Variability of ozone air quality indicators in selected metropolitan statistical areas

This article analyzes numerical variability in ozone air quality data to understand how this variability affects the number of violations seen each year in metropolitan statistical areas (MSAs). Three commonly cited violation indices are used: 1) the annual number of expected exceedances averaged over 3 years is greater than 1; 2) the n+ 1th hourly value in n years of data is greater than 0.12 ppm; and 3) the annual number of expected exceedances is greater than 1. Only the first index is consistent with applicable regulations. The analyses indicate that about 23 percent of all MSAs with valid data had one or more change in their ozone violation status between 1979 and 1987. This change in status occurred for approximately 7 percent of all MSA-years of available data. This statistic was about one-third of the value usually obtained when the two incorrect, but commonly used, criteria of ozone violations are used.     

Annual and Multi-Year Variability in Ozone Expected Exceedances

Ozone data from 184 metropolitan statistical areas (MSAs) for the time period 1979-1989 were analyzed to: (1) put the large number of O₃ exceedances observed in 1988 into perspective, and (2) determine if increasing the averaging time used to calculate the O₃ NAAdS violation rate would significantly decrease the number of MSAs that experienced a change in their annual attainment status. The analyses indicate that 1988 had more exceedances in more MSAs than any other year, but 1979-1981 had more O₃ NAAQS violations. The analyses also indicate that perceived attainment status flip-flops are significantly reduced by increasing the averaging period from the current three-year period to a longer time period.     

Issues Regarding the Ozone Air Quality Standard and the Design Value

Abstract

Two problems exist in the form and the compliance test of the present National Ambient Air Quality Standard (NAAQS) for ozone. One is the use of the number of exceedances in the form of the standard, which generates confusion and unnecessary complexity when the form is translated to the design value. The other is the requirement of a zero percent chance of violation in the compliance test, which makes the NAAQS considerably more stringent than generally assumed. There are also two sample-size problems in the estimation procedure for the design value. One is the upward creeping of the (n+l)th highest value in n years as n increases from one in the table look-up approach. The other is the infinite-sample-size assumption instead of the number of high-ozone season days per year for the daily maximum ozone concentrations in the distribution fitting approach. Both problems lead to an exaggeration of the design value.

The above problems can be removed in a revised NAAQS by (1) using a statistic that is identical to the design value itself in the form of the standard, (2) defining the design value as an n-year mean of, say, the annual mth highest values rather than the xth highest value in n years, and (3) using a simple compliance test like the t test that compares the design value with the level of the standard, taking into account the year-to-year fluctuation of the annual mth highest values. When the design value of an area is close to the level of the standard, the test provides a natural “too close to call” interval, which adjusts itself with the fluctuation of the annual mth highest values, so that as the fluctuation increases, the ability to assign the compliance status of the area decreases. The inclusion of a “too close to call” interval or category in the standard is critical to reduce the tendency toward ozone attainment flip-flops in areas approaching attainment and to assure that the ozone NAAQS is not more or less stringent than it appears.     

Vertical Ozone Profile in the Lower Troposphere over Chicago

ABSTRACT

A 15-year (1981-95) climatology for the diurnal maximum ozone concentration (DMOC) was developed using 1-hr average ozone concentrations in the Baltimore-Washington area, which was made up of four regions: Baltimore, Washington, non-urban Maryland, and non-urban northern Virginia. The DMOC time series for each of these regions were divided into four terms representing different behavioral time scales: the long-term mean; the mean in-tra-annual perturbation; the interannual perturbation; and the synoptic perturbation. The urban regions had smaller values of the long-term mean ozone, but the annual range was larger. The values of the interannual perturbation were largest in the summer, when ozone production is significant, and smallest in the late winter and early spring. The interannual perturbation in the summer in the four regions consistently had positive departures in 1983, 1988, and 1991, and it had negative departures in 1981, 1984, 1985, 1989, 1990, and 1992. Summers with large positive interannual departures experienced a large number of ozone exceedances (i.e., relative to the 1-hr National Ambient Air Quality Standard of 125 parts per billion [ppb]), and summers with large negative departures experienced few or no exceedances. About 50% of the exceedances had concentrations ranging in value from 125-135 ppb, and about 75% had concentrations from 125-145 ppb.     

