Search

Several Working Groups have been established within the frame of the second phase of the Tropospheric Ozone Assessment Report (TOAR-II). The Tropospheric Ozone Precursors focus Working Group (TOP WG) aims to examine the current regional and global distribution, variability and trends of ozone precursors. Part of our strategy has been to analyze in greater detail different regions of the globe. In particular, this work addresses South America, a region characterized by densely populated urban areas with high air pollution levels.We use data from air quality monitoring networks that measure surface-level ozone, nitrogen oxides, carbon monoxide and meteorological variables. To date, we have validated and incorporated monitoring stations from Brazil (51), Chile (18), Colombia (13) and Ecuador (6) into our central database. To evaluate short- and long-term ozone exposure, we use the maximum daily 8-hour average (MDA8) and the peak season guideline proposed by the World Health Organization (WHO) set at 51 and 31 ppbv, respectively. We applied the Quantile Regression (QR) method to analyze ozone and precursor network trends. We also identified points in time series that mark changes in trends through a piecewise function.The highest MDA8 ozone for 2015-2021 was found in São Paulo (52 ppbv) and Santiago (51 ppbv). In São Paulo, the short-term ozone exposure decreased by 7% compared to the average of the years prior to 2015 (period analyzed in TOAR phase I), while in Santiago, it increased by 10%. In Bogota and Quito, the MDA8 complied with the WHO guidelines (33 and 32 ppbv, respectively). Similarly, the long-term ozone exposure guideline was exceeded in São Paulo (39 ppbv) and Santiago (40 ppbv), while Bogota (25 ppbv) and Quito (26 ppbv) complied. The trend analysis showed that Quito was the only city with a negative ozone trend of -0.10 ppb/year (50th percentile) for the analyzed period. In turn, the São Paulo trend increased after 2008 (0.43 ppbv/year), while Santiago and Bogota have increased since 2017 (0.93 ppbv/year and 1.3 ppbv/year, respectively). We highlight that the positive trend in Santiago is driven mainly by the high percentiles (>70th). Underlying processes that explain trends involve more efficient photochemical ozone formation (e.g., NO2/NOx trend) and meteorological factors.This ongoing work aims to include more South American cities and background stations already available in the new TOAR database. Finally, we will project ozone trends for the next decade using machine learning techniques (random forest) under precursor emission scenarios and temperature projections.

This study delves into the photochemical atmospheric changes reported globally during the pandemic by analyzing the change in emissions from mobile sources and the contribution of local meteorology to ozone (O3) and particle formation in Bogotá (Colombia), Santiago (Chile), and São Paulo (Brazil). The impact of mobility reductions (50%–80%) produced by the early coronavirus-imposed lockdown was assessed through high-resolution vehicular emission inventories, surface measurements, aerosol optical depth and size, and satellite observations of tropospheric nitrogen dioxide (NO2) columns. A generalized additive model (GAM) technique was also used to separate the local meteorology and urban patterns from other drivers relevant for O3 and NO2 formation. Volatile organic compounds, nitrogen oxides (NOx), and fine particulate matter (PM2.5) decreased significantly due to motorized trip reductions. In situ nitrogen oxide median surface mixing ratios declined by 70%, 67%, and 67% in Bogotá, Santiago, and São Paulo, respectively. NO2 column medians from satellite observations decreased by 40%, 35%, and 47%, respectively, which was consistent with the changes in mobility and surface mixing ratio reductions of 34%, 25%, and 34%. However, the ambient NO2 to NOx ratio increased, denoting a shift of the O3 formation regime that led to a 51%, 36%, and 30% increase in the median O3 surface mixing ratios in the 3 respective cities. O3 showed high sensitivity to slight temperature changes during the pandemic lockdown period analyzed. However, the GAM results indicate that O3 increases were mainly caused by emission changes. The lockdown led to an increase in the median of the maximum daily 8-h average O3 of between 56% and 90% in these cities.

The ozone mixing ratio spatiotemporal variability in the pristine South Pacific Ocean is studied, for the first time, using 21-year-long ozone (O3) records from the entire southern tropical and subtropical Pacific between 1994 and 2014. The analysis considered regional O3 vertical observations from ozonesondes, surface carbon monoxide (CO) observations from flasks, and three-dimensional chemistry-transport model simulations of the global troposphere. Two 21-year-long numerical simulations, with and without biomass burning emissions, were performed to disentangle the importance of biomass burning relative to stratospheric intrusions for ambient ozone levels in the region. Tagged tracers of O3 from the stratosphere and CO from various biomass burning regions have been used to track the impact of these different regions on the southern tropical Pacific O3 and CO levels. Patterns have been analyzed based on atmospheric dynamics variability. Considering the interannual variability in the observations, the model can capture the observed ozone gradients in the troposphere with a positive bias of 7.5 % in the upper troposphere/lower stratosphere (UTLS) as well as near the surface. Remarkably, even the most pristine region of the global ocean is affected by distant biomass burning emissions by convective outflow through the mid and high troposphere and subsequent subsidence over the pristine oceanic region. Therefore, the biomass burning contribution to tropospheric CO levels maximizes in the UTLS. The Southeast Asian open fires have been identified as the major contributing source to CO from biomass burning in the tropical South Pacific, contributing on average for the study period about 8.5 and 13 ppbv of CO at Rapa Nui and Samoa, respectively, at an altitude of around 12 km during the burning season in the spring of the Southern Hemisphere. South America is the second-most important biomass burning source region that influences the study area. Its impact maximizes in the lower troposphere (6.5 ppbv for Rapa Nui and 3.8 ppbv for Samoa). All biomass burning sources contribute about 15–23 ppbv of CO at Rapa Nui and Samoa and account for about 25 % of the total CO in the entire troposphere of the tropical and subtropical South Pacific. This impact is also seen on tropospheric O3, to which biomass burning O3 precursor emissions contribute only a few ppbv during the burning period, while the stratosphere–troposphere exchange is the most important source of O3 for the mid troposphere of the South Pacific Ocean, contributing about 15–20 ppbv in the subtropics.

