Session Presentations

Impact of COVID-19 related shutdown on onroad and aircraft transportation emissions and their impacts on O3 and PM2.5 in the U.S. using Sensitivity Modeling Techniques
Saravanan Arunachalam, Calvin Arter and Gavendra Pandey
Institute for the Environment, University of North Carolina at Chapel Hill

During the peak of the pandemic, 3.6 billion humans went into hiding globally. This led to drastic reductions in economic activity, specifically related to fossil fuel combustion focused on the transportation sector. Daily global CO2 emissions decreased by 17% by April compared to 2019 levels, with onroad transportation sector seeing a 50% decrease and the aviation sector seeing a 75% decrease. Within the U.S., these reductions vary by state and based on when social distancing measures were put in place. In the onroad emissions sector, while passenger vehicle activity reduced a lot due to large segment of the society working from home, vehicular activity due to freight did not see as much of a reduction. In fact, in some urban areas of the U.S., freight activity actually saw a modest increase compared to pre-pandemic period. Overall, weekly reductions in onroad activity during the pandemic was as high as 48% nationally, with individual states ranging from 35% (Arkansas) to 63% (New Jersey) during specific weeks. Similarly, U.S. wide aviation operations saw steep declines as well (> 70%), with some of the largest airports like Atlanta Hartsfield and New York John F. Kennedy seeing reductions of > 75 - 80% compared to pre-pandemic period. With such large scale reductions in transportation - related activity and their associated emissions, we find motivation to assess the impacts of these on ambient air quality and potential gains in public health due to reduced air pollution exposures. We use CMAQ instrumented with the Decoupled Direct Method (DDM) to compute sensitivities of predicted O3 and PM2.5 from individual states and airports to three input variables - light duty vehicles, heavy duty vehicles and aircraft operations. We will show illustrative results from select states and airports to show how O3 and PM2.5 and associated health risk changed due to reduced activity from the transportation sector. We also use information from this forced natural experiment to assess the ability of the model to capture such changes compared to observations.

Evaluating Correspondence of Changes in Indicators of Driving Behavior and TROPOMI Tropospheric Nitrogen Dioxide Column Density during COVID-19 in US Cities

Scott Natter, Drexel University, Computer Science, Philadelphia, Pennsylvania, USA

Shannon Capps, Drexel University, Civil, Architectural, and Environmental Engineering, Philadelphia, Pennsylvania, USA

In this study, we evaluate how tropospheric nitrogen dioxide (NO$$_2$$) column number densities, as reported by TROPOMI, have changed as a result of COVID-19 responses and policies. We use high temporal and spatial resolution data indicative of human behavior including mobility and driving direction requests as representative of the transportation sector, which contributes more than 50% of emissions of oxides of nitrogen. Specifically, we analyze how these representative values vary with NO$$_2$$ column density changes over the pre-pandemic and pandemic period. As we are analyzing the reduction in passenger vehicle emissions, the geographic foci are urban Combined Statistical Areas where traffic is typically more dense. Policy indicators such as stay-at-home orders are also considered. Though meteorology also plays a role in NO$$_2$$ column densities, clear correspondence between indicators of driving behavior and this pollutant are identified in select cities.

Disentangling the impact of the COVID-19 lockdowns on urban NO2 from natural variability

Daniel Goldberg, Susan Anenberg, Gaige Kerr, Debora Griffin, Chris McLinden, Zifeng Lu, David Streets

TROPOMI satellite data show substantial drops in nitrogen dioxide (NO2) during COVID-19 physical distancing. To attribute NO2 changes to NOX emissions changes over short timescales, one must account for meteorology. We find that meteorological patterns were especially favorable for low NO2 in much of the U.S. in spring 2020, complicating comparisons with spring 2019. Meteorological variations between years can cause column NO2 differences of ~15% over monthly timescales. After accounting for sun angle and meteorological considerations, we calculate that NO2 drops ranged between 9.2 - 43.4% among twenty cities in North America during spring 2020, with a median of 21.6%. Of the studied cities, largest NO2 drops (>30%) were in San Jose, Los Angeles, and Toronto, and smallest drops (<12%) were in Miami, Minneapolis, and Dallas. These normalized NO2 changes can be used to highlight locations with greater activity changes and better understand the sources contributing to adverse air quality in each city.

