Comparison of three approaches to ~1km resolution CO2 and CO emission estimates before and during the COVID-19 shelter-in-place in the SF bay Area

Abstract:

Emissions of CO2 and CO changed dramatically during the COVID-19 shelter-in-place. We combine observations in the SF Bay Area from ~30 nodes in the BErkeley Air quality and CO₂ Network (BEACO₂N) with measurements from ~1800 sites in the Caltrans Performance Measurement System (PeMS), a real-time traffic monitoring system reporting both car and truck flow to assess traffic related emissions inventories. Using PeMS flow rates, we estimate a decrease in vehicle miles traveled (VMT) to be 41% for cars and 34% for trucks from the six weeks before to six weeks after shelter in place orders were issued. Mapping these emissions estimates, we observe both spatial and temporal heterogeneity in the decline and recovery of emissions across the region. We compare an a priori emissions inventory based on historical measures of traffic and emission factors to emissions estimates constrained by PeMS observations and estimates of 2020 emission factors. We also compare the spatial allocation of CO2 emission changes derived from a formal inversion of the BEACO2N data (Turner et al, submitted manuscript) with the spatial allocation of changes in VMT reported by the PeMS network.