We present the results of a neural network model designed for the forecasting of hourly PM2.5 concentrations in Santiago, Chile. The study focuses on the observed values at two of the monitoring stations, which are located in the south-west zone of the city and are among the stations that register the highest concentrations during the period between April and August. This is the season when air quality is very often in ranges that are harmful to the population and some restrictions on emissions become useful. The forecasting model is a multilayer neural network. The input variables are observed values of hourly PM10 and PM2.5 concentrations measured at the station of interest and at a neighboring station at 7 PM of the present day and some observed and forecasted meteorological variables. NO2 concentrations during the morning and afternoon hours, which may be associated with secondary particle formation, are also used as input. The implemented models are trained with 2014 and 2015 data and tested with 2016 values. Information is collected until 7 PM of the present day, and the largest forecasting error up to 21 hours in advance is 32%. The accuracy of this forecasting is better than that obtained with a neural model previously used for the forecasting of hourly PM2.5 concentrations in the north-west zone in Santiago. Our neural models show better results than those obtained with linear models with the same input variables. The developed models provide a tool for anticipating episodes in Santiago and other cities with similarly unfavorable conditions for pollutant dispersion.

Air pollution has been shown to have a direct effect on human health. In particular, PM2.5 has been proven to be related to cardiovascular and respiratory problems. Therefore, it is important to have accurate models to predict high pollution events for this and other pollutants. We present different models that forecast PM2.5 maximum concentrations using a Long Short-Term Memory (LSTM) based neural network and a Deep Feedforward Neural Network (DFFNN). Ten years of air pollution and meteorological measurements from the network of monitoring stations in the city of Santiago, Chile were used, focusing on the behaviour of three zones of the city. All missing values were rebuilt using a method based on discrete cosine transforms and photochemical predictors selected through unsupervised clustering. Deep learning techniques provide significant improvements compared to a traditional multi-layer neural networks, particularly the LSTM model configured with a 7-day memory window (synoptic scale of pollution patterns) can capture critical pollution events at sites with both primary and secondary air pollution problems. Furthermore, the LSTM model consistently outperform deterministic models currently used in Santiago, Chile.

International audience; The changing composition of the atmosphere, driven by anthropogenic emissions, is the cause of anthropogenic climate change as well as deteriorating air quality. Emission inventories are essential to understand the contribution of various human activities, model and predict the changing atmospheric composition, and design cost-effective mitigation measures. At present, national emission inventories in South America (SA) focus on Greenhouse Gases (GHG) as part of their obligation to the United Nations Framework Convention for Climate Change (UNFCC) within the framework of their national communications. Emission inventories other than GHG in SA focus mainly on growing urban areas and megacities. Therefore, studies examining air quality at national, regional or continental scales in SA depend on (down-scaled) global emission inventories. This paper examines the emission estimates of air pollutants from various global inventories for five SA countries, namely Argentina, Brazil, Chile, Colombia and Peru. A more detailed analysis is conducted for the EDGAR and ECLIPSE emission inventories, in particular comparing local city-scale inventories of a major city in each country. Although total emissions between down-scaled global inventories and local city inventories are often comparable, large discrepancies exist between the sectoral contributions. This is critical, as the mitigation of poor air quality will depend on addressing the right sources. Potential sources of discrepancies between global and local inventories include the spatial distribution proxies, difference in emission factors used and/or the use of generic statistical country data when estimating emissions. This highlights the importance of using local information when generating national emission inventories, especially for air quality modeling and development of effective mitigation measures. This study represents the first step towards an increased understanding of the strengths and weaknesses of emissions information in SA.

Coyhaique is a southern Chilean city with a population of approximately 64,000 habitants. In spite of its small size, Coyhaique has been identified as the city with highest annual PM2.5 concentrations of the Americas (including south America, central America and north America). Episodes of high pollution are concentrated on the fall- winter season when meteorological conditions do not favor atmospheric particle dispersion and extended use of wood stoves is responsible for more than 99% of the emissions. In Chile, the 24 h average of PM2.5 concentration is classified in four ranges: fair, bad, very bad and critical. We have developed a neural network model and a linear model aimed to forecast the maximum of the 24 h moving average one day in advance. Input variables for the models are hourly values of PM2.5 at 18 h and 19 h of the present day, measured and forecasted temperature, wind speed and precipitation and measured values of NO2, CO and O3 concentrations. The neural network model is slightly more accurate than the linear model. We are able to anticipate the observed range in 75% of the cases, and critical days in 84% of the cases.