Time-resolved emission reductions for atmospheric chemistry modelling in Europe during the COVID-19 lockdowns

Marc Guevara¹, Oriol Jorba¹, Albert Soret¹, Herve Petetin¹, Dene Bowdalo¹, Kim Serradell¹, Carles Tena¹, Hugo Denier van der Gon², Jeroen Kuenen², Vincent-Henri Peuch³, and Carlos Perez Garcia-Pando^1,4

¹Earth Sciences Department, Barcelona Supercomputing Center, Barcelona, 08034, Spain

²TNO, Department of Climate, Air and Sustainability, Utrecht, the Netherlands

³European Centre for Medium-Range Weather Forecasts, Reading, UK

⁴ICREA, Catalan Institution for Research and Advanced Studies, 08010 Barcelona, Spain

To control the spread of the COVID-19 disease, European governments implemented emergency measures going from light social distancing to strict lockdowns, depending on the country. As a consequence, many industries, businesses and transport networks were forced to either close down or drastically reduce their activity, which resulted in an unprecedented drop of anthropogenic emissions. This work presents the Copernicus Atmosphere Monitoring Service (CAMS) European regional emission reduction factors associated to the COVID-19 lockdowns (CAMS-REG_ERF-COVID19), an open source dataset of daily-, sector- and country-dependent emission reduction factors for Europe linked to the COVID-19 restrictions. The resulting dataset covers a total of six emission sectors, including: road transport, energy industry, manufacturing industry, residential and commercial combustion, aviation and shipping. The time period covered by the dataset starts from 21 February, when the first European localised lockdown was implemented in the region of Lombardy (Italy), and is updated on a regular basis to provide the most recent information. The CAMS-REG_ERF-COVID19 dataset is based on a wide range of information sources and approaches, including open access and measured activity data and meteorological data, as well as the use of machine learning techniques. We combine the computed emission reduction factors with the Copernicus CAMS European gridded emission inventory to spatially and temporally quantify reductions in emissions from both criteria pollutants and greenhouse gases, as well as to assess the contribution of each pollutant sector and country to the overall reductions. The resulting gridded and time-resolved emission reductions are used to perform an air quality modelling study to evaluate its potential on reproducing observed air pollutant concentration changes in selected regions across Europe. Clear improvements of the modelled results are found when considering the emission reduction factors.

Traffic, Precursor Emissions, and Ozone in the South Coast Air Basin During California's COVID-19 Shutdown

Cesunica E. Ivey^1,2, Ziqi Gao³, Shams Tanvir², Khanh Do^1,2, Arash Kashfi Yeganeh^1,2,  Matt Barth², Armistead Russell³, Charles L. Blanchard⁴, Sang-Mi Lee⁵

1. Department of Chemical and Environmental Engineering, University of California, Riverside, Riverside, CA
2. Center for Environmental Research and Technology, Riverside, CA
3. Department of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA
4. Envair, Albany, CA
5. South Coast Air Quality Management District, Diamond Bar, CA
   

In the spring of 2020, unprecedented shifts in human activity in response to the COVID-19 pandemic led to observable changes in the natural environment, specifically air pollutant concentrations. In Spring of 2020, the South Coast Air Basin of California (USA) experienced noticeable declines in on-road activity and primary traffic-related pollutant emissions. However, secondary ozone concentration trends were not consistent across the basin. The upwind site in Pasadena, CA experienced overall increases in maximum daily 8-hour ozone (MDA8) during the shutdown, whereas the downwind site in Crestline, CA experienced an overall decrease in MDA8. Typically, the highest MDA8 concentrations are observed at locations downwind of the Los Angeles city center, indicating a shift in the spatial peak of ozone production due to major decreases in precursor emissions during the COVID-19 shutdown. Higher temperatures in late April led to higher than average MDA8 concentrations in both locations. Further, by aggregating traffic-related emissions at census tract level, we related air quality to local transportation emissions and disproportionate impacts on environmental justice communities. The COVID-19 shutdown provided a preview of the potential impacts of large-scale emissions reductions on ozone formation in the South Coast Air Basin.

Air Quality Changes under COVID-19 Social Distancing in the United States: Observational Analysis and Modeling Sensitivity Study

Daiwen Kang, Christian Hogrefe, Rohit Mathur, Benjamin Murphy, Vlad Isakov, George Pouliot, Robert Gilliam, Fahim Sidi, and Tanya Spero

