Estimates of ozone AOT40 from passive sampling in forest sites in South-Western Europe

Weekly-fortnightly ozone (O3) concentrations measured by passive sampling at 81 forest monitoring plots in France, Italy, Spain and Switzerland over the period 2000-2002 were used to estimate the cumulative exposure index AOT40. The estimation method is based on a deterministic model which describes the O3 daily profile as a function of relative altitude (the difference between the altitude of the site and the lowest altitude within a 5 km radius) and the time of the day. Estimated AOT40 values (AOT40(e)) were evaluated against co-located automatic measurement stations and with 14 independent automatic stations located throughout Italy whose weekly mean O3 values were used to simulate passive samplers. AOT40 can be predicted by modelling passive sampling data (R2: 0.90; P<0.0001, SE of estimates: 3271 ppb h), although considerable deviations can occur for individual sites. Estimated AOT40 shows a distinct, significant latitudinal and altitudinal gradient. Taking the 3-year average as a whole, exceedance of critical level of 5000 ppb h occurs at 77-100% of the monitored sites, respectively.     

Analyses of ozone in urban and rural sites in Málaga (Spain)

Ozone concentrations were measured at two (urban and a rural) sites near the city of Málaga (Spain). The aim of this study was to determine the daily, monthly and seasonal variation patterns of ozone concentrations at both sites and to study the possible regional influences. The daily variations mostly have the usual features with the afternoon maximum and the night minimum being more pronounced in the urban area. The average monthly concentrations throughout the year start to increase in March reaching their maximum values in July for the urban site. However, in the rural area, the monthly variations are smaller reaching their maximum value in June. The hourly evolution of the ozone concentrations in both sampling sites is well defined in spring and summer and not so well defined in autumn and winter. Taking into account the four seasons, the rural concentrations are higher than the urban ones. Summer is the season when there are similar concentrations at both sampling sites. Average hourly summer afternoon ozone for the hours 12:00-20:00 LST exceeded the 110 microg m(-3) European Union guidelines for human health for 8 h ozone exposure at the urban and rural sites.     

Stochastic model to forecast ground-level ozone concentration at urban and rural areas

Stochastic models that estimate the ground-level ozone concentrations in air at an urban and rural sampling points in South-eastern Spain have been developed. Studies of temporal series of data, spectral analyses of temporal series and ARIMA models have been used. The ARIMA model (1,0,0) x (1,0,1)24 satisfactorily predicts hourly ozone concentrations in the urban area. The ARIMA (2,1,1) x (0,1,1)24 has been developed for the rural area. In both sampling points, predictions of hourly ozone concentrations agree reasonably well with measured values. However, the prediction of hourly ozone concentrations in the rural point appears to be better than that of the urban point. The performance of ARIMA models suggests that this kind of modelling can be suitable for ozone concentrations forecasting.     

Formaldehyde personal exposure measurements and time weighted exposure estimates in children

Residential concentrations of formaldehyde have been associated with poor respiratory health in children, where formaldehyde has been measured using stationary monitors inside homes. Although children spend most of their time indoors at home, there are few studies of children’s personal exposure to formaldehyde. The aim of this study was to investigate the relationship between personal exposure formaldehyde concentrations, microenvironmental concentrations and time weighted exposure estimates in children.Forty-one primary school children (aged between 9 and 12 years) wore a personal passive sampler over two 24 h periods in two seasons and completed 24 h daily activity diaries and a questionnaire about lifestyle and behaviour. Samplers were co located indoors at home, outdoors at centralised locations and indoors at school for the corresponding period.Personal exposure formaldehyde concentrations in this group of children were generally low with a geometric mean concentration of 9.1 ppb (range <detection limit to 27.3 ppb). There were strong correlations between personal exposure concentrations and both domestic indoor (r_s = .779, p < 0.001) and time weighted estimated (r_s = .802, p < 0.001) concentrations. The time weighted model did not improve the estimate of personal exposure compared with stationary indoor concentrations. Indoor air concentration measured with a single stationary monitor was a suitable surrogate for personal exposure.     

