Abstract: Point sources are often major contributors of greenhouse gas and air pollutant emissions in urban areas. Dense air monitoring networks provide a unique avenue for studying point source emissions over long time periods. Here, we use the Berkeley Environmental Air-quality and CO2 Network (BEACO2N) to study CO2 and air pollutant emissions from an oil refinery in the city of Richmond, CA. We identify 266 plumes crossing one or more sites in the BEACO2N network during 2022–2023 as having a source at the refinery and quantify CO2 emissions using the Gaussian plume model. The refinery is modeled as two point sources, and total CO2 emissions are found to be 61.3 ± 6.3 kg s−1, in close agreement with the EPA Facility Level Information on GreenHouse gases Tool inventory and Carbon Mapper measurements. Additionally, plume composition was found to vary, with CO/CO2 enhancement ratios ranging from 0 to 5 ppb/ppm. Taking the average CO/CO2 ratio, CO emissions are estimated to be roughly 3,000 mt, a sizable amount compared to the city's estimated non-refinery CO emissions of 6,800 mt. In addition, we show that dense monitoring networks provide a unique resource for point source emissions quantification with broad application to plumes of varying size and composition, and recommendations are laid out for future applications.Link to Article

Abstract: We present the seasonal variations of enhancement ratios (ERs, i.e., ΔNOx/ΔCO2 and ΔCO/ΔCO2) as a function of distance from highways in the San Francisco Bay Area, using observations from the Berkeley Environmental Air Quality and CO2 Network (BEACO2N) at 40 locations. The spatial patterns exhibit exponential distance-decay relationships, with higher NOx and CO ERs near highways and more uniform ERs at distances beyond 3 km. These patterns are used to infer emission factors (EFs) for transportation and residential buildings. BEACO2N-derived EFs for CO (7.8 ± 0.6 ppbv/ppmv) and NOx (1.0 ± 0.02 ppbv/ppmv) from transportation agree with inventory estimates. In contrast, the residential NOx EF (0.15 ± 0.01 ppbv/ppmv) is four times lower than inventory estimates, and the residential CO EF (4.3 ± 0.3 ppbv/ppmv) is 33% lower than the California state inventory estimate.Link to Article

Abstract: Home heating preferences vary dramatically with regional climate. The temperature at which residents turn on natural gas home heating systems (critical temperature) varies by as much as 25°C from the northern to southern United States (U.S.). Here we derive temperature dependent CO2 emissions in three U.S. cities using a dense ground-based CO2 observation network. A Bayesian inverse modeling methodology is used to update a 1-km emission inventory in each of the three cities. This method is able to correctly identify the critical temperature of home heating even when this information is withheld from the prior inventory, as verified by natural gas distribution data. Variance in regional heating practices has not been previously demonstrated with ground-based networks of CO2 observations. This result provides evidence that a Bayesian inverse modeling framework is sensitive to emissions of the home heating sector.Link to Article

Abstract: Over the past decade, technological advancements and the growing demand for higher spatial resolution in air quality measurements have driven the widespread deployment of low-cost electrochemical gas sensors. These sensors are inexpensive, compact, easy to deploy, and have potential to detect variability in ambient air. While short-term sensor performance and calibration assessments have been well-documented, research on sensor reliability in long-term deployments remains limited. The Berkeley Environmental Air Quality and CO2 Network (BEACO2N) is a dense, long-term sensor network measuring O3, NO, NO2, and CO using electrochemical gas sensors, in addition to CO2 and PM2.5 with other sensors. Here, we evaluate the performance of the Alphasense NO2-B43F, Ox-B431, NO-B4, and CO-B4 sensors over the course of 3 years in the San Francisco Bay Area. We find consistent performance in the calibrated data over the 3 years. These findings highlight the reliability of long-term sensor use with appropriate calibration techniques and suggest the sensors remain suitable for extended deployments without requiring replacement.Link to Article