A Monte-Carlo Simulation of the Ozone Attainment Process

A Monte-Carlo simulation of the approach to attainment of the National Ambient Air Quality Standard for ozone has been performed for the California Bay Area Air Quality Management District. Four compliance tests together with different design values are used in the simulation. The results show that the present compliance test requiring a zero-percent chance of violation and the design value represented by the fourth highest value in three years makes both the standard and the control requirement much more stringent than generally assumed. In fact, to attain the standard on a long-term basis would require annual means and annual second-highest values that are close to those of the rural background ozone. The simulation also shows that by taking into account the fluctuation of ozone concentrations in the compliance test, such as a t test, and by using a design value consistent with the test, a standard defined in terms of the three-year mean of the annual second-highest values not only is more consistent with the currently- perceived stringency of the present standard, but may also be attainable with a more reasonable control requirement.     

Long-term meteorologically independent trend analysis of ozone air quality at an urban site in the greater Houston area

The Houston-Galveston-Brazoria (HGB) area of Texas has a history of ozone exceedances and is currently classified under moderate nonattainment status for the 2008 8-hr ozone standard of 75 ppb. The HGB area is characterized by intense solar radiation, high temperature, and high humidity, which influence day-to-day variations in ozone concentrations. Long-term air quality trends independent of meteorological influence need to be constructed for ascertaining the effectiveness of air quality management in this area. The Kolmogorov-Zurbenko (KZ) filter technique, used to separate different scales of motion in a time series, is applied in the current study for maximum daily 8-hr (MDA8) ozone concentrations at an urban site (U.S. Environmental Protection Agency [EPA] Air Quality System [AQS] Site ID: 48-201-0024, Aldine) in the HGB area. This site, located within 10 miles of downtown Houston and the George Bush Intercontinental Airport, was selected for developing long-term meteorologically independent MDA8 ozone trends for the years 1990–2016. Results from this study indicate a consistent decrease in meteorologically independent MDA8 ozone between 2000 and 2016. This pattern could be partially attributed to a reduction in underlying nitrogen oxide (NO_x) emissions, particularly lowering nitrogen dioxide (NO₂) levels, and a decrease in the release of highly reactive volatile organic compounds (HRVOCs). Results also suggest solar radiation to be most strongly correlated to ozone, with temperature being the secondary meteorological control variable. Relative humidity and wind speed have tertiary influence at this site. This study observed that meteorological variability accounts for a high of 61% variability in baseline ozone (low-frequency component, sum of long-term and seasonal components), whereas 64% of the change in long-term MDA8 ozone post 2000 could be attributed to NO_x emission reduction. Long-term MDA8 ozone trend component was estimated to be decreasing at a linear rate of 0.412 ± 0.007 ppb/yr for the years 2000–2016 and 0.155 ± 0.005 ppb/yr for the overall period of 1990–2016.

Implications: The effectiveness of air emission controls can be evaluated by developing long-term air quality trends independent of meteorological influences. The KZ filter technique is a well-established method to separate an air quality time series into short-term, seasonal, and long-term components. This paper applies the KZ filter technique to MDA8 ozone data between 1990 and 2016 at an urban site in the greater Houston area and estimates the variance accounted for by the primary meteorological control variables. Estimates for linear trends of MDA8 ozone are calculated and underlying causes are investigated to provide a guidance for further investigation into air quality management of the greater Houston area.     

Ozone patterns for three metropolitan statistical areas in North Carolina, USA

As part of an effort by the state of North Carolina to develop a State Implementation Plan (SIP) for 1-h peak ozone control, a network of ozone stations was established to monitor surface ozone concentrations across the state. Between 19 and 23 ozone stations made continuous surface measurements between 1993 and 1995 surrounding three major metropolitan statistical areas (MSAs): Raleigh/Durham (RDU), Charlotte/Mecklenburg (CLT), and Greensboro/High Point/Winston-Salem (GSO). Statistical averages of the meteorological and ozone data were performed at each Metropolitan Statistical Area (MSA) to study trends and/or relationships on high ozone days (days in which one of the MSA sites measured an hourly ozone concentration90.0 ppbv). County emission maps of precursor gases, wind roses, total area averages of ozone, total downwind averages of ozone deviations, upwind averages of ozone, and a modified delta ozone analysis were all obtained and analyzed. The results of this study show a reduction in the delta ozone relative to an earlier study at RDU, but no average significant change at CLT (no comparison can be made for GSO). The statistical data analyses in this study are used to quantify the importance of local contributions and regional transport, to ozone air pollution in the MSAs.     