Evaluation of non-methane hydrocarbon (NMHC) emissions based on an ambient air measurement in Tokyo area,Japan

Non-methane hydrocarbons (NMHCs) are known to have an important role on air quality due to their high reactivity. NMHC analysis has been performed on 148 ambient air samples collected at five different sites in the Kanto area (Tokyo metropolitan area and surrounding six prefectures) of Japan in summer and winter of 2008, and fifty NMHCs have been determined and quantified. A field measurement campaign has been conducted at one of the busiest intersections in Tokyo metropolitan area in winter of 2008. NMHC emissions are evaluated through comparison of distributions of individual NMHCs emitted from motor vehicles, which are estimated from the measurements, with those determined from the current emissions inventory. The comparison revealed that the measured distributions of acetylene, ethylene and toluene showed a good agreement with those estimated from the emissions inventory (the values estimated from the measurements are a factor of 1.5, 0.56 and 2.3 larger than the emissions inventory in median, respectively), however, propane and isobutane are found to be significantly underestimated in the emissions inventory (the measured values were a factor of 18 and 5.1 larger than the emissions inventory, respectively). The significant underestimate of propane can be explained by that the current emissions inventory does not consider emissions from liquefied propane gas (LPG) fueled vehicles. However, for isobutane, reasons for the underestimate are still unclear. Another field measurement has been conducted in summer of 2008, where the air samples have been collected at three different sites on the ground and by a helicopter as well. Remarkable high concentrations of 1-butene and cis- and trans-2-butenes have been sporadically observed in the samples collected at Urayasu in the coastal area of Tokyo bay. Calculated propylene equivalent (PE) concentrations of butenes revealed that those have a significantly important role in ozone formation when the air plume is affected by emissions from their emission sources. The PE concentrations of butenes varied from 0.1 to 39 ppbC, and accounted for 1.5–75% of total PE concentrations at Urayasu. Most of the continuous air quality monitoring stations does not record concentrations of individual hydrocarbons, therefore, the importance of reactive and low concentration hydrocarbon such as butenes might be overlooked in the current emissions inventory and/or air quality model. In this paper, the reliability of NMHC emissions is evaluated based on the field measurements. Their possible impacts on air quality in the Kanto area are discussed as well, based on the calculated propylene equivalent concentrations.     

Verification of a source-oriented externally mixed air quality model during a severe photochemical smog episode

The CIT/UCD three-dimensional source-oriented externally mixed air quality model is tested during a severe photochemical smog episode (Los Angeles, 7–9 September 1993) using two different chemical mechanisms that describe the formation of ozone and secondary reaction products. The first chemical mechanism is the secondary organic aerosol mechanism (SOAM) that is based on SAPRC90 with extensions to describe the formation of condensable organic products. The second chemical mechanism is the caltech atmospheric chemistry mechanism (CACM) that is based on SAPRC99 with more detailed treatment of organic oxidation products.The predicted ozone concentrations from the CIT/UCD/SOAM and the CIT/UCD/CACM models agree well with the observations made at most monitoring sites with a mean normalized error of approximately 0.4–0.5. Good agreement is generally found between the predicted and measured NO_x concentrations except during morning rush hours of 6–10 am when NO_x concentrations are under-predicted at most locations. Total VOC concentrations predicted by the two chemical mechanisms agree reasonably well with the observations at three of the four sites where measurements were made. Gas-phase concentrations of phenolic compounds and benzaldehyde predicted by the UCD/CIT/CACM model are higher than the measured concentrations whereas the predicted concentrations of other aromatic compounds approximately agree with the measured values.The fine airborne particulate matter mass concentrations (PM_2.5) predicted by the UCD/CIT/SOAM and UCD/CIT/CACM models are slightly greater than the observed values during evening hours and lower than observed values during morning rush hours. The evening over-predictions are driven by an excess of nitrate, ammonium ion and sulfate. The UCD/CIT/CACM model predicts higher nighttime concentrations of gaseous precursors leading to the formation of particulate nitrate than the UCD/CIT/SOAM model. Elemental carbon and total organic mass are under-predicted by both models during morning rush hour periods. When this latter finding is combined with the NO_x under-predictions that occur at the same time, it suggests a systematic bias in the diesel engine emissions inventory. The mass of particulate total organic carbon is under-predicted by both the UCD/CIT/SOAM and UCD/CIT/CACM models during afternoon hours. Elemental carbon concentrations generally agree with the observations at this time. Both the UCD/CIT/SOAM and UCD/CIT/CACM models predict low concentrations of secondary organic aerosol (SOA) (<3.5 μg m⁻³) indicating that both models could be missing SOA formation pathways. The representation of the aerosol as an internal mixture vs. a source-oriented external mixture did not significantly affect the predicted concentrations during the current study.     

