High‐Resolution Lagrangian Inverse Modeling of CO₂ Emissions Over the Paris Region During the First 2020 Lockdown Period

Abstract

Stringent mobility restrictions across the world during the COVID 19 pandemic have impacted local economies and, consequently, city carbon budgets, offering a unique opportunity to evaluate the capability of scientific approaches to quantify emissions changes. Our study aims to quantify and map CO2 emissions from fossil fuel and biogenic CO2 fluxes over the Paris metropolitan area during the first lockdown period (March‐May 2020) in France, in comparison with the same period in 2019. Our inversion system relies on transport model simulations initiated with the Weather Research and Forecasting chemistry transport model combined with a high‐resolution fossil fuel CO2 emissions inventory, and biogenic CO2 fluxes from a vegetation model. The inversion with atmospheric observations from a network of six towers resulted in a positive re‐adjustment of fossil fuel CO2 emissions in 2019 and 2020 compared to prior. In 2020, the inversion resulted in a large emission reduction (43%) compared to 2019, while the reductions were estimated to be 37% based on the prior inventory itself. By assimilating CO mixing ratios in addition to CO2, the traffic emission estimates were reduced by 68% in 2020, compared to nontraffic (29%). Various sensitivity tests show that prior emission uncertainty and different background conditions significantly impacted the emissions estimates. We conclude that our current inversion system with atmospheric CO2 monitoring makes it possible to identify the emission decrease in 2020 partly over the urban region. However, additional information on prior emission errors and a dense network will be needed to map emissions precisely.

Can We Detect Urban‐Scale CO2 Emission Changes Within Medium‐Sized Cities?

“…Downtown SLC is co-located with significant commercial activity with commuting workers, which would have been most impacted by the COVID-19 lockdown. It is worth noting that the COVID-19 emission reductions estimated in SLC are lower than other cities such as Los Angeles, Paris, the Bay Area (Lian et al, 2022;Nalini et al, 2022;Turner et al, 2020;Yadav et al, 2021). It is suspected that less stringent lockdown measures in SLC relative to other cities across the U.S were the driving factors behind the smaller emission reductions.…”

“…Furthermore, the COVID‐19 lockdown measures were less stringent in SLC (Hallas et al., 2021), which would have presumably lowered CO 2 emission reductions in SLC relative to other cities. Cities, such as Paris and San Francisco observed much larger reductions in CO 2 emissions that ranged between 30% and 50% (Lian et al., 2022; Nalini et al., 2022; Turner et al., 2020; Yadav et al., 2021), where it would be easier to detect emission changes with a CO 2 monitoring network. Finally, there has been limited work on determining whether alternative CO 2 measurement strategies, such as mobile observations, can detect urban‐scale emission changes.…”

“…The COVID‐19 lockdown provided a unique opportunity to determine whether hybrid modeling approaches could identify changes in urban CO 2 emissions at the city‐scale based on observed differences in CO 2 during the lockdown period. On average, major metropolitan areas observed emission reductions during the COVID‐19 lockdown that ranged between 30% and 55% (Lian et al., 2022; Nalini et al., 2022; Turner et al., 2020; Yadav et al., 2021) depending on the city. Reductions in traffic were thought to be the primary driver in emission reductions in Paris and the San Francisco Bay Area during the COVID‐19 lockdown (Lian et al., 2022; Turner et al., 2020).…”

“…Given uncertainties surrounding atmospheric transport, instrument uncertainties, and background estimates of CO 2 , it was suggested that top‐down modeling approaches are limited to detecting emission reductions greater than 20% for urban areas (Lian et al., 2022). Network density was also highlighted as a factor that may limit urban‐scale mapping of emission adjustments (Nalini et al., 2022). However, alternative observation techniques such as mobile‐based measurement platforms have been shown to provide more detailed emission adjustments that can resolve city‐scale emission patterns (Mallia et al., 2020).…”

Near-real-time CO2 fluxes from CarbonTracker Europe for high-resolution atmospheric modeling

“…Anthropogenic emissions are highly variable in time (e.g., Nassar et al, 2013;Mues et al, 2014) and are subject to socioeconomic factors as well as meteorology. Moreover, they are very heterogeneous in space; in this section, we elaborate on the spatiotemporal distribution of the anthropogenic (Agustí-Panareda et al, 2014); CT-NRT: CarbonTracker-Near Real-Time (Chen et al, 2019); EDGAR: Emission Database for Global Atmospheric Research (Janssens-Maenhout et al, 2019); CDIAC: Carbon Dioxide Information and Analysis Center (Andres et al, 1996); BFAS: Biosphere Flux Adjustment Scheme (Agustí-Panareda et al, 2016); clim.…”