The rapid spread of coronavirus disease 2019 (COVID-19), declared as a pandemic by the World Health Organization (WHO), prompted the national and local governments around the world to take unprecedented measures to restrict human activity such as lockdowns, social distancing, and strict travel bans. The dramatic and varying scales of responses to COVID-19 offer a rare opportunity to assess the impacts of real-world reductions in emissions on local, regional and global air quality and to investigate the relative contributions of different source sector emission perturbations. In this study, fine particulate matter (PM_2.5) and ozone (O₃) measured from U.S. EPA's AIRNOW network before and after the start of social distancing are analyzed to assess the change in air quality during the period of social distancing. The temporal and spatial variations will be analyzed in response to publicly available mobility data collected since the start of social distancing. Another objective of this study is to examine how predictions of various air pollutant concentrations from the Community Multiscale Air Quality (CMAQ) modeling system simulates the estimated reduction of anthropogenic emissions induced as a result of changing activity patterns. Sensitivity studies based on the 2016 modeling platform are performed by incorporating real-time observed mobility data such as Google mobility data and vehicle miles traveled (VMT) data through CMAQ's Detailed Emission Scaling Isolation and Diagnostic (DESID) module and analyzed to develop an initial estimate of the likely impacts. The strengths and weaknesses of the modeling system in response to quick and dramatic emission changes from different sectors will be identified and presented.

DISCLAIMER

The views expressed in this paper are those of the authors and do not necessarily represent the views or policies of the U.S. EPA.

Association between the exposure to long-term PM2.5 concentration and COVID-19 death rate at the provincial level in Iran

Hossein Khajehpour, Farzaneh Taksibi, Mansour Shamsipour, Mohammad Sadegh Hassanvand

In this study, we investigated the association between exposure to long-term ambient PM2.5 and the COVID-19 death rate at the sub-national (provincial) levels in Iran. Daily averaged PM2.5 concentrations from the ambient air quality monitoring stations were collected and processed to calculate the annually-averaged concentrations. Also, the most effective number of lag days is studied to account for the most consistent mortality percentage of the infected cases hospitalized due to COVID-19. Then the resulted death percentage and the long-term exposures are regressed to check the correlation among the variables. Our results support the conclusion of the previous research works in other countries that the long-term exposure to higher concentrations of PM2.5 increases the risk of the lethality of the COVID-19 cases.

Abrupt decline in tropospheric nitrogen dioxide after the outbreak of COVID-19

Fei Liu, Joanna Joiner, Bryan Duncan, Nickolay A. Krotkov, Lok N. Lamsal, Aaron Page, Sarah A. Strode, Yasuko Yoshida, Sungyeon Choi, Bo Zheng, Can Li, Henk Eskes, Ronald van der A, Pepijn Veefkind, Pieternel F. Levelt, Oliver P. Hauser, and the OMI NO2 team

To combat the spread of COVID-19, government sealed off cities reporting large numbers of infected people, starting January 23, 2020 in Wuhan, China; this included halting public transportation and closing local businesses. Policy interventions to reduce the spread of COVID-19 have environmental and economic impacts. Tropospheric nitrogen dioxide (NO2) indicates economic activities, as nitrogen dioxide is primarily emitted from fossil fuel consumption. Satellite measurements show dramatic drop in tropospheric NO2 vertical column densities after the outbreak of COVID-19. We estimate the impact of COVID-19 control measures on global anthropogenic emissions by taking advantages of the changes of observed NO2 from space. We also relate to this reduction to government's actions, including the announcement of the first report and the date of lockdown. Our analysis offers insights into the unintended environmental and economic consequences through reduced economic activities.

Simulation of particulate matter over mega-cities of India during COVID-19 : Preliminary results

Sujit Maji*^, Gufran Beig*, Saroj K. Sahu**

* Indian Insititute of Tropical Meteorology, Pune, India, 411008 ** Utkal University, Bhubaneswar, Odisha, India, 751004 ^ email address: sujit.cat@tropmet.res.in

With the advent of COVID-19 strong control measures had been taken (by applying lock-down) which significantly affected the vehicular, industrial, and other outdoor activities resulting in reduced air pollution. We have calculated percentage change of pollutants during lock-down from observation and implemented a percent reduction for different pollutants in line of that in emission inventory. We have simulated PM2.5 (particulate matter of aerodynamic radius less than 2.5 micron) over India at 12 Km horizontal resolution using WRF-Chem (Weather and Research Forecast model ) version 3.9.1 during 25th Mar 2020 - 14th Apr 2020 for two scenarios (a) normal emission , and (b) reduced emission during lock-down due to COVID-19. Two different sets of meteorology : FNL with a horizontal resolution of 1 degree, and IITMGEFS with a horizontal resolution of 12 Km have been utilized for comparison. Model outputs were evaluated with the observation data from SAFAR (System of Air Quality Forecasting and Research) observational network over four cities : Delhi, Ahmedabad, Mumbai and Pune. Model simulation using the reduced emission captures the observed reduction and pattern of PM2.5 concentration well. Site to site comparison indicates possibility of stepwise implementation of lock-down across different areas. Calculated statistical metrics  shows higher Index of agreement (IOA) values at some stations (> 0.5) compared to the all stations (< 0.5). This is due to inability of model to capture spatial heteorologenity/emission variability at very high resolution. Mean biases indicates model have well predicted PM2.5 for the whole period (1.1 at Ahmedabad, -5.48 at Mumbai  and, -9.7 at Pune). Model simulation using normal emission have over-predicted the observed values and differed significantly from predicted values using reduced emission. From our above study we can conclude: (1) There was a significant reduction in particulate pollutants during implementation of first lock-down (25 March -14 April 2020) over India and, (2) Our calculated emission inventory represents the reduced emission over mega-cities.

