Ensemble and bias-correction techniques for air quality model forecasts of surface O3 and PM2.5 during the TEXAQS-II experiment of 2006

Several air quality forecasting ensembles were created from seven models, running in real-time during the 2006 Texas Air Quality (TEXAQS-II) experiment. These multi-model ensembles incorporated a diverse set of meteorological models, chemical mechanisms, and emission inventories. Evaluation of individual model and ensemble forecasts of surface ozone and particulate matter (PM) was performed using data from 119 EPA AIRNow ozone sites and 38 PM sites during a 50-day period in August and September of 2006. From the original set of models, two new bias-corrected model data sets were built, either by applying a simple running mean average to the past 7 days of data or by a Kalman-Filter approach. From the original and two bias-corrected data sets, three ensembles were created by a simple averaging of the seven models. For further improvements three additional weighted model ensembles were created, where individual model weights were calculated using the singular value decomposition method. All six of the ensembles are compared to the individual models and to each other in terms of root mean square error, correlation, and contingency and probabilistic statistics. In most cases, each of the ensembles show improved skill compared to the best of the individual models. The over all best ensemble technique was found to be the combination of Kalman-Filtering and weighted averaging. PM_2.5 aerosol ensembles demonstrated significant improvement gains, mostly because the original model's skill was very low.     

Long-term trend and spatial pattern of PM2.5 induced premature mortality in China

With rapid economic growth, China has witnessed increasingly frequent and severe haze and smog episodes over the past decade, posing serious health impacts to the Chinese population, especially those in densely populated city clusters. Quantification of the spatial and temporal variation of health impacts attributable to ambient fine particulate matter (PM_2.5) has important implications for China's policies on air pollution control. In this study, we evaluated the spatial distribution of premature deaths in China between 2000 and 2010 attributable to ambient PM_2.5 in accord with the Global Burden of Disease based on a high resolution population density map of China, satellite retrieved PM_2.5 concentrations, and provincial health data. Our results suggest that China's anthropogenic ambient PM_2.5 led to 1,255,400 premature deaths in 2010, 42% higher than the level in 2000. Besides increased PM_2.5 concentration, rapid urbanization has attracted large population migration into the more developed eastern coastal urban areas, intensifying the overall health impact. In addition, our analysis implies that health burdens were exacerbated in some developing inner provinces with high population density (e.g. Henan, Anhui, Sichuan) because of the relocation of more polluting and resource-intensive industries into these regions. In order to avoid such national level environmental inequities, China's regulations on PM_2.5 should not be loosened in inner provinces. Furthermore policies should create incentive mechanisms that can promote transfer of advanced production and emissions control technologies from the coastal regions to the interior regions.     

A comparison of chemical mechanisms based on TRAMP-2006 field data

A comparison of a model using five widely known mechanisms (RACM, CB05, LaRC, SAPRC-99, SAPRC-07, and MCMv3.1) has been conducted based on the TexAQS II Radical and Aerosol Measurement Project (TRAMP-2006) field data in 2006. The concentrations of hydroxyl (OH) and hydroperoxy (HO₂) radicals were calculated by a zero-dimensional box model with each mechanism and then compared with the OH and HO₂ measurements. The OH and HO₂ calculated by the model with different mechanisms show similarities and differences with each other and with the measurements. First, measured OH and HO₂ are generally greater than modeled for all mechanisms, with the median modeled-to-measured ratios ranging from about 0.8 (CB05) to about 0.6 (SAPRC-99). These differences indicate that either measurement errors, the effects of unmeasured species or chemistry errors in the model or the mechanisms, with some errors being independent of the mechanism used. Second, the modeled and measured ratios of HO₂/OH agree when NO is about 1 ppbv, but the modeled ratio is too high when NO was less and too low when NO is more, as seen in previous studies. Third, mechanism–mechanism HO_x differences are sensitive to the environmental conditions – in more polluted conditions, the mechanism–mechanism differences are less. This result suggests that, in polluted conditions, the mechanistic details are less important than in cleaner conditions, probably because of the dominance of reactive nitrogen chemistry under polluted conditions.     