Variability of ozone concentration in a montane environment,northern Italy

To evaluate the spatial variability of ozone concentrations, two studies were undertaken in the montane environment of Trentino region, northern Italy, in 2007. In the first study, a 225 km² area was considered. Here, a randomized design was used to evaluate the variability of ozone concentration at 1 and 225 km² scale. Measurements were carried out by passive samplers between May and June 2007. In a second study, the whole 6207 km² area of Trentino was considered. The area is covered by five grid cells of the European Monitoring and Evaluation Programme (EMEP). A systematic 15 × 15 km grid was used to allocate 15 passive samplers over the entire province, resulting into 1–4 samplers for each of the 5 EMEP grid cells (2500 km² each) overlapping the study area. Measurements were carried out between June and September 2007. Accuracy of passive samplers was checked by direct comparison with conventional ozone analysers. Significant differences (P = 0.034) were found in ozone concentration among 1 × 1 km quadrates within the 225 km² study area, while variability within the 1 × 1 km grid cells (coefficient of variation, CV′ = 0.12) slightly exceed the measurement error (CV′ = 0.08). At larger scales (225, 2500 and 6207 km²), ozone concentration shows much higher variability (CV′ from 0.18 to 0.28, with peak values at 0.40). Reported differences lead to very different AOT40 estimates even within the same EMEP grid cell. These findings suggest that 1 × 1 km resolution seems appropriate for ozone concentration modelling. On the other hand, significant sub-grid variation may exist at the resolution adopted by the EMEP model. Coupled with the likely variability of other important meteorological, soil and vegetation variables, our findings suggest that ozone risk assessment for vegetation based on large-scale modelled AOT40 and flux needs to be considered with great caution. The evidence reported in this paper asks for more detailed national-scale modelling, and the development of methods to incorporate local scale variations into large-scale models.     

Evaluation of air quality models for the simulation of a high ozone episode in the Houston metropolitan area

A high ozone event in the Houston–Galveston–Brazoria area was utilized to study the shortcomings of the current air quality models. To improve the baseline simulations with the Comprehensive Air quality Model with Extensions (CAMx) for developing the state implementation plan, the Texas Commission on Environmental Quality (TCEQ) imputed emissions of highly reactive volatile organic compounds (HRVOCs) by scaling the amount of fugitive emissions of olefins to co-emitted NO_x from selected point sources, effectively multiplying by 3–12 times over the regular inventory values. In this paper, CAMx and the Community Multiscale Air Quality (CMAQ) model were used to determine if the imputed HRVOC emissions were consistent with the observed atmospheric conditions. With the base emissions, CMAQ and CAMx both with the Carbon-Bond 4 (CB-4) mechanism simulated similar ozone concentrations. But with the imputed HRVOC emissions, CMAQ predicted lower ozone peaks than CAMx in the vicinity and downwind of the Ship Channel and other highly HRVOC-rich areas. Based on analyses of sensitivity simulations of CMAQ with different emission inputs and vertical diffusion algorithms in the model, we found that the modeled atmosphere lacked reactivity to produce the observed high ozone event. Although the imputed HRVOC emissions improved ozone prediction at the surface sites, but the ethylene concentrations were not consistent with the measurements at the super sites (La Porte and Clinton) and by NOAA aircraft. Several sensitivity tests designed to provide additional radicals into the system and other research results suggested that the lack of reactivity may need to be corrected by targeted, and probably of episodic, increase of HRVOC emissions, from the sources in the Houston Ship Channel. Additional investigation of the ozone production efficiency for different chemical mechanisms is necessary to pinpoint the emissions uncertainty issues.     

Indoor Ozone Exposures

Indoor and outdoor ozone concentrations were measured from late May through October at three office buildings with very different ventilation rates. The indoor values closely tracked the outdoor values, and, depending on the ventilation rate, were 20 to 80 percent of those outdoors. The Indoor/outdoor data are adequately described with a mass balance model. The model can also be coupled with reported air exchange rates to estimate indoor/outdoor ratios for other structures. The results from this and previous studies indicate that Indoor concentrations are frequently a significant fraction of outdoor values. These observations, and the fact that most people spend greater than 90 percent of their time indoors, indicate that indoor ozone exposure (concentration × time) is greater than outdoor exposure for many people. Relatively Inexpensive strategies exist to reduce indoor ozone levels, and these could be implemented to reduce the public’s total ozone exposure.     