Quantifying the impact of COVID-19 on air quality in major Canadian cities

Rabab Mashayekhi¹, Radenko Pavlovic¹, Jacinthe Racine¹, Michael D. Moran², Patrick M. Manseau¹, Annie Duhamel¹, Ali Katal¹, Jessica Miville¹, David Niemi¹, Si Jun Peng¹, Mourad Sassi¹, Debora Griffin² and Chris McLinden²

¹ Air Quality Policy-Issue Response Section, Canadian Meteorological Center, Environment and Climate Change Canada, Dorval, Quebec, Canada

² Air Quality Research Division, Environment and Climate Change Canada, Toronto, Ontario, Canada

In this study, we have investigated the impact of reduced activity levels due to COVID-19 lockdown measures on air quality in four major Canadian cities: Montreal, Toronto, Calgary, and Vancouver. First, near-real-time hourly surface measurements of NO₂, PM_2.5, and O₃ were analysed for two periods, a "pre-lockdown" period from 15 February to 14 March 2020, and a "lockdown" period from 22 March to 2 May 2020, during which lockdown measures were in full force everywhere in Canada. The observed mean diurnal concentrations during both periods in 2020 were compared to the same periods in the previous decade (2010-2019). Higher-than-usual seasonal declines were observed in average NO₂ from the pre-lockdown to lockdown periods in 2020. For PM_2.5 Montreal was the only city that had a higher-than-usual seasonal decline, while for O₃ all four cities remained within the previous decadal range. In order to isolate the impact of lockdown-related emissions changes from other factors such as seasonal changes in meteorology and emissions and meteorological variability, a modelling scenario analysis was conducted with the GEM-MACH air quality model for the lockdown period on a 10-km grid that covered most of Canada and the U.S. Two simulations were performed and compared, the first a counterfactual simulation with Business-As-Usual (BAU) baseline emissions and the second a simulation with "lockdown scenario" emissions, which were based on an estimated 80% decrease in air traffic, 60% decrease in on-road transportation, 30% decrease in industrial activity, and 20% increase in emissions from residential heating. Compared to the BAU simulation, NO₂ surface concentrations in the four metropolitan areas for the COVID-19 emissions scenario were lower by 31% to 33% on average. For PM_2.5 surface concentrations less pronounced average decreases ranging from 6% to 17% were predicted. Maximum decreases of up to 41% for NO₂ and 22% for PM_2.5 were predicted for grid cells near major airports and highway intersections or close to city centers. O₃ surface concentrations, on the other hand, showed increases up to a maximum of 18% close to city centers and slight decreases over the suburbs, while O_x (odd oxygen) decreased over these cities.

Impacts of COVID-19 Pandemic on Activity and Emissions

Zach Rewolinski, James Li, Anthony Nguyen, Daniel R. Kowal, and Daniel S. Cohan

Responses to the COVID-19 pandemic have led to major declines in economic activity, but there is a lack of research quantifying how this has impacted activity and emissions on a sector-by-sector basis. Analyzing the changes in activity and emissions across sectors allows us to comprehend the unique impact that COVID-19 has had on the United States. Moreover, by examining local policies which have impacted emissions in specific regions, we can gain knowledge regarding the relationships between restrictions, transportation, power generation and consumption, and associated emissions on a local level. By bringing these relationships to light, we can offer insight into why emissions are changing, and how to lower emissions further.

In this study, we use publicly available data that describes power plant activity, electricity demand, and vehicle activity to quantify changes in emissions and energy consumption across various regions of the United States. We incorporate data on policies and restrictions enacted in order to examine how the timing and stringency of those policies may have influenced activity and emissions. In comparing 2020 data to historical data, we find that 1) vehicle activity dropped sharply and then rebounded; 2) electricity demand fell in the industrial and commercial sectors but rose for residential consumers; and 3) power plant emissions began falling even before the pandemic due to diminished coal use amid low natural gas prices and a growing prevalence of renewable energy.

