. This paper is dedicated to Joost Businger, who had strong ties with the atmospheric boundary-layer community in Australia over the past 40 years. It was while visiting CSIRO in Aspendale, Victoria, in 1965–1966 that Joost determined the stability dependence of the Monin-Obukhov surface-layer profiles. He immediately walked over to Arch Dyer’s office to show Arch his results. Arch carefully examined them, and then opened his desk drawer and pulled out his own plots of the stability dependence that he and Bruce Hicks had obtained. They showed the same curves, and thus the Businger-Dyer-Hicks stability functions were born. Arch and Bruce at the time were struggling with how to handle the internal politics; they needed Bill Swinbank’s approval, as Assistant Divisional Chief, before they could submit their results for publication. Bill had his own very strong ideas that conflicted with observations. Joost’s independent confirmation of their results provided a way forward (Bruce Hicks, personal communication, 2003). Joost has contributed significantly, either directly or indirectly, to experimental field programs both within Australia and overseas and to the development of parameterisations of turbulent transfer processes in the boundary layer.     

基于注意力机制与自适应时序分解的气温预报模型

针对传统方法在捕捉气象序列长期依赖关系及泛化性能上的不足,提出了一种基于稀疏注意力与自适应时序分解的气温预报模型（ATFSAS）。该模型整体采用编码器 解码器架构,结合稀疏注意力机制以有效捕捉气象观测数据间的长期依赖性。为减少编码过程中造成的冗余,提出了一种信息蒸馏方法。通过结合多层解码器与自适应时序分解单元,逐步细化预报信号中的周期性和趋势性分量,实现了较为精准的气温预报。基于德国耶拿气象数据集,进行24 h精细化气温预报,其平均绝对误差为1.7108℃。基于中国地面气候资料日值数据集,进行中短期日平均气温预报和多地区单日平均气温预报,相比传统模型LSTM,ATFSAS模型预报结果的平均绝对误差分别提升了35.56%和23.66%。     

Laboratory Studies of Bromide Oxidation in the Presence of Ozone: Evidence for a Glass-Surface Mediated Reaction

The reaction of sodium bromide particles in the presence of ozone was studied in a flow system both under dark conditions and with 254 nm radiation. We found that there was significant formation of gaseous bromine (probably Br₂) in the presence of ozone in the dark, and that bromide deposited to the walls of the Pyrex reaction flask was its source. The observed rate of gaseous bromine formation in these experiments was approximately 100–1000 times faster than expected based on the knownrate constant for aqueous reaction of bromide with ozone. While the mechanism responsible for this enhanced reactivity was not identified, based on previous reports we suggest that the glass surface converted ozone to more reactive species, such as hydroxyl radical, which in turn oxidized bromide. In the presence of 254 nm radiation, rates of gaseous bromine collection were further enhanced, likely as a result of increased radical production in the system, and wall-deposited bromide was also the source of the gaseous bromine. In these `light' experiments, there was a significant decline in ozone mixing ratios, consistent with bromine radical chemistry. These results suggest the possibility that ozone reacting with internally mixed silicate/sea-salt particles might be a significant mechanism for the oxidation of particulate halides, and subsequent release of photoactive halogen species, in the marine boundary layer.