Abstract: Densely spaced sensor networks provide a unique opportunity for describing emissions from stationary and moving point sources and from intermittent events like fires or industrial flaring. As an example of what sensor networks can achieve, we describe quantification of emissions from a small urban fire in the Bay Area of California using the Berkeley Environmental Air-quality and CO2 Network (BEACO2N), a dense air quality and greenhouse gas monitoring network. Pollutant enhancements are measured at multiple sites, and the ensemble of observations are fit to a 2-D Gaussian model to characterize the extent of dilution prior to observation and derive emissions at the location of the fire. Distinct approaches are used for calibration of the CO2, air quality gases, and PM2.5 instruments. Consistency of the ratios at multiple locations downwind of the fire supports the precision of the network. We find that the fire emitted approximately 770 ± 30 kg of PM2.5, 70,000 ± 20,000 kg of CO2, 2500 ± 300 kg of CO, and 28 ± 9 kg of NOx. The emission ratios are in the range of typical wildland fires. Using this example, we explore the minimum plume emissions that could be observed and quantified by the network.Link to Article

Abstract: Urban areas are major contributors to greenhouse gas emissions, necessitating effective monitoring systems to evaluate mitigation strategies. A dense sensor network, such as the Berkeley Environmental Air-quality & CO2 Observation Network (BEACO2N), offers a unique opportunity to monitor urban emissions at high spatial resolution. Here, we describe a simple approach to quantifying urban emissions with sufficient precision to constrain seasonal and annual trends. Measurements from 12 BEACO2N sites in Los Angeles (called the USC Carbon Census) are analyzed within a box model framework. By combining CO2 and CO observations, we partition total CO2 emissions into fossil fuel and biogenic emissions. We infer temporal changes in biogenic emissions that correspond to the MODIS enhanced vegetation index (EVI) and show that net biogenic exchange can consume up to 60% of fossil fuel emissions in the growing season during daytime hours. While we use the first year of observations to describe seasonal variation, we demonstrate the feasibility of this approach to constrain annual and longer trends.Link to Article

Abstract: Deployment of large numbers of low capital cost sensors to increase the spatial density of air quality measurements enables applications that build on mapping air at neighborhood scales. Effective deployment requires not only low capital costs for observations but also a simultaneous reduction in labor costs. The Berkeley Environmental Air Quality and CO2 Network (BEACO2N) is a sensor network measuring O3, CO, NO, and NO2, particulate matter (PM2.5), and CO2 at dozens of locations in cities where it is deployed. Here, we describe a low labor cost in situ field calibration for the BEACO2N O3, CO, NO, and NO2 sensors. This method identifies and leverages uniform periods in concentrations across the network for calibration. The calibration achieves high accuracy and low biases with respect to temperature, humidity, and concentration, with coefficients of determination and root mean square errors of 0.88 and 3.70 ppb for O3, 0.66 and 3.16 ppb for NO2, and 0.79 and 1.58 ppb for NO. Performance of the CO sensor is 0.90 and 33.3 ppb at a site colocated with reference measurements. The method is a crucial step toward lowering operational costs of delivering accurate measurements in dense networks employing large numbers of inexpensive air quality sensors.Link to Article

Abstract: Cities represent a significant and growing portion of global carbon dioxide (CO2) emissions. Quantifying urban emissions and trends over time is needed to evaluate the efficacy of policy targeting emission reductions as well as to understand more fundamental questions about the urban biosphere. A number of approaches have been proposed to measure, report, and verify (MRV) changes in urban CO2 emissions. Here we show that a modest capital cost, spatially dense network of sensors, the Berkeley Environmental Air Quality and CO2 Network (BEACO2N), in combination with Bayesian inversions, result in a synthesis of measured CO2 concentrations and meteorology to yield an improved estimate of CO2 emissions and provide a cost-effective and accurate assessment of CO2 emissions trends over time. We describe nearly 5 years of continuous CO2 observations (2018–2022) in a midsized urban region (the San Francisco Bay Area). These observed concentrations constrain a Bayesian inversion that indicates the interannual trend in urban CO2 emissions in the region has been a modest decrease at a rate of 1.8 ± 0.3%/year. We interpret this decrease as primarily due to passenger vehicle electrification, reducing on-road emissions at a rate of 2.6 ± 0.7%/year. Link to Article