Effects of COVID-19 on Transportation and Air Quality in the North Central Texas Region

Nicholas Van Haasen and Vivek Thimmavajjhala

The novel coronavirus, COVID-19, was first confirmed in North Texas on March 9, 2020. By March 24, 2020, counties (Collin, Dallas, Denton, and Tarrant) issued orders to mitigate the spread of COVID-19. Some independent school districts were closed indefinitely. By the end of March, the governor issued a one-month executive order that only allowed Texans to leave their homes for essential activities. On May 1, 2020, the region started to reopen businesses. Later in June, as confirmed COVID-19 cases continued to increase, the reopening was systematically halted and slowly started to reverse. The effects of such a closure on the economy, the constraints placed on individuals to social distance, and the stay-at-home orders all culminated in profoundly impacting the North Central Texas region. We, NCTCOG, monitored the impacts of the COVID-19 pandemic on our regional transportation and air quality. We collected and analyzed several metrics including, but not limited to, highway traffic volume, vehicle speed, and crash statistics; transit ridership, airport passengers, bike/pedestrian; and air quality levels. Since the North Central Texas region is in nonattainment for the pollutant ozone, tracking ozone measures is a key component of our current and ongoing analysis, which we plan to continue in the forthcoming months.

The Improvement of the Air Quality due to Traffic Halting in Los Angeles during the COVID-19 Outbreak

Jiani Yang ^a,*, Yuan Wang ^a,b, Joseph Pinto ^c

^a Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, 91125 USA

^b Jet Propulsion Laboratory,4800 Oak Grove Dr, Pasadena, CA 91109 USA

^c Environmental Sciences and Engineering, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, 135 Dauer Drive, Chapel Hill, NC 27599 USA

Correspondence to: Jiani Yang (yjn@caltech.edu)

Background: On March 19, 2020, the government of California ordered all 40 million Californians to stay at home in the coming weeks as the result of the escalation of the coronavirus disease 2019 (COVID-19) pandemic. As lockdowns were implemented, the significant changes caused by these restrictions brought the dramatic improvement in air quality in metropolitan cities such as Los Angeles (LA Basin).

Methods: We use real-time data from The South Coast Air Quality Management District (South Coast AQMD), and California Department of Transportation (PeMS) to evaluate the anthropogenic drivers of the pollution sources. We fit the regression analysis to compare the correlation of 7 variables including traffic flow, truck flow, traffic speed, NO₂, CO, PM_2.5, and O₃. We also mapped the monthly spatial variation and hourly heatmap of those 7 variables in 2020 to understand the impacts of the lockdown on different locations in the LA Basin.

Results: In the Los Angeles Basin, the traffic flow on highways started to drop an intensely by 20.86 % when initiating the stay at home order and it continued decreasing by 32.92% the first week, 30.94% the second week, 37.9% the third week, and 33.57% at the fourth week following the lockdown compared to corresponding dates during the year of 2019. The average truck flow for each sensor is generally higher in 2020 than 2019 before lockdown. The general weekly drops have also been observed for the truck flow on the highways by 1.63% the first week, then it raises back to 18.31% the second week, and declined again by 10.98% the third week, and declined by 2.55% at the fourth week following the county-wide lockdown. Accordingly, the change of traffic trigged the intensive decline of NO₂ by 44.23% the first week, 12.96% the second week, 50.64% the third week, 32.65% the fourth week following the lockdown; We found a dramatic drop in PM_2.5, NO₂, CO during the first week after initiating the stay at home order. The correlation (Pierson r) between truck flow change and changes of NO₂, CO, and PM_2.5 is 0.91(****),0.88(****),0.74(**); The correlation between traffic flow change and changes of NO₂ is 0.87(****), CO is 0.81(***), and PM_2.5 is 0.62(**). The correlation between traffic speed change and changes of NO₂ is -0.84(****), CO is -0.78 (***), and PM_2.5 is -0.59(*). We found that a decline of 1% in NO₂, CO and PM_2.5 is associated with the decline of 15.79%, 17.15% and 9.43% in the truck flow; A decline of 1% in NO₂, CO and PM_2.5 is associated with the decline of 11.26%, 9.43% and 20.96% in traffic flow; A decline of 1% in NO₂, CO and PM_2.5 is associated with the increase of 3.45%, 3.13% and 5.21% in traffic speed. The results are all statistically significant.

Conclusion: The drop in truck flow is mainly responsible for the drop of NO₂ and CO. The lockdowns provided a large-scale experiment into air quality research. The result of this research would provide an important reference for the policy markers regarding truck management in light of air quality control to prepare the 2028 Summer Olympics in LA.