Machine-learning-based “weather-normalization” algorithms to establish the An emerging approach to address this problem is to develop Thus, simply comparing theĬoncentration of pollutants during the COVID-19 period to those immediatelyīeforehand or to the same period in previous years is not sufficient to indicateĬausality. Pollution due to variations in synoptic conditions (weather), seasonalĮffects, and long-term emission trends, as well as the nonlinear responsesīetween emissions and concentrations. Magnitude of these impacts is complicated by the natural variability of air While this may provideīoth air quality and climate benefits, a quantitative assessment of the AtmosphericĬoncentrations of nitrogen dioxide (NO 2) thus readily respond to localĬhanges in NO x emissions (Lamsal et al., 2011). NO x has a short atmospheric lifetime and are predominantlyĮmitted during the combustion of fossil fuel for industry, transport, andĭomestic activities (Streets et al., 2013 Duncan et al., 2016). Pollutants, notably nitrogen oxides (NO x= NO + NO 2) (Liu et al.,Ģ020a Dantas et al., 2020 Petetin et al., 2020 Tobias et al., 2020 Le etĪl., 2020). The associated decrease in industrial production, energy consumption,Īnd transportation resulted in a reduction in the emissions of air Syndrome coronavirus 2 (SARS-CoV-2, hereafter COVID-19) led to a sharpĭecline in human activities across the globe (Le Quéré et al.,Ģ020). Hemisphere spring of 2020 to slow the spread of the severe acute respiratory The stay-at-home orders imposed in many countries during the Northern With declines in surface O 3 forecasted if NO x emission The O 3 response is dependent on season, timescale, and environment, Reflecting a reduction in photochemical production. Nighttime ozone due to reduced titration and a decrease in daytime ozone, Indicates a flattening of the O 3 diurnal cycle with an increase in O 3 between February–June 2020 to be small. While surface O 3 increased by up toĥ0 % in some locations, we find the overall net impact on daily average Response of surface O 3 is complicated by competing influences of Reduction during the first 6 months of 2020 amounted to 3.1 (2.6–3.6) TgN,Įquivalent to 5.5 (4.7–6.4) % of the annual anthropogenic total. We estimate that the global NO x (NO + NO 2) emission The US have been more gradual, with a halting recovery starting in late China experienced the earliest and steepest decline,īut concentrations since April have mostly recovered and remained withinĥ % of the business-as-usual estimate. On average, NO 2Ĭoncentrations were 18 (13–23) % lower than business as usual fromįebruary 2020 onward. Little change (e.g., Rio de Janeiro, Taipei). Ranging from 60 % in severely affected cities (e.g., Wuhan, Milan) to In NO 2 coincide with the timing and intensity of COVID-19 restrictions, Observation sites in 46 countries from January through June 2020. We use a machine learning algorithm driven by information from the NASA GEOS-CF model toĪssess changes in nitrogen dioxide (NO 2) and ozone (O 3) at 5756 Synoptic and seasonal variability of air pollutants. Quantifying theseĬhanges requires a business-as-usual counterfactual that accounts for the To widespread reductions in air pollutant emissions.