Changes in Diurnal Patterns Related to Changes in Ozone Levels

Ozone is an ubiquitous air pollutant that affects both human health and vegetation. There is concern about the number of hours human populations in nonattainment areas in the United States are exposed to levels of 03 at which effects have been observed. As improvement in air quality is achieved, it is possible that 03 control strategies may produce distributions of 1-h 03 concentrations that result in different diurnal profiles that produce greater potential exposures to 03 at known effects levels for multiple hours of the day. These concerns have prompted new analysis of aerometric data. In this analysis, the change in the seasonally averaged diurnal pattern was investigated as changes in 03 levels occurred. For the data used in this analysis, 25 of the 36 sites that changed compliance status across years showed no statistically significant change in the shape of the average diurnal profile (averaged by 03 season). For 71 percent (10 out of 14) of the sites in southern California and Dallas-Fort Worth, Texas, that showed improvement in O3 levels (i.e., reductions in the number of exceedances over the years), but still remained in nonattainment, a statistically significant change in the shape of the seasonally averaged diurnal profile occurred. Based on the results obtained in this study, the evaluation of diurnal patterns may be useful for identifying the influence of changes in emission levels versus meteorological variation on attainment status. Using data from the southern California and Dallas-Fort Worth sites, which showed improvements in 03 levels, changes were observed in the seasonally averaged diurnal profiles. On the other hand, for the sites moving between attainment and nonattainment status, such a change in shape was generally not observed and it was possible that meteorology played a more important role than changes in emission levels relative to attainment status.     

The Need for a More Robust Ozone Air Quality Standard

The present National Ambient Air Quality Standard for ozone has many statistical problems, including use of extreme values which have inherent large fluctuations, a compliance test that can gradually lower the target of the design value below the standard level, and inconsistencies between the number-of-exceedances criterion and the design value. The above problems can be avoided or minimized by using a more robust statistic, such as the 95th percentile, and applying a statistical compliance test, without sacrificing the stringency of the standard. Analysis of EPA’s ozone data shows that the annual 95th percentiles and their three-year means have less variability than the annual second highest values and the fourth highest values in three years, respectively. A t test for the mean of the annual 95th percentiles is proposed for compliance testing not only to preserve the averaging concept of the present standard, but also to take account of ozone concentration fluctuations in order to increase the stability of the compliance status of a site or a Metropolitan Statistical Area. A procedure is provided to adjust the level of the 95th-percentile standard so that the stringency of the present standard is preserved.     

Scientific Basis for the VOC Reactivity Issues Raised by Section 183(e) of the Clean Air Act Amendments of 1990

A methodology for determining regional ozone design values and the expected number of exceedances is described. The methodology was applied to data bases for one year or less from four U.S. urban areas: Houston, Los Angeles, Philadelphia, and St. Louis. The effects of reducing numbers of stations in a network were tested, and it was concluded that networks of nine or ten appropriately selected stations are adequate for estimating design values. Using the methodology described, the expected number of exceedances tends to be underestimated when using smaller networks; however, this appears to be an artifact of the conservative approach taken in developing the methodology.     

