TransClim (v1.0): a chemistry–climate response model for assessing the effect of mitigation strategies for road trafﬁc on ozone

. Road trafﬁc emits not only carbon dioxide (CO 2 ) and particulate matter, but also other pollutants such as nitrogen oxides (NO x ), volatile organic compounds (VOCs) and carbon monoxide (CO). These chemical species inﬂuence the atmospheric chemistry and produce ozone (O 3 ) in the troposphere. Ozone acts as a greenhouse gas and thus contributes to anthropogenic global warming. Technological trends and political decisions can help to reduce the O 3 effect of road trafﬁc emissions on climate. In order to assess the O 3 response of such mitigation options on climate, we developed a chemistry–climate response model called TransClim (Modelling the effect of surface Trans portation on Clim ate). The current version considers road trafﬁc emissions of NO x , VOC and CO and determines the O 3 change and its corresponding stratosphere-adjusted radiative forcing. Using a tagging method, TransClim is further able to quantify the contribution of road trafﬁc emissions to the O 3 concentration. Thus, TransClim determines the contribution to O 3 as well as the change in total tropospheric O 3 of a road trafﬁc emission scenario. Both quantities are essential when assessing mitigation strategies. The response model is based on lookup tables which are generated by a set of emission variation simulations performed with the global low


SI-text: OA source apportionment
OA source apportionment was performed on high resolution mass spectra for m/z 12-120 using PMF combined with the multilinear engine (ME-2) in Igor Pro (Paatero, 1999;Canonaco et al., 2013).First, a range of solutions with two to ten factors from unconstrained runs were examined.
The OA factors are mixed with each other seriously in less factor number solutions, and increasing the number of factors can improve the separation.As shown in Fig. S4, four POA factors including HOA, COA, CCOA and BBOA were resolved in the 6-factor solution, while high signal of CO 2 + was mixed in CCOA and COA factors.Meanwhile, two SOA factors with different oxidation states were also resolved, referred to as LO-OOA with O/C ratio of 0.42 and MO-OOA with O/C ratio of 1.02, respectively.However, much hydrocarbon-like ions in higher m/z were observed in LO-OOA, also suggesting the mixture between POA factors and SOA factors.In seven factors solution, the mixture of CO 2 + in COA was reduced, while the mixture in CCOA and LO-OOA were still observed, and MO-OOA was separated into MO-OOA1 and MO-OOA2.When further increasing the factor into eight, clear POA sources including CCOA, BBOA, COA, and HOA were resolved.and the mixture in LO-OOA was also reduced.However, the MO-OOA is still separated into MO-OOA1 and MO-OOA2, and there was no new meaningful factor resolved.Therefore, six sources including HOA, COA, CCOA, BBOA, LO-OOA and MO-OOA should be resolved.Note that reasonable POA sources could not be resolved in six PMF solution, while SOA was separated in the seven and eight factor solutions.
To optimize the source apportionment, we further conducted 6-factor solution using ME-2 approach in order to obtain a more environmentally meaningful result (Canonaco et al., 2013).
The clear profiles of HOA, COA, CCOA and BBOA resolved in 8-factor PMF solution were used to constrain these four sources, and a-value of 0.1 was applied for a reasonable variation range.Finally, six OA sources including HOA, COA, BBOA, CCOA, LO-OOA, and MO-OOA were identified (Fig. S5).
