Pandemic Restrictions in 2020 highlight the significance of non-road NO x sources in central London

. Fluxes of nitrogen oxides (NO x = NO + NO 2 ) and carbon dioxide (CO 2 ) were measured using eddy covariance at the BT Tower in central London during the coronavirus pandemic. Comparing fluxes to those measured in 2017 prior to the pandemic restrictions and the introduction of the Ultra-Low Emissions Zone (ULEZ) highlighted a 75 % reduction in NO x emissions between the two periods but only a 20 % reduction in CO 2 emissions and a 32 % reduction in traffic load. Use of a footprint model and the London Atmospheric Emissions Inventory (LAEI) identified transport and heat and power 5 generation to be the two dominant sources of NO x and CO 2 but with significantly different relative contributions for each species. Application of external constraints on NO x and CO 2 emissions allowed the reductions in the different sources to be untangled identifying that transport NO x emissions had reduced by > 75 % since 2017. This was attributed in part to the success of air quality policy in central London, but crucially due to the substantial reduction in congestion that resulted from pandemic reduced mobility. Spatial mapping of the fluxes suggests that central London was dominated by point source heat 10 and power generation emissions during the period of reduced mobility. This will have important implications on future air quality policy for NO 2 which until now, has been primarily focused on the emissions from diesel exhausts.

estimated to cost the UK's National Health Service (NHS) and social care £1.6bn between 2017 and 2025, rising to £5.6bn if diseases with less robust evidence for an association are included (Public Health England, 2018). This has become particularly relevant since the start of the coronavirus pandemic whereby long-term exposure to air pollution has been associated with the severity of COVID-19 cases (Imperial College London, 2021). 25 In 2008, countries in the EU were set legally binding limits for NO 2 concentrations in line with the World Health Organisation (WHO) recommendations. These are 40 µg m −3 for the annual mean with no more than 18 exceedances of the 200 µg m −3 hourly limit every year (European Parliament, 2008). This target was expected to be met by 2010. In 2021, the WHO reduced the recommended annual mean limit by 75 % to 10 µg m −3 .
London is a megacity in the UK with extensive NO 2 air quality issues. Almost all roadside locations exceeded the European 30 Limit Value for NO 2 every year between 2010 and 2016 (Font et al., 2019). Being in a highly developed position with significant resources, it has acted as a testing bed for policy intervention to try and curb emissions and achieve these air quality targets. These have been focused largely on traffic pollution and congestion charging and have the primary goal of reducing NO x concentrations via reduced road transport emissions, either through reduced traffic numbers or through reduced average emission per vehicle per unit distance. Most notable is the introduction of the world's first ultra-low emissions zone (ULEZ), 35 launched on 8th April 2019 with the zone spatially shown in Figure 1 a). This operates 24 hours a day, 364 days a year (excludes Christmas Day) and requires a daily payment if the vehicle driven inside the zone does not meet the most stringent emissions standards (currently Euro III for motorbikes, Euro IV for petrol cars and Euro VI for diesel cars and larger vehicles), in addition to the congestion charging payment within the same area. The ULEZ was expanded on 25th October 2021 up to the north and south circular roads in an 18-fold increase in size. In addition to policy, the coronavirus pandemic had significant 40 implications on NO x emissions in the UK through reduced mobility. During 2020 and 2021 the UK staged three lockdowns with "stay at home" orders. Full details on the timings and the severity of lockdown restrictions in London can be found in Figure A1.
Assessment of the impact of policy intervention and other external stimulus like the coronavirus pandemic on NO x emissions is crucial for the future design and implementation of air quality policy in the UK. Eddy-covariance is a technique used to 45 quantify the surface atmosphere exchange of an atmospheric pollutant. The calculated flux coupled with a footprint model provides information on surface emissions, allowing for changes to be studied and for direct comparison to the emissions inventories used in policy development. Whilst most frequently used for measuring carbon dioxide exchange with ecosystems from stationary towers (Baldocchi et al., 2001;Griffis et al., 2008;Butterbach-Bahl et al., 2013), the technique has been extended to the urban canopy for both greenhouse gases and air pollutants (Langford et al., 2010;Lee et al., 2014;Helfter 50 et al., 2016;Karl et al., 2017), as well as to airborne measurements for the assessment of fluxes at a much greater spatial extent (Vaughan et al., 2021;Metzger et al., 2013;Vaughan et al., 2017).
Here we present the first year of data from the long-term NO x flux measurement programme at the BT Tower (London, UK). As the only long-term measurements of NO x emissions from a megacity in the world this is a highly unique and potentially powerful data-set. The NO x emissions measurements are combined with additional CO 2 emissions measurements, a 55 2 https://doi.org/10.5194/egusphere-2022-956 Preprint. Telecommunications tower (BT Tower) located in central London, UK (51 • 31'17.4"N,0 • 8'20.04"W). The measurement height is 190 m above street level, with a mean building height of 8.8 ± 3.0 m in the 10 km radius surrounding the tower (Lee et al., 2014). The gas inlet and ultrasonic anemometer are attached to a mast that extends 3 m above the top of the tower. Air is pumped down a 45 m Teflon tube (3/8" OD) in a turbulent flow of 20-25 L min −1 to the gas instruments, which are situated in a small air conditioned room inside the tower on the 35th floor.