Ozone and UV-B variations at Ispra from 1993 to 1997

An analysis of the variability of the total ozone column at Ispra (Italy) has been performed to ascertain if, even in a short-time interval of 5 years (1993–1997), a decline of the monthly mean ozone values could be demonstrated. A linear fit of the data displays a decrease of 0.21% per year with a mean value equal to 319±2 D.U. and an amplitude of the annual cycle of about 10% of the mean. A linear regression of the surface monthly mean ozone values has also been performed showing a decreasing trend (−1% per year) that could contribute, even if for a very small amount, to the decline of the total ozone values. Ispra monthly mean total ozone data have been compared with those of three stations located within 2° latitude and 3° longitude from Ispra (Haute Provence, Hohenpeissenberg and Arosa). A linear fit of the data shows some discrepancies in the ozone changes, which can be attributed to the limited length of the observational period.An analysis has been performed to verify if the variation of ozone at Ispra is in agreement with that of the solar UV measured at a wavelength (305 nm) where the ozone absorption is still remarkable. The results, taken at a fixed solar zenith angle of 68°, show a clear anticorrelation between the monthly mean values of UV and the corresponding values of the total ozone column; the linear fit of the UV data displays an increase of 2.0% per year, much higher than expected from the ozone decrease, and a mean value of 1.4±0.1 mW m^-2 nm^-1.     

Design of Afterburners for Varnish Cookers

Currently, outdoor ozone levels in many U.S. cities exceed the primary health-based national ambient air quality standard. While outdoor ozone levels are an important measure of the severity of those exceedances, people typically spend more than 80 percent of their time Indoors, where ozone levels are lower. Indoor ozone levels range from 10 to 80 percent of outdoor levels, with many people receiving a substantial portion of their ozone exposure while indoors. This paper uses an Indoor air quality model (IAQM) to estimate indoor ozone levels by mlcroenvlronment type (home, office, and vehicle) and configuration (windows open, windows closed, older construction, weatherized, and air conditioned). The formulation of IAQM is discussed, along with specification of model parameters for ozone. The multicompartment version of IAQM is described, with a single-compartment version used for the analyses. IAQM-calculated ozone indoor-outdoor ratios compare well with research-reported values. Results indicate that ozone peak-concentration indoor-outdoor ratios range as follows: home—0.65 (windows open), 0.36 (air conditioned), 0.23 (typical construction, windows closed), and 0.05 (energy-efficient construction, windows closed); office—0.82 (heat-Ing, ventilation and air conditioning systems supplying 100 percent outdoor air), 0.60 (typical HVAC), and 0.32 (energy-efficient HVAC); and vehicle—0.41 (85 mph), 0.33 (55 mph), and 0.21 (10 mph). Analysis results are presented to characterize IAQM’s sensitivity to assumed model parameters.     

Use of an indoor air quality model (IAQM) to estimate indoor ozone levels

Currently, outdoor ozone levels in many U.S. cities exceed the primary health-based national ambient air quality standard. While outdoor ozone levels are an important measure of the severity of those exceedances, people typically spend more than 80 percent of their time indoors, where ozone levels are lower. Indoor ozone levels range from 10 to 80 percent of outdoor levels, with many people receiving a substantial portion of their ozone exposure while indoors. This paper uses an indoor air quality model (IAQM) to estimate indoor ozone levels by microenvironment type (home, office, and vehicle) and configuration (windows open, windows closed, older construction, weatherized, and air conditioned). The formulation of IAQM is discussed, along with specification of model parameters for ozone. The multicompartment version of IAQM is described, with a single-compartment version used for the analyses. IAQM-calculated ozone indoor-outdoor ratios compare well with research-reported values. Results indicate that ozone peak-concentration indoor-outdoor ratios range as follows: home--0.65 (windows open), 0.36 (air conditioned), 0.23 (typical construction, windows closed), and 0.05 (energy-efficient construction, windows closed); office--0.82 (heating, ventilation and air conditioning systems supplying 100 percent outdoor air), 0.60 (typical HVAC), and 0.32 (energy-efficient HVAC); and vehicle--0.41 (85 mph), 0.33 (55 mph), and 0.21 (10 mph). Analysis results are presented to characterize IAQM's sensitivity to assumed model parameters.     