Filtration of Radioactive Aerosols By Glass Fibers

Abstract

An ozone (O₃) exposure assessment study was conducted in Toronto, Ontario, Canada during the winter and summer of 1992. A new passive O₃ sampler developed by Harvard was used to measure indoor, outdoor, and personal O₃ concentrations. Measurements were taken weekly and daily during the winter and summer, respectively. Indoor samples were collected at a total of 50 homes and workplaces of study participants. Outdoor O₃ concentrations were measured both at home sites using the passive sampler and at 20 ambient monitoring sites with continuous monitors. Personal O₃ measurements were collected from 123 participants, who also completed detailed time-activity diaries. A total of 2,274 O₃ samples were collected. In addition, weekly air exchange rates of homes were measured.

This study demonstrates the performance of our O₃ sampler for exposure assessment. The data obtained are further used to examine the relationships between personal, indoor (home and workplace), and outdoor O₃ concentrations, and to investigate outdoor and indoor spatial variations in O₃ concentrations. Based on home outdoor and indoor, workplace, and ambient O₃ concentrations measured at the Ontario Ministry of the Environment (MOE) sites, the traditional microenvironmental model predicts 72% of the variability in measured personal exposures. An alternative personal O₃ exposure model based on outdoor measurements and time-activity information is able to predict the mean personal exposures in a large population, with the highest R² value of 0.41.     

Nonparametric functional data estimation applied to ozone data: prediction and extreme value analysis

The study of extreme values and prediction of ozone data is an important topic of research when dealing with environmental problems. Classical extreme value theory is usually used in air-pollution studies. It consists in fitting a parametric generalised extreme value (GEV) distribution to a data set of extreme values, and using the estimated distribution to compute return levels and other quantities of interest. Here, we propose to estimate these values using nonparametric functional data methods. Functional data analysis is a relatively new statistical methodology that generally deals with data consisting of curves or multi-dimensional variables. In this paper, we use this technique, jointly with nonparametric curve estimation, to provide alternatives to the usual parametric statistical tools. The nonparametric estimators are applied to real samples of maximum ozone values obtained from several monitoring stations belonging to the Automatic Urban and Rural Network (AURN) in the UK. The results show that nonparametric estimators work satisfactorily, outperforming the behaviour of classical parametric estimators. Functional data analysis is also used to predict stratospheric ozone concentrations. We show an application, using the data set of mean monthly ozone concentrations in Arosa, Switzerland, and the results are compared with those obtained by classical time series (ARIMA) analysis.     

The characterization of ozone concentrations at a select set of high-elevation sites in the eastern United States

Hourly averaged data for ozone collected in 1986 and 1987 were analyzed and characterized for a select set of high-elevation sites in the eastern United States. Pressure-corrected adjustments may be necessary when comparing ozone concentrations measured at two different elevations. When unadjusted concentrations (i.e. in units of parts per million) were used, the Whiteface Mountain sites showed what appeared to be an ozone elevational gradient. A gradient was not observed for the two MCCP Shenandoah National Park sites (SH1 and SH2). When adjusted ozone values (i.e. in units of micrograms per cubic meter) were used, the elevational gradient reported for Whiteface Mountain was no longer observed. When unadjusted concentrations were used, in most cases the high-elevation sites appeared to be receiving greater ozone exposure than the nearby, lower elevation sites. When adjusted ozone values were used, a consistent conclusion was not evident. On a regional basis for the period May through September 1987, when unadjusted concentrations were used, the high-elevation sites in the South appeared to experience higher cumulative ozone exposures than sites in the North. When adjusted ozone values were used, the geographic gradient was not strong. Assuming that target sensitivity remains nearly constant as elevation changes, adjusted concentrations should be taken into consideration when evaluating the relationship between ozone exposures at high-elevation sites and biological effects.     

Distribution of ozone and other air pollutants in forests of the Carpathian Mountains in central Europe.