NO x measurements
Long-term measurements of NO and NO 2 fluxes began in September 2020 with data presented here up to September 2021.
Both chemical species were measured using a dual channel chemiluminescence analyser (Air Quality Design Inc., Boulder Colorado, USA; 5 Hz). The number of photons measured by the photomultiplier tube were converted into a part per trillion (ppt) mixing ratio using a five point calibration curve produced through dilutions of a 5 ppm NO in N 2 calibration standard 70 (BOC Ltd., UK; traceable to the scale of the UK National Physical Laboratory, NPL) into NO x free air. NO 2 was calculated by conversion of NO 2 into NO using a photolytic blue light converter (BLC). Here, both NO and NO 2 were measured, from which NO 2 can be quantified by subtracting the NO mixing ratio and applying a correction factor for the conversion efficiency of the BLC. The instrument was calibrated every 37 hours in addition to an hourly zero measurement to subtract the temperature dependent background signal of each channel.

CO 2 measurements
Long-term measurements of CO 2 have been ongoing at the BT Tower since 2011 as part of UKCEH's National Capability programme. Dry mass fractions were measured initially using a cavity ringdown spectrometer (Model 1301-f, Picarro Inc., Santa Clara, California, USA; 10 Hz) as described by Helfter et al. (2016). Unfortunately, instrumental failure means data is not available between February and June 2021, after which a closed path infrared gas analyser (Li-7000, LI-COR Environmental,

Meteorological measurements
Meteorological measurements were made at the BT Tower as described by Lane et al. (2013). Wind speed, wind direction and sonic temperature were measured using a ultrasonic anemometer (Gill R3-50, Gill Instruments, Lymington, UK; 20 Hz), along with pressure and relative humidity measurements using a weather station (WXT520, Vaisala Corp. Helsinki, Finland; 1 Hz).

Flux Calculations
The flux, F, is defined in this context as the vertical transport of a chemical species per unit area per unit time. Hourly fluxes were calculated using eddy covariance theory as described by Eq. (1), where F is equal to the covariance between the instantaneous change in vertical wind speed, w ′ , and the instantaneous change in species concentration, c ′ , averaged over the hour.
Eddy-covariance calculations were performed using the modular software packages in eddy4R adopting the same processing settings described in Drysdale et al. (2022) as adapted from Squires et al. (2020) This was to allow a direct comparison to be made to the previous measurements made in 2017. Data was filtered such that the friction velocity (u*) is > 0.2 to ensure sufficiently developed turbulence and using eddy4R's quality control flagging scheme based on both a combination of input data validation, stationarity and integrated turbulence characteristics Smith and Metzger, 2013).