Modelling critical levels of ozone for the forested area of austria modifications of the aot40 concept

GOAL, SCOPE AND BACKGROUND: Ozone is the most important air pollutant in Europe for forest ecosystems and the increase in the last decades is significant. The ozone impact on forests can be calculated and mapped based on the provisional European Critical Level (AOT40 = accumulated exposure over a threshold of 40 ppb, 10,000 ppb x h for 6 months of one growing season calculated for 24 h day(-1)). For Norway spruce, the Austrian main tree species, the ozone risk was assessed in a basis approach and because the calculations do not reflect the health status of forests in Austria, the AOT40 concept was developed. METHODS: Three approaches were outlined and maps were generated for Norway spruce forests covering the entire area of Austria. The 1st approach modifies the AOT40 due to the assumption that forests have adapted to the pre-industrial levels of ozone, which increase with altitude (AOTalt). The 2nd approach modifies the AOT40 according to the ozone concentration in the sub-stomata cavity. This approach is based on such factors as light intensity and water vapour saturation deficit, which affect stomatal uptake (AOTsto). The 3rd approach combines both approaches and includes the hemeroby. The pre-industrial ozone level approach was applied for autochthonous ('natural') forest areas, the ozone-uptake approach for non-autochthonous ('altered') forest areas. RESULTS AND DISCUSSION: The provisional Critical Level (AOT40) was established to allow a uniform assessment of the ozone risk for forested areas in Europe. In Austria, where ozone risk is assessed with utmost accuracy due to the dense grid of monitoring plots of the Forest Inventory and because the continuously collected data from more than 100 air quality measuring stations, an exceedance up to the five fold of the Critical Level was found. The result could lead to a yield loss of up to 30-40% and to a severe deterioration in the forest health status. However, the data of the Austrian Forest Inventory and the Austrian Forest Damage Monitoring System do not reflect such an ozone impact. Therefore, various approaches were outlined including the tolerance and avoidance mechanisms of Norway spruce against ozone impact. Taking into consideration the adaptation of forests to the pre-industrial background level of ozone, the AOT40 exceedances are markedly reduced (1st approach). Taking into account the stomatal uptake of ozone, unrealistic high amounts of exceedances up to 10,000 ppb x h were found. The modelled risk does not correspond with the health status and the wood increment of the Austrian forests (2nd approach). Consolidating the forgoing two approaches, a final map including the hemeroby was generated. It became clear that the less natural ('altered') forested regions are highly polluted. This means, that more than half of the spruce forests are endangered by ozone impact and AOT40 values of up to 30,000 ppb x h occur (3rd approach). CONCLUSIONS: The approaches revealed that a plausible result concerning the ozone impact on spruce forests in Austria could only be reached by combining pre-industrial ozone levels, ozone flux into the spruce needles and the hemeroby of forests.     

Forecasting air quality time series using deep learning

This paper presents one of the first applications of deep learning (DL) techniques to predict air pollution time series. Air quality management relies extensively on time series data captured at air monitoring stations as the basis of identifying population exposure to airborne pollutants and determining compliance with local ambient air standards. In this paper, 8 hr averaged surface ozone (O₃) concentrations were predicted using deep learning consisting of a recurrent neural network (RNN) with long short-term memory (LSTM). Hourly air quality and meteorological data were used to train and forecast values up to 72 hours with low error rates. The LSTM was able to forecast the duration of continuous O₃ exceedances as well. Prior to training the network, the dataset was reviewed for missing data and outliers. Missing data were imputed using a novel technique that averaged gaps less than eight time steps with incremental steps based on first-order differences of neighboring time periods. Data were then used to train decision trees to evaluate input feature importance over different time prediction horizons. The number of features used to train the LSTM model was reduced from 25 features to 5 features, resulting in improved accuracy as measured by Mean Absolute Error (MAE). Parameter sensitivity analysis identified look-back nodes associated with the RNN proved to be a significant source of error if not aligned with the prediction horizon. Overall, MAE's less than 2 were calculated for predictions out to 72 hours.

Implications: Novel deep learning techniques were used to train an 8-hour averaged ozone forecast model. Missing data and outliers within the captured data set were replaced using a new imputation method that generated calculated values closer to the expected value based on the time and season. Decision trees were used to identify input variables with the greatest importance. The methods presented in this paper allow air managers to forecast long range air pollution concentration while only monitoring key parameters and without transforming the data set in its entirety, thus allowing real time inputs and continuous prediction.     