As for POAs, both the mass spectrum of HOA and COA are characterized by prominent hydrocarbon ion series of C n H 2n−1 and C n H 2n+1 , while the COA contains higher signal at C n H 2n−1 than C n H 2n+1 , an especially much higher signal ratio of C 4 H 7 + /C 4 H 9 + which is the typical characteristic of COA profile as reported in previous studies at various urban sites (He et al., 2011;Ng et al., 2011).The HOA was emitted mainly from local traffic with mass peaks in traffic hours (Huang et al., 2010;Sun et al., 2011;Ge et al., 2012), and COA was dominantly associated with the cooking emissions (He et al., 2010;Xu et al., 2016), exhibiting mass peaks at around breakfast, lunch and dinner times (Fig. S6).Meanwhile, the time series of COA correlated well with that of C 6 H 10 O + (r=0.85), which is a tracer fragment of cooking emissions (Sun et al., 2011;Hu et al., 2016) (Fig. S6).The mass spectrum of CCOA is dominated by unsaturated hydrocarbons, particularly polycyclic aromatic hydrocarbons (PAH)-related ion peaks (e.g., m/z 77, 91 and 115) (Dall'Osto et al., 2013;Hu et al., 2013).In comparison, BBOA was mainly emitted from the combustion of biomass fuel such as wood or straw (Wang et al., 2009;Zhang et al., 2015), with the tracer signal at m/z 60 (dominantly C 2 H 4 O 2 + ), which is mainly enhanced by biomass-burning tracer levoglucosan and related species (mannosan, galactosan, etc.) (Cubison et al., 2011).Consistently, a good correlation between the time series of BBOA and the C 2 H 4 O 2 + fragment was also observed (r=0.95) (Fig. S6).
As for SOAs, the mass spectrum of LO-OOA shows a higher peak at m/z 43 (mainly and a lower O/C ratio (0.57) than that of MO-OOA (0.82).The tight correlation between the time series of C 2 H 3 O + and LO-OOA (r = 0.83) also suggests its less oxidized property (Fig. S6).
In addition, obvious m/z 60 (C 2 H 4 O 2 + ) and m/z 73 (C 3 H 5 O 2 + ) peaks were found in the mass spectrum of LO-OOA and the time variation of LO-OOA correlated well with that of BBOA (R 2 =0.59) (Fig. S7), suggesting the possible influence of BBOA source on the formation of LO-OOA.Therefore, LO-OOA was further referred to as OOA-BB and discussed in detail in section 3.2.In comparison, MO-OOA has a mass spectrum with a much higher peak at m/z 44 (mainly CO 2 + ) than m/z 43, and a largely different time series with LO-OOA, implying different formation processes.Meanwhile, the time series of MO-OOA was correlated well with the typical fragment ions of aqueous-phase processing products including CH 2 O 2 + (R 2 = 0.91), CH 3 SO + (R 2 = 0.89), and CH 3 SO 2 + (R 2 = 0.75) (Fig. S8) (Tan et al., 2009;Chhabra et al., 2010;Ge et al., 2012;Sun et al., 2016), suggesting the dominant role of aqueous-phase processes in its formation.Therefore, MO-OOA was further referred to as aq-OOA and discussed in detail in section 3.3.

Fig. S1
Fig. S1 Comparisons of frequencies of temperature(a), Ox concentration(b), WS(c), WD(d), RH(e) and ALWC(f) between reference days and SIA-enhanced periods (SIA_P1 and SIA_P2), and variation of NO 3 -/SO 4 2-ratio as a function of PM 2.5 mass, colored by RH during winter campaign (g).

Fig. S3
Fig. S3 Correlation between RH and SOR (a) or NOR (b) during the winter campaign in Xi'an.

Fig. S4
Fig. S4 Profiles (HR) of OA sources for six, seven and eight factors PMF solutions during the winter campaign in Xi'an.

Fig. S5
Fig. S5 Profiles (HR) of OA sources for six factors ME-2 solution during the winter campaign in Xi'an.

Fig. S6
Fig. S6 Diurnal cycles of OA sources (a), and the time series of each source and their tracers during the winter campaign (b).

Fig. S9
Fig. S9 Mass concentrations of OA sources, OA oxidation ratios (H/C, O/C and OM/OC) during different periods including reference days, SIA_P1 and SIA_P2.

Fig. S11
Fig. S11 Correlation between the concentration of aq-OOA and sulfate or nitrate during summer (a), and winter (b) in Xi'an.