Minimal Climate Impacts From Short‐Lived Climate Forcers Following Emission Reductions Related to the COVID‐19 Pandemic

Abstract We present an assessment of the impacts on atmospheric composition and radiative forcing of short‐lived pollutants following a worldwide decrease in anthropogenic activity and emissions comparable to what has occurred in response to the COVID‐19 pandemic, using the global composition‐climate model United Kingdom Chemistry and Aerosols Model (UKCA). Emission changes reduce tropospheric hydroxyl radical and ozone burdens, increasing methane lifetime. Reduced SO2 emissions and oxidizing capacity lead to a decrease in sulfate aerosol and increase in aerosol size, with accompanying reductions to cloud droplet concentration. However, large reductions in black carbon emissions increase aerosol albedo. Overall, the changes in ozone and aerosol direct effects (neglecting aerosol‐cloud interactions which were statistically insignificant but whose response warrants future investigation) yield a radiative forcing of −33 to −78 mWm−2. Upon cessation of emission reductions, the short‐lived climate forcers rapidly return to pre‐COVID levels; meaning, these changes are unlikely to have lasting impacts on climate assuming emissions return to pre‐intervention levels.


Introduction
The following is supporting information for the main text. It consists of additional plots of parameters calculated from processing UKCA output data, additional emission reduction information, further detail as to the calculation of methane concentration perturbation and additional information about the species emitted in the UKCA model.    . Zonal mean change in O3 mixing ratio and zonal mean percentage change in OH mixing ratio between. In each case, the plot show the effect of reducing the emissions of the sector of interest (e.g. A1-A2 shows impact of reducing aircraft NOx emissions from a 25% reduction to a 50% reduction). All changes for mid March -mid May averaged over 3 years 2012-2014 Black lines show the tropopause.      Hatching indicates regions where the radiative forcing is not statistically significant (95% confidence level). Values above each plot show the area-weighted mean with the mean 95% confidence interval. Figure S12. Radiative forcing from aerosol-cloud interactions for the period of lowest emissions, averaged over 3 years. Hatching indicates regions where the radiative forcing is not statistically significant (95% confidence level). Values above each plot show the area-weighted mean with the mean 95% confidence interval.

Text S1: Emission Scenario Description
The emissions scenarios were conceptualised in late March/early April 2020 when verified data concerning the impact of lockdowns on anthropogenic sector emissions was not plentiful or widely available. In order to best estimate the reductions, we compiled information from several sources which are detailed below: Lockdown measures resulted in an 88% decline in car use in the EU and a 60% decrease in industrial carbon emissions by 25th March (Mallet, 2020), the EEA reported that NO2 concentrations in several cities in southern Europe were around 50% lower than 2019 (European Environment Agency, 2020). In the UK, there was a 60% reduction in all motor vehicle use in the UK (UK Department of Transport, 2020).
International flights from the UK, USA, China, Germany and Japan have decreased 75% from January to the end of March this year (Kommenda, 2020), and european internal flights are estimated to have decreased by 86%. Data from Flightradar (FlightRadar, 2020) was also used to estimate a change in the total global flight by around 50%. Some uncertainty was present early on due to the 'ghost flights' berth requirements law, but the law was later suspended (Morgan, 2020).
The industrial sector was also hit by the COVID-19 lockdowns, but was a lot harder to quantify. It was suggested that EU industrial emissions decreased up to 60% [FT 2020], Whilst it is likely that many other sectors were affected by the lockdowns, the data at the time provided insufficient evidence to come up with perturbations, so we did not attempt to estimate any of these changes. In this manner, our scenarios most likely represent a lower bound on the actual effect.

Text S2: Methane Concentration Evolution
To estimate the transient change in methane as a result of its lifetime perturbation, a simple kinetic model is considered with an instantaneous 4% increase in methane lifetime. This produces an upper bound estimate for methane concentration as the lifetime change in scenarios A1-A4 are not instantaneous and only one scenario, A3, reaches 4% (Fig. S5). Nevertheless, the results are informative.
In this model, the initial steady state concentration of methane, , is defined in terms of methane flux, , and its lifetime, : Upon an instantaneous perturbation of methane lifetime to perturbed value, , with an unchanged flux, the methane concentration ceases to be that given by the steady state expression and can be described by the following differential equation: Solving this via separation of variables yields: Where is the constant integration.
Noting that at = 0, [ 4 ] = [ 4 ] 0 , the constant of integration, , can be written as: Dispensing with the moduli and multiplying by − yields: The ratio, , of perturbed [ 4 ]to original [ 4 ] 0 can be expressed as: This ratio satisfies the requirements: The ratio depends weakly on the initial methane lifetime but it is clear that several decades are needed for the model to reach a new steady state concentration (Fig. S18). A 4% instantaneous increase in methane lifetime after 3 months, the length of the simulated perturbation, will result in a 0.1% increase in methane concentrations. Figure S17. Percentage change in perturbed methane concentration after an instantaneous increase in methane lifetime of 4% using the simple kinetic framework over the first 5 months (b) and 40 years (b). Three initial methane lifetimes were considered.