Establishment of a Box-Jenkins Multivariate Time-Series Model to Simulate Ground-Level Peak Daily One-Hour Ozone Concentrations at Ta-Liao in Taiwan

Abstract

Box-Jenkins univariate autoregressive integrated moving average (ARIMA) and regression with time-series error (RTSE) models were established to simulate historical peak daily 1-hr ozone concentrations at Ta-Liao, Taiwan, 1997– 2001. During 1995–2003, the 600 days of Pollution Standard Index (PSI) more than 100 (peak daily 1-hr ozone concentrations detected by greater than 120 ppm) at Tao-Liao showed the highest ozone exceedances among the six monitoring stations in Kaohsiung County. To improve the predictability of extremely high ozone, two different principal components, PC1 and PC(1 + 2), were introduced in the RTSE model. Four typical predictors (particular matter with an aerodynamic diameter less than or equal to 10 μm, temperature, wind speed, and wind direction) plus a PC trigger remained signi?cant in the RTSE model. The model performance statistics concluded that the RTSE model with PC1 was optimal, compared with the univariate ARIMA, the RTSE model without PC, and RTSE model with PC(1 + 2). The contingency table shows that the successful predictions of the univariate model were only 12.9% of that of the RTSE model with PC1. Also, the POD value was improved approximately 5-fold when the univariate model was replaced by the RTSE model, and almost 8-fold when it was replaced by the RTSE model with PC1. Moreover, introducing the PC trigger indeed enhanced the ozone predictability. After the PC trigger was introduced in the RTSE model, the POD was increased 69.9%, and the FAR was reduced 8.3%. The overall correlation between the observed and simulated ozone was improved 9.6%. Also, the ?rst principal component was more useful than the ?rst two components in playing the “trigger” role, though it counted only for<br/>58.62% of the environmental variance during the high ozone days.     

A climatology of regional background ozone at different elevations in Switzerland (1992–1998)

The ozone records of several monitoring stations in Switzerland from 1992 to 1998 are investigated with respect to the variability observed during regional background conditions, i.e. conditions with little detectable local or regional-scale influences as evident by NO_x and CO concentrations. The sites cover different altitudes between 490 and 3600 m asl. They are characteristic of near-surface conditions, the top of the planetary boundary layer or residual layer, the complex atmosphere in an alpine valley, and the free troposphere. The results reveal a distinctly different ozone variability (diurnal cycles, seasonal cycles, trends) during regional background conditions compared to all days. The estimated annual average ozone concentration under these conditions is between 33 and 50 ppb, dependent on altitude, with a spring maximum and an autumn/winter minimum. Differences in background ozone are found depending on the synoptic weather type. For all sites a positive ozone trend is calculated for background conditions that is larger than for all data. For the latter, the trends appear to be stronger positive for the last 7 years than for the last 11 years.     

Hydrocarbon emissions from industrial release events in the Houston-Galveston area and their impact on ozone formation

Ambient measurements have shown that ozone formation in the Houston-Galveston area of Texas is frequently much more rapid than in other urban areas. One of the contributing factors is believed to be short-term episodic or “event” emissions from industrial facilities, particularly releases that contain significant mass fractions of highly reactive volatile organic compounds (HRVOCs). In this work, time series analyses are used to compare average annual flow rates for air pollutant emissions with those released during reported emission events. The results indicate that the magnitude and frequency of HRVOC event emissions are an important element in accurately reflecting ozone precursor emission patterns in the Houston-Galveston area, particularly in Harris, Brazoria, Galveston, and Chambers counties. More than 50% of the reported episodic (event) emissions of HRVOCs are ethene and approximately a third are propene; the remainders are isomers of butene and 1,3-butadiene. Most events last less than 24 h. The mass released in an event can vary from a few hundred to more than 100,000 lb, and the dominant type of industrial source is chemical manufacturers (SIC 2869). Daily emissions from a single facility can vary from annual average emissions by multiple orders of magnitude at a frequency of several times a year. Because there are so many facilities in the Houston-Galveston area, HRVOC emission variability of this magnitude can be expected daily, at some time and some location in the Houston-Galveston area. If the emission variability occurs at times and locations where atmospheric conditions are conducive to ozone formation, both ambient data and photochemical modeling indicate that industrial emission events can lead to elevated concentrations of ozone. Specifically, peak, area-wide ozone concentration can be increased by as much as 100 ppb for large HRVOC emission events.