Abstract: Low-cost particulate matter (PM) sensors continue to grow in popularity, but issues such as aerosol-size-dependent sensitivity drive the need for effective calibration schemes. Here we devise a time-evolving calibration method for the Plantower PMS5003 PM2.5 mass concentration measurements. We use 2 years of measurements from the Berkeley Environmental Air-quality and CO2 Network sensors deployed in San Francisco and Los Angeles in our analysis. The calibration uses a hygroscopic growth correction factor derived from κ-Köhler theory, where the calibration parameters are determined empirically using US Environmental Protection Agency Air Quality System (EPA AQS) reference data at co-location sites during the period from 2021–2022. The parameters are found to vary cyclically through the seasons, and the seasonal cycles match changes in sulfate and elemental carbon PM composition fractions throughout the year. In both regions, the seasonal RH dependence calibration performs better than the uncalibrated data and data calibrated with the EPA's national Plantower calibration algorithm. In the San Francisco Bay Area, the seasonal RH dependence calibration reduces the root mean square error (RMSE) by ∼40 % from the uncalibrated data and maintains a mean bias much smaller than the EPA national calibration scheme (−0.90 vs −2.73 µg m−3). We also find that calibration parameters forecasted beyond those fit with the EPA reference data continue to outperform the uncalibrated data and EPA calibration data, enabling real-time application of the calibration scheme even in the absence of reference data. While the correction greatly improves the data accuracy, non-Gaussian distribution of the residuals suggests that other processes besides hygroscopic growth can be parameterized for future improvement of this calibration. Link to Article

Abstract: Transportation emissions are the largest individual sector of greenhouse gas (GHG) emissions. As such, reducing transportation-related emissions is a primary element of every policy plan to reduce GHG emissions. The Berkeley Environmental Air-quality and CO2 Observation Network (BEACO2N) was designed and deployed with the goal of tracking changes in urban CO2 emissions with high spatial (∼1 km) and temporal (∼1 hr) resolutions while allowing the identification of trends in individual emission sectors. Here, we describe an approach to inferring vehicular CO2 emissions with sufficient precision to constrain annual trends. Measurements from 26 individual BEACO2N sites are combined and synthesized within the framework of a Gaussian plume model. After removing signals from biogenic emissions, we are able to report normalized annual emissions for 2018–2020. A reduction of 7.6 ± 3.5% in vehicular CO2 emissions is inferred for the San Francisco Bay Area over this 2 year period. This result overlaps with, but is slightly larger than, estimates from the 2017 version of the California Air Resources Board EMFAC emissions model, which predicts a 4.7% decrease over these 2 years. This demonstrates the feasibility of independently and rapidly verifying policy-driven reductions in GHG emissions from transportation with atmospheric observations in cities. Link to Article

Abstract: Transportation represents the largest sector of anthropogenic CO2 emissions in urban areas in the United States. Timely reductions in urban transportation emissions are critical to reaching climate goals set by international treaties, national policies, and local governments. Transportation emissions also remain one of the largest contributors to both poor air quality (AQ) and to inequities in AQ exposure. As municipal and regional governments create policy targeted at reducing transportation emissions, the ability to evaluate the efficacy of such emission reduction strategies at the spatial and temporal scales of neighborhoods is increasingly important; however, the current state of the art in emissions monitoring does not provide the temporal, sectoral, or spatial resolution necessary to track changes in emissions and provide feedback on the efficacy of such policies at the abovementioned scale. The BErkeley Air Quality and CO2 Network (BEACO2N) has previously been shown to provide constraints on emissions from the vehicle sector in aggregate over a ∼ 1300 km2 multicity spatial domain. Here, we focus on a 5 km, high-volume, stretch of highway in the San Francisco Bay Area. We show that inversion of the BEACO2N measurements can be used to understand two factors that affect fuel efficiency: vehicle speed and fleet composition. The CO2 emission rate of the average vehicle (in grams per vehicle kilometer) is shown to vary by as much as 27 % at different times of a typical weekday because of changes in these two factors. The BEACO2N-derived emission estimates are consistent to within ∼ 3 % of estimates derived from publicly available measures of vehicle type, number, and speed, providing direct observational support for the accuracy of the EMission FACtor model (EMFAC) of vehicle fuel efficiency.Link to Article