Ozone (O3) concentrations were monitored during the 1997-1999 growing seasons in 32 forest sites of the Carpathian Mountains. At all sites (elevation between 450 and 1320 m) concentrations of O3, nitrogen dioxide (NO2), and sulfur dioxide (SO2) were measured with passive samplers. In addition, in two western Carpathian locations, Vychodna and Gubalówka, ozone was continuously monitored with ultraviolet (UV) absorption monitors. Highest average hourly O3 concentrations in the Vychodna and Guba?ówka sites reached 160 and 200 microg/m3 (82 and 102 ppb), respectively (except for the AOT40 values, ozone concentrations are presented as microg/m3; and at 25 degrees C and 760 mm Hg, 1 microg O3/m3 = 0.51 ppb O3). These sites showed drastically different patterns of diurnal 03 distribution, one with clearly defined peaks in the afternoon and lowest values in the morning, the other with flat patterns during the entire 24-h period. On two elevational transects, no effect of elevation on O3 levels was seen on the first one, while on the other a significant increase of O3 levels with elevation occurred. Concentrations of O3 determined with passive samplers were significantly different between individual monitoring years, monitoring periods, and geographic location of the monitoring sites. Results of passive sampler monitoring showed that high O3 concentrations could be expected in many parts of the Carpathian range, especially in its western part, but also in the eastern and southern ranges. More than four-fold denser network of monitoring sites is required for reliable estimates of O3 distribution in forests over the entire Carpathian range (140 points). Potential phytotoxic effects of O3 on forest trees and understory vegetation are expected on almost the entire territory of the Carpathian Mountains. This assumption is based on estimates of the AOT40 indices for forest trees and natural vegetation. Concentrations of NO2 and SO2 in the entire Carpathian range were typical for this part of Europe and below the expected levels of phytotoxicity.     

Use of passive ambient ozone (O3) samplers in vegetation effects assessment

A stochastistic, Weibull probability model was developed and verified to simulate the underlying frequency distributions of hourly ozone (O3) concentrations (exposure dynamics) using the single, weekly mean values obtained from a passive (sodium nitrite absorbent) sampler. The simulation was based on the data derived from a co-located continuous monitor. Although at the moment the model output may be considered as being specific to the elevation and location of the study site, the results were extremely good. This effort for the approximation of the O3 exposure dynamics can be extended to other sites with similar data sets and in developing a generalized understanding of the stochastic O3 exposure-plant response relationships, conferring measurable benefits to the future use of passive O3 samplers, in the absence of continuous monitoring.     

Formation and Transport of the Madrid Ozone Plume

Abstract

The main results of an experimental study focusing on the formation and transport of photochemical pollution in the Madrid air basin are presented. This southern European, heavily populated urban area is located on an elevated plateau at a height of 700 m, near a mountain range with maximum heights of around 2,400 m. Daily and seasonal cycles of ozone were documented during a one-year survey at three semi-rural sites located 30 km away from the urban center. Maximum hourly values of up to 140 ppb were measured, and the ozone generated within the urban plume on polluted days (when values exceeded 90 ppb) has been estimated at around 40-50 ppb.A meteorological characterization of these smoggy days pointed out the influence of thermally induced local wind flows on the concentration daily cycles at the measuring sites, denoting a preferred advection of the urban plume. Moreover, during intensive summer field campaigns, the use of meteorological and ozone sondes, as well as an instrumented aircraft, revealed some features about the horizontal and vertical distribution of the polluted air masses, as well as their evolution within the planetary boundary layer. Ozone plumes have been detected up to 100 km away from the city, usually mixed in a layer that reaches a height of 1,000-1,500 m in the afternoon. On some occasions, ozone-enriched layers have been detected as high as 4,000 m during morning hours, suggesting possible tropospheric injection induced by topographydriven flows or convective mesoscale systems that are usually present in the center of the Iberian Peninsula in the summer.     