Chemical and optical properties of aerosols and their interrelationship in winter in the megacity Shanghai of China

A field campaign on air quality was carried out in Shanghai in winter of 2012. The concentrations of NO, NO₂, NO_x, SO₂, CO, and PM_2.5 increased during haze formation. The average masses of SO₄^2-, NO₃^- and NH₄⁺ were 10.3, 11.7 and 6.7 μg/m³ during the haze episodes, which exceeded the average (9.2, 7.9, and 3.4 μg/m³) of these components in the non-haze days. The mean values for the aerosol scattering coefficient (b_sp), aerosol absorption coefficient (b_ap) and single scattering albedo (SSA) were 288.7, 27.7 and 0.91 Mm^-1, respectively. A bi-peak distribution was observed for the mass concentrations of CO, NO, NO₂, and NO_x. More sulfate was produced during daytime than that in the evening due to photochemical reactions. The mass concentration of NH₄⁺ achieved a small peak at noontime. NO₃^- showed lower concentrations in the afternoon and higher concentrations in the early morning. There were obvious bi-peak diurnal patterns for bsp and bap as well as SSA. bsp and bap showed a positive correlation with PM_2.5 mass concentration. (NH₄)₂SO₄, NH₄NO₃, organic mass, elemental carbon and coarse mass accounted for 21.7%, 19.3%, 31.0%, 9.3% and 12.3% of the total extinction coefficient during non-haze days, and 25.6%, 24.3%, 30.1%, 8.1% and 8.2% during hazy days. Organic matter was the largest contributor to light extinction. The contribution proportions of ammonium sulfate and ammonium nitrate to light extinction were significantly higher during the hazy time than during the non-haze days.     

Impacts of Russian biomass burning on UK air quality

Unusually high levels of PM₁₀ were observed in the UK in May 2006 and September 2002. This paper investigates the possible contribution of long-range transport of smoke from widespread agricultural burning and forest fires in western Russia to these air pollution episodes. The Lagrangian dispersion model NAME is run in both forwards and backwards modes to determine the transport and sources of the polluted air masses for the two incidents. Comparison of the model results to satellite data and ground observations from across Europe demonstrates good agreement for both the timing and magnitude of the episodes and suggests that fires in western Russia were the primary cause of both incidents. Secondary contributions to the 2006 episode may have come from European anthropogenic pollution and pollen released in northern Europe. The occurrence and timing of both pollution episodes were strongly controlled by the meteorological situation at the time. Scaling of model results to observations suggests that 0.5–0.7 Mtonnes of biomass per day could have been burnt during periods when winds reaching the UK were from the east. The newly reported 2006 episode means that Russian fires have affected UK air quality at least twice since 2000 and it is suggested that, without changes in current practice, such events are likely to occur again in the future with implications for UK and European air quality.     

Monthly Mean Afternoon Mixing-Layer Depths “Tuned” to the Eco-Climatic Regions of the Canadian Prairie Provinces

A simple diagnostic model was used to derive afternoon (0000 UTC) mixing-layer depths from gridded zero hour output from the Canadian Global Environmental Multiscale (GEM) numerical weather model interpolated to approximately 200 locations in the Prairie provinces of Canada. Knowledge of the spatial pattern of monthly mean afternoon mixing-layer depths was enhanced by associating the annual cycles of monthly means with eco-climatic regions - regions with fairly uniform terrain and soils, and where similar vegetation indicate that the area experiences, on average, similar synoptic-scale weather and produces similar surface fluxes of heat and moisture. When the mixing-layer depth is used to assess the dilution of airborne pollutants or to determine the specific humidity of the boundary layer, the eco-climatic region rather than proximity to the location-of-interest should be the primary factor in the selection of an appropriate radiosonde site.     