Abstract: The majority of global anthropogenic CO2 emissions originate in cities. We have proposed that dense networks are a strategy for tracking changes to the processes contributing to urban CO2 emissions and suggested that a network with ∼ 2 km measurement spacing and ∼ 1 ppm node-to-node precision would be effective at constraining point, line, and area sources within cities. Here, we report on an assessment of the accuracy of the Berkeley Environmental Air-quality and CO2 Network (BEACO2N) CO2 measurements over several years of deployment. We describe a new procedure for improving network accuracy that accounts for and corrects the temperature-dependent zero offset of the Vaisala CarboCap GMP343 CO2 sensors used. With this correction we show that a total error of 1.6 ppm or less can be achieved for networks that have a calibrated reference location and 3.6 ppm for networks without a calibrated reference. Link to Article

Abstract: Governments restricted mobility and effectively shuttered much of the global economy in response to the COVID‐19 pandemic. Six San Francisco Bay Area counties were the first region in the United States to issue a “shelter‐in‐place” order asking non‐essential workers to stay home. Here we use CO2 observations from 35 Berkeley Environment, Air‐quality and CO2 Network (BEACO2N) nodes and an atmospheric transport model to quantify changes in urban CO2 emissions due to the order. We infer hourly emissions at 900‐m spatial resolution for 6 weeks before and 6 weeks during the order. We observe a 30% decrease in anthropogenic CO2 emissions during the order and show that this decrease is primarily due to changes in traffic (–48%) with pronounced changes to daily and weekly cycles; non‐traffic emissions show small changes (–8%). These findings provide a glimpse into a future with reduced CO2 emissions through electrification of vehicles. Link to Article

Abstract: Urban carbon dioxide comprises the largest fraction of anthropogenic greenhouse gas emissions, but quantifying urban emissions at subnational scales is highly challenging, as numerous emission sources reside in close proximity within each topographically intricate urban dome. In attempting to better understand each individual source's contribution to the overall emission budget, there exists a large gap between activity-based emission inventories and observational constraints on integrated, regional emission estimates. Here we leverage urban CO2 observations from the BErkeley Atmospheric CO2 Observation Network (BEACO2N) to enhance, rather than average across or cancel out, our sensitivity to these hyperlocal emission sources. We utilize a method for isolating the local component of a CO2 signal that accentuates the observed intra-urban heterogeneity and thereby increases sensitivity to mobile emissions from specific highway segments. We demonstrate a multiple-linear-regression analysis technique that accounts for boundary layer and wind effects and allows for the detection of changes in traffic emissions on scale with anticipated changes in vehicle fuel economy – an unprecedented level of sensitivity for low-cost sensor technologies. The ability to represent trends of policy-relevant magnitudes with a low-cost sensor network has important implications for future applications of this approach, whether as a supplement to existing, sparse reference networks or as a substitute in areas where fewer resources are available. Link to Article