Footprint Modelling
A parameterised version of the backwards Lagrangian stochastic particle dispersion model implemented in eddy4R was used to estimate the footprint for each hourly flux measurement at the BT Tower. The model is described by Kljun et al. (2004) and has been parameterised for a range of meteorological conditions and receptor heights to reduce the computational expense of running it. The original model aims to produce a cross-wind integrated footprint function as a function of its along-wind 100 distance, which has now been further extended into two dimensions using a Gaussian distribution driven by the standard deviation in the cross-wind component (Metzger et al., 2012;Kljun et al., 2015). Meteorology statistics from the eddy covariance calculations are used in combination with modelled boundary layer height from ERA5 (Copernicus Climate Change Service Climate Data Store (CDS), 2021), and a surface roughness length of 1.1 m to produce a weighted matrix of 100 m x 100 m grid cells. Each output weighted matrix was then scaled and aligned to an appropriate coordinate reference system to allow 105 each matrix to be plotted onto a map.

Traffic Data
Hourly traffic loads surrounding the BT Tower were calculated by summing the traffic load from each of the 24 Automatic Traffic Counters (ATCs) within the flux footprint, as shown in Figure 1. This gave an indication of the magnitude of traffic load for both measurement periods and allowed relative changes to be studied between the two years. In addition, daily vehicle 110 length breakdown was examined from which vehicles were separated into three length classes: < 5.2 m, indicating the number of passenger cars, 5.2 m-12 m, indicating the number of vans and rigid lorries and > 12 m, indicating the number of buses and arctic lorries. As the LAEI estimates that almost all lorry emissions in central London are due to the rigid class, the >12 m class is assumed to solely be made up from buses. Data was provided by the Operational Analysis Department, Transport for London (TFL) via a freedom of information request (Transport for London, 2021). Average median diurnal profiles with error bars for the data are shown to the right in blue for 2020/21 in comparison to those generated from the 2017 data in red.
Of the 8760 hours in the year, 7034 hours of NO x fluxes were calculated. Data loss was largely due to instrument or sample pump failure. Of these 7034 hours, a further 1341 were removed by the quality control flagging to leave 5693 hours or 65 % of high-quality fluxes to be analysed. This data is displayed in Figure 2 along with measured CO 2 flux, traffic load around the tower and the UK's restrictions stringency index as calculated by the Oxford COVID-19 Government Response Tracker (Hale et al., 2021). Traffic load around the tower was strongly anti-correlated with stringency index as expected. However, there was of the average diurnal profiles for each of NO x flux, CO 2 flux and traffic load around the tower in Figure 2 for this data, and the 2017 data described by Drysdale et al. (2022), highlighted an 75 % reduction in NO x flux since 2017. However, only a corresponding 20 % reduction in CO 2 flux and 32 % reduction in traffic load was observed.

Calculation of inventory estimated emissions
Estimated emissions from the London Atmospheric Emissions Inventory (LAEI) for the measurement footprint were calculated 140 to aid understanding of these observations. The hourly footprint weighted matrix output from eddy4R was used to select the relevant areas of the LAEI. The theoretical contribution to the flux was extracted from each footprint grid cell and scaled for hour of day, day of week and month of year for each emissions sector using a set of anthropogenic scaling factors described by Drysdale et al. (2022). An excellent agreement between the diurnal profiles and measurement footprint (shown in Figure   A2) for the 2017 and 2020/21 measurement periods was seen. This gave us confidence that any changes in emissions were not

Source apportionment of emissions reduction
The inventory breakdown for each species and the different percentage reductions in measured emissions since 2017 were used to disentangle changes in emissions of each sector. This was done simultaneously using a number of assumptions. Table 1. A summary of the data used in the formation of simultaneous Eqs. (2) and (3). representation of diesel vehicle emissions and/or congestion. Therefore, rather than using the inventory predicted 42:58 split for heat and power generation:transport the relative contributions were varied. Labelled as α : β, different scenarios between this assumption is considered reasonable. However, this is is likely to be untrue for transport. Policy implemented between 165 the two measurement periods specifically targeted NO x emissions and NO x emissions are disproportionately higher in higher traffic loads due to the ineffectiveness of exhaust treatment systems in that environment. Additionally, the modernisation of the vehicle fleet will have introduced more vehicles with lower NO x /CO 2 emission ratios. Therefore, the relative change in the emissions of NO x and CO 2 from traffic sources may not have been the same, and different values are given here as y and y'.
This information is all summarised in Table 1 with the two independent constraints displayed in Eq.'s (2) for CO 2 and (3) The different scenarios are visualised in Figure