Urban air pollution modelling and measurements of boundary layer height

An urban field trial has been undertaken with the aim of assessing the performance of the boundary layer height (BLH) determination of two models: the Met Office Unified Model (UM) and a Gaussian-type plume model, ADMS. Pulsed Doppler lidar data were used to measure mixing layer height and cloud base heights for a variety of meteorological conditions over a 3 week period in July 2003. In this work, the daily growth and decay of the BLH from the lidar data and model simulations for 5 days are compared. The results show that although the UM can do a good job of reproducing the boundary layer growth, there are occasions where the BLH is overestimated by 30–100%. Within dispersion models it is the BLH that effectively limits the height to which pollution disperses, so these results have very important implications for pollution dispersion modelling. The results show that correct development of the boundary layer in the UM is critically dependant on morning cloud cover. The ADMS model is used routinely by local authorities in the UK for local air-quality forecasting. The ADMS model was run under three settings; an ‘urban’ roughness, a ‘rural’ roughness and a ‘transition’ roughness. In all cases, the ‘urban’ setting over estimated the BLH and is clearly a poor predictor of urban BLH. The ‘transition’ setting, which distinguishes between the meteorological data input site and the dispersion modelling site, gave the best results under the well mixed conditions of the trial.     

Trend,seasonal and multivariate analysis study of total gaseous mercury data from the Canadian atmospheric mercury measurement network (CAMNet)

Long-term monitoring data of total gaseous mercury (TGM) concentrations from the Canadian Atmospheric Mercury Measurement Network (CAMNet) were analysed for temporal trends, seasonality and comparability within the network and compared to other network and model results. Data collected from 11 Canadian measurement sites between 1995 and 2005 were analysed. Sites within CAMNet were characterized by principle component analysis (PCA) into four main categories. For the first time since automated TGM measurements have been made within CAMNet, this paper reveals statistically significant decreasing TGM concentrations from rural locations in Canada during this time period. The largest declines were observed close to the urban areas of Toronto and Montreal, where levels fell by 17% at Point Petre, and 13% at St. Anicet, respectively. Many of the TGM changes are comparable with the overall trends observed in total mercury concentrations in precipitation, for similar time periods, at co-located or nearby National Atmospheric Deposition programme's Mercury Deposition Network (NADP-MDN) sites. The results show that these changes are mostly driven by local or regional changes in mercury emissions. Other sites within CAMNet reflect reported changes in hemispherical global background concentrations of airborne mercury, where slight decreases or no statistically significant trend in TGM concentrations exist over the same time period.     

Study of relationship between water-soluble Ca2+ and lidar depolarization ratio for spring aerosol in the boundary layer

The backscattering ratio, depolarization ratio and water-soluble ions of aerosols inside the boundary layer were measured by a depolarization lidar and an in situ ion chromatography (IC) in the spring of 2004 and 2005. To study the relationship between depolarization and aerosol compositions, depolarization ratios were compared with mixing ratio of soluble ions like calcium, potassium, and sodium from surface ground measurements. About 70% of the daily maximum depolarization ratios were distributed between 1.5% and 3.5% with a mean value of 2.14±0.83%. High correlation coefficients (R>0.8) were found between depolarization and ion calcium for all of the depolarization episodes (DP≈5–12%), which suggest the existence of mineral dust. As the tracer of biomass burning and sea salt, the mixing ratio of K⁺ and Na⁺ usually are greater than Ca²⁺ but did not show dependences with depolarization implying that biomass burning and sea salt aerosols may transport along with Asian dust but not exhibit significant non-spherical properties.     

Recent decline of methyl bromide in the troposphere

Atmospheric methyl bromide (CH₃Br) measured at a remote ground station in the Arctic (mid-1996 to early 2002) and in the free troposphere at mid-latitude (early 1999 to early 2002) showed a steady annual average decrease of 4–6%. The trend was consistent with a simulation of the response to the phase-out schedule of anthropogenic emissions under the Montreal Protocol and its amendments, suggesting that a decrease in CH₃Br abundance in the Northern Hemisphere of as much as 40% from the level of the early 1990s would be possible by the completion date of the program.