Abstract: The newest generation of air quality sensors is small, low cost, and easy to deploy. These sensors are an attractive option for developing dense observation networks in support of regulatory activities and scientific research. They are also of interest for use by individuals to characterize their home environment and for citizen science. However, these sensors are difficult to interpret. Although some have an approximately linear response to the target analyte, that response may vary with time, temperature, and/or humidity, and the cross-sensitivity to non-target analytes can be large enough to be confounding. Standard approaches to calibration that are sufficient to account for these variations require a quantity of equipment and labor that negates the attractiveness of the sensors' low cost. Here we describe a novel calibration strategy for a set of sensors, including CO, NO, NO2, and O3, that makes use of (1) multiple co-located sensors, (2) a priori knowledge about the chemistry of NO, NO2, and O3, (3) an estimate of mean emission factors for CO, and (4) the global background of CO. The strategy requires one or more well calibrated anchor points within the network domain, but it does not require direct calibration of any of the individual low-cost sensors. The procedure nonetheless accounts for temperature and drift, in both the sensitivity and zero offset. We demonstrate this calibration on a subset of the sensors comprising BEACO2N, a distributed network of approximately 50 sensor nodes, each measuring CO2, CO, NO, NO2, O3 and particulate matter at 10 s time resolution and approximately 2 km spacing within the San Francisco Bay Area. Link to Article

Abstract: The majority of anthropogenic CO2 emissions are attributable to urban areas. While the emissions from urban electricity generation often occur in locations remote from consumption, many of the other emissions occur within the city limits. Evaluating the effectiveness of strategies for controlling these emissions depends on our ability to observe urban CO2 emissions and attribute them to specific activities. Cost-effective strategies for doing so have yet to be described. Here we characterize the ability of a prototype measurement network, modeled after the Berkeley Atmospheric CO2Observation Network (BEACO2N) in California's Bay Area, in combination with an inverse model based on the coupled Weather Research and Forecasting/Stochastic Time-Inverted Lagrangian Transport (WRF-STILT) to improve our understanding of urban emissions. The pseudo-measurement network includes 34 sites at roughly 2 km spacing covering an area of roughly 400 km2. The model uses an hourly 1 × 1 km2 emission inventory and 1 × 1 km2 meteorological calculations. We perform an ensemble of Bayesian atmospheric inversions to sample the combined effects of uncertainties of the pseudo-measurements and the model. We vary the estimates of the combined uncertainty of the pseudo-observations and model over a range of 20 to 0.005 ppm and vary the number of sites from 1 to 34. Link to Article

Abstract: With the majority of the world population residing in urban areas, attempts to monitor and mitigate greenhouse gas emissions must necessarily center on cities. However, existing carbon dioxide observation networks are ill-equipped to resolve the specific intra-city emission phenomena targeted by regulation. Here we describe the design and implementation of the BErkeley Atmospheric CO2 Observation Network (BEACO2N), a distributed CO2 monitoring instrument that utilizes low-cost technology to achieve unprecedented spatial density throughout and around the city of Oakland, California. We characterize the network in terms of four performance parameters — cost, reliability, precision, and systematic uncertainty — and find the BEACO2N approach to be sufficiently cost-effective and reliable while nonetheless providing high-quality atmospheric observations. First results from the initial installation successfully capture hourly, daily, and seasonal CO2 signals relevant to urban environments on spatial scales that cannot be accurately represented by atmospheric transport models alone, demonstrating the utility of high-resolution surface networks in urban greenhouse gas monitoring efforts. Link to Article

Publications

M.Y. Patel, N.G. Asimow, A.R. Winter, Y. Zhu, and R.C. Cohen. Using Network Observations to Constrain CO and CO2 Emissions From an Oil Refinery in the San Francisco Bay Area, JGR Atmospheres, https://doi.org/10.1029/2025JD045035, 2026.

Y. Zhu, M.Y. Patel, A.R. Winter, N.G. Asimow, and R.C. Cohen. Observational Inferences of NOx and CO Emission Factors for Vehicles and Homes in the San Francisco Bay Area, ACS ES&T Air, https://doi.org/10.1021/acsestair.5c00004, 2025.

N.G. Asimow, M.Y. Patel, Y. Zhu, A.R. Winter, K.R. Gurney, W.M. Berelson, A.J. Turner, and R.C. Cohen. Differences in Regional Home Heating Behavior in Three U.S. Cities Revealed by Ground-Based Sensor Network, Geophysical Research Letters, https://doi.org/10.1029/2025GL115772, 2025.

