COVID lockdowns significantly affect statewide atmospheric fine aerosols in India after excluding long-term pollution patterns and time-lag effect

Abstract

Atmospheric fine particulate matter (PM_2.5, near-surface concentrations when size is ≤ 2.5 μm) affects global climate and human health. India alone accounts for a quarter of the global PM_2.5-related health burden. Studies in India, mostly in urban areas, have reported significant declines in PM_2.5 concentrations because of COVID-19 lockdown. These studies did not consider the long-term PM_2.5 patterns and time-lag effect (inter-yearly variations in PM_2.5 concentrations carried forward from one period to another due to interannual shifts in meteorological conditions). Since the studies focused primarily on urban areas, not covering rural/remote areas where pollution may rise during lockdown, it is still unclear what impact lockdown had on statewide pollution levels in India. Here, we examine whether significant changes in fine-mode aerosol optical depth (AOD_f: columnar PM_2.5) occurred statewide across India because of lockdown after excluding the confounding variables. We found a substantial decrease in AOD_f in a few (28%) states/territories. The declines were significant (ANCOVA; α = 0.05) in some Northeastern states/territories: Sikkim (29%), Arunachal Pradesh (24%), Nagaland (5%), Mizoram (4%) and Uttarakhand (3%). However, in most states/territories, AOD_f increased significantly because of lockdown. The lockdown-associated hardship caused more people to rely on polluting cooking fuels, thereby increasing residential emissions, particularly in rural areas. At city-level, we found significant reductions in near-surface PM_2.5 concentrations due to lockdown. These declines were comparable or greater than previously reported. Also, there were significant reductions in AOD_f (PM_2.5 concentrations) at state (city) levels resulting from previous environmental intervention measures. If not accounted, previous environmental intervention measures can significantly bias lockdown effect estimates in India.