Comparison of the performances of land use regression modelling and dispersion modelling in estimating small-scale variations in long-term air pollution concentrations in a Dutch urban area

The performance of a Land Use Regression (LUR) model and a dispersion model (URBIS – URBis Information System) was compared in a Dutch urban area. For the Rijnmond area, i.e. Rotterdam and surroundings, nitrogen dioxide (NO₂) concentrations for 2001 were estimated for nearly 70 000 centroids of a regular grid of 100 × 100 m.A LUR model based upon measurements carried out on 44 sites from the Dutch national monitoring network and upon Geographic Information System (GIS) predictor variables including traffic intensity, industry, population and residential land use was developed. Interpolation of regional background concentration measurements was used to obtain the regional background. The URBIS system was used to estimate NO₂ concentrations using dispersion modelling. URBIS includes the CAR model (Calculation of Air pollution from Road traffic) to calculate concentrations of air pollutants near urban roads and Gaussian plume models to calculate air pollution levels near motorways and industrial sources. Background concentrations were accounted for using 1 × 1 km maps derived from monitoring and model calculations.Moderate agreement was found between the URBIS and LUR in calculating NO₂ concentrations (R = 0.55). The predictions agreed well for the central part of the concentration distribution but differed substantially for the highest and lowest concentrations. The URBIS dispersion model performed better than the LUR model (R = 0.77 versus R = 0.47 respectively) in the comparison between measured and calculated concentrations on 18 validation sites. Differences can be understood because of the use of different regional background concentrations, inclusion of rather coarse land use category industry as a predictor variable in the LUR model and different treatment of conversion of NO to NO₂.Moderate agreement was found between a dispersion model and a land use regression model in calculating annual average NO₂ concentrations in an area with multiple sources. The dispersion model explained concentrations at validation sites better.     

Passive sampler for PM10-2.5 aerosol

This study investigates the use of a small passive sampler for aerosol particles to determine particulate matter (PM)10-2.5 concentrations in outdoor air. The passive sampler collects particles by gravity, diffusion, and convective diffusion onto a glass coverslip that is then examined with an optical microscope; digital images are processed with free software and the resultant PM10-2.5 concentrations determined. Both the samplers and the analyses are relatively inexpensive. Passive samplers were collocated with Federal Reference Method (FRM) samplers in Chapel Hill, NC; Phoenix, AZ; and Birmingham, AL; for periods from 5 to 15 days. Particles consisted primarily of inorganic dusts at some sites and a mix of industrial and inorganic materials at other sites. Measured concentrations ranged from < 10 microg/m3 to approximately 40 microg/m3. Overall, PM10-2.5 concentrations measured with the passive samplers were within approximately 1 standard deviation of concentrations measured with the FRM samplers. Concentrations determined with passive samplers depend on assumptions about particle density and shape factors and may also depend somewhat on local wind speed and turbulence; accurate values for these parameters may not be known. The degree of agreement between passive and FRM concentrations measured here suggests that passive measurements may not be overly dependent on accurate knowledge of these parameters.     

Contribution to ambient benzene concentrations in the vicinity of petrol stations: Estimation of the associated health risk

This work examines the contribution of petrol stations to the ambient benzene concentrations and attempts to estimate the possible health risks for the people living in the vicinity of such installations.Three monitoring sites (urban, suburban and rural) were used as reference points and the benzene concentrations were recorded at several distances along their perimeter.In order to evaluate the net contribution of the petrol station to the ambient benzene concentrations, the urban background concentration, measured by passive samplers and the contribution of the roads, estimated with both the COPERT and the linear source model CALINE 4, were deduced. Validation and optimization of the modeling system COPERT and CALINE4 was done in advance to ensure the reliability of the results. It seems that petrol stations have a significant contribution to ambient benzene concentrations in their vicinity.Finally, a risk assessment evaluation was attempted in terms of increased cancer risk due to the presence of the petrol stations in an area. The results show remarkable increase of the population risks in the vicinity, ranging from 3% to 21% in comparison to the population in the rest of the town.     

Some aspects of boundary layer evolution in Mexico City

Measurements of chemical species and meteorological parameters were made at a site located 440 m above the mean basin level of Mexico City, over a two-week period in November during Project Azteca. Data from three of the stations of Mexico City's air quality monitoring network (Red Automática de Monitoreo Ambiental, RAMA) were also used to estimate the dilution in concentration experienced by pollutants as they are transported upslope during the course of the day. Both carbon monoxide and sulfur dioxide show a dilution of up to 50%, while ozone is usually more concentrated at the elevated site. These comparisons clearly highlight the intrinsic differences between primary and secondary gases, which are supported also by time–space, cross correlation analysis. The thermal mesoscale wind circulation dominates concentrations of pollutants at the research site: upslope during the day and downslope during the night. The data present clear evidence that downslope flows during the night contribute to ozone concentration at basin sites.