A.R. Winter, Y. Zhu, N.G. Asimow, M.Y. Patel, and R.C. Cohen. Sustained Performance of Low-Cost Air Quality Sensors in Long-Term Deployments, ACS Sensors, https://doi.org/10.1021/acssensors.5c00566, 2025.

M.Y Patel, Y. Zhu, A.R. Winter, N.G. Asimow, and R.C. Cohen. Plume Detection and Emissions Quantification Potential Using a Dense Sensor Network, ACS ES&T Air, https://doi.org/10.1021/acsestair.5c00069, 2025.

J. Kim, W.M. Berelson,N.E. Rollins, N.G. Asimow, C. Newman, R.C. Cohen, J.B. Miller, B.C. McDonald, J. Peischl, and S.J. Lehman. Observing Anthropogenic and Biogenic CO2 Emissions in Los Angeles Using a Dense Sensor Network, Environ. Sci. Technol., https://doi.org/10.1021/acs.est.4c11392, 2025.

A.R. Winter, Y. Zhu, N.G. Asimow, M.Y. Patel, and R.C. Cohen. A Scalable Calibration Method for Enhanced Accuracy in Dense Air Quality Monitoring Networks, Env Sci & Tech., https://doi.org/10.1021/acs.est.4c08855, 2025.

N.G. Asimow, A.J. Turner, and R.C. Cohen, Sustained Reductions of Bay Area CO2 Emissions 2018–2022, Env Sci & Tech., https://pubs.acs.org/doi/10.1021/acs.est.3c09642, 2024.

M.Y. Patel, P.F. Vannucci, J. Kim, W.M. Berelson, and R.C. Cohen, Towards a Universal Hygroscopic Growth Calibration for Lost-Cost PM2.5 Sensors, Atmos. Meas. Tech., 17, 1051–1060, https://doi.org/10.5194/amt-17-1051-2024, 2024.

J. Kim, A.J. Turner, H.L. Fitzmaurice, E.R. Delaria, C. Newman, P.J. Wooldridge, and R.C. Cohen, Observing annual trends in vehicular CO2 emissions, Env Sci & Tech., https://doi.org/10.1021/acs.est.1c06828, 2022.

H.L. Fitzmaurice, A.J. Turner, J. Kim, K. Chan, E.R. Delaria, C. Newman, P.J Wooldridge, R.C. Cohen, Assessing vehicle fuel efficiency using a dense network of CO2 observations, Atmos. Chem. Phys., 22, 3891–3900, https://doi.org/10.5194/acp-22-3891-2022, 2022.

A.J. Turner, J. Kim, H.L. Fitzmaurice, C. Newman, K. Worthington, K. Chan, P.J. Wooldridge, P. Köhler, C. Frankenbergand R.C. Cohen, Observed impacts of COVID-19 on urban CO2 emissions, Geophys. Res. Lett., https://doi.org/10.1029/2020GL090037, 2020.

A.A. Shusterman, J. Kim, K.J. Lieschke, C. Newman, P.J. Wooldridge, and R.C. Cohen, Observing local CO2 sources using low-cost, near-surface urban monitors, Atmos. Chem. Phys. 18, 13773-13785, 2018.

J. Kim, A.A. Shusterman, K.J. Lieschke, C. Newman, and R.C. Cohen, The BErkeley Atmospheric CO2 Observation Network: field calibration and evaluation of low-cost air quality sensors, Atmos. Meas. Tech. 11, 1937-1946, 2018.

A.J. Turner, A.A. Shusterman, B.C. McDonald, V. Teige, R.A. Harley, and R.C. Cohen, Network design for quantifying urban CO2 emissions: assessing trade-offs between precision and network density, Atmos. Chem. Phys. 16, 13465-13475, 2016.

A.A. Shusterman, V.E. Teige, A.J. Turner, C. Newman, J. Kim, and R.C. Cohen, The BErkeley Atmospheric CO2 Observation Network: initial evaluation, Atmos. Chem. Phys. 16, 13449-13463, 2016.