Flux correlations with traffic load
Examining how the NO x flux correlated with traffic load for both measurement time periods gives further insight into the un- highly dependent on several variables including the fleet composition, type of exhaust treatment system and the actual level of congestion (Ko et al., 2019). It is thought that for individual roads, excess emissions from congestion can be anything up to 75 % greater than non-congested roads (Gately et al., 2017). Therefore, it is thought that reducing the peak traffic load below 25000 vehicles hr −1 has had a large impact on traffic NO x emissions, more than accounting for the remaining emissions 210 reduction.

Spatial Mapping
This change in emissions is clearly seen in the spatial mapping of the NO x fluxes in Figure  ). However, the requirement for CHP to be in urban areas risks an increase in air pollution. Indeed it has been shown that CHP can "substantially" impact air quality due to NO x , the highest criteria judged by Environment Protection UK and the Institute of Air Quality Management (Kings College London, 2018). Here, the heat and power generation source stands out and dominates over transport but is only seen due to the drastic reduction in transport NO x emissions during the period of pandemic reduced mobility. The greatly reduced correlation with traffic load for the easterly 2020/21 data in Figure 5 is further 235 evidence that the dominant source in this direction is heat and power generation.
Eddy covariance emissions measurements at the unique BT Tower site in central London provide an opportunity to study the evolution of air pollutant emissions in a megacity and the part that policy and other external stimuli play in improving air quality. Here, the direct emissions measurements have shown that reducing congestion could be an even more effective way of 240 reducing NO x emissions from road transport than the ULEZ. However, this is not the direction in which the UK is heading.
With much cheaper mileage, the continued uptake of electric vehicles is predicted to increase congestion. Reducing the number of vehicles on the road by improving infrastructure for other greener methods of travel such as cycling would not only achieve reduced congestion but give additional benefits to health further reducing costs of treatment at health services (Fishman et al., 2015). A more targeted approach to simultaneously reduce congestion as well as emissions per vehicle per unit distance is 245 therefore recommended to other cities looking to implement policies to tackle high traffic NO x emissions.
The observation that NO x emissions in central London during this continuing period of reduced mobility were thought to be dominated by heat and power generation is an important one. This is a transition which was expected to occur in the coming years but was brought forward in time by the pandemic, providing a glimpse into future air quality. As of 2020, there were 2659 CHP sites in the UK with additional widespread usage in Europe (Department for Business, Energy and Industrial 250 Strategy (BEIS), 2021). Due to their increased efficiency and the push towards NetZero economies, they are expected to increase in popularity. Despite this period of drastically reduced transport emissions, all air quality monitoring sites (urban background, urban traffic and curbside) in London far exceeded the new WHO NO 2 air quality target. To achieve these targets it is therefore clear that legislation is required to reduce NO x emissions from heat and power generation. The heat and power generation source has been somewhat neglected due to the prominence of issues with diesel vehicle emissions. But with the 255 planned use of hydrogen combustion in decarbonisation, which currently has major uncertainties due to a lack of experimental data, now is the critical time to start thinking about policy intervention for this sector (Lewis, 2021). This makes the lack of acknowledgement for gas combustion in boilers in the UK clean air strategy highly disappointing. This is the first indication from a megacity which shows heat and power emissions will need to be regulated to achieve the new air quality NO x targets.
As more and more of the world's population is expected to live in urban areas, it is essential that compliance with WHO targets 260 is achieved to minimise health and economic impacts. The conclusions derived from this work will therefore be of interest to other nations, especially with air quality improvements being increasingly sought in the developing world.