**Appendix A**

#### *Appendix A.1. Spatial Distribution of Exposure Concentration*

Figure A1 shows the spatial distribution of exposure concentrations estimated by Tier-model 2, Tier-model 3 and Tier-model 4.

**Figure A1.** *Cont*.

**Figure A1.** Map of GLA showing the spatial distribution of the annual mean exposure concentration (μg/m3): (**a**) Tier 2, (**b**) Tier 3 and (**c**) Tier 4.

In Tiers 2 and 3, although the exposures range similarly, we can see different distribution patterns across GLA. The inclusion of the I/O ratios of the different dwelling types increased the spatial resolution of the exposure concentration, while in Tier 2 (Figure A1a) the spatial distribution was similar to our baseline Tier-model 1 (Figure 3a) due to the average infiltration factor that was used. Figure A1c is presenting higher exposure concentrations than Figure A1a,b due to the incorporation of the highly polluted transportation MEs such as LU. In all three maps, central GLA showed the highest exposure concentrations because of the high ambient PM2.5 concentrations in that area.

#### *Appendix A.2. Percentage of Exposure Concentration Reduction across GLA*

The percentage of reduction in exposure concentration after using Tier 3 and Tier 4 are shown in Figure A2a,b.

After including the I/O ratios of the different dwelling types (Tier 3), Central and South-eastern GLA had the highest reduction (between 45.5% and 47.6%), while in Figure A2b, Central and parts of the Southern GLA showed the highest percentage of mitigation (up to approximately 39.4%). As shown in Figure 1, Central and South-eastern GLA had the lowest I/O ratios. Because in Tier 3 transportation and outdoor MEs were not considered and people were assumed to spend 100% of their time indoors, the percentage of the reduction was mainly driven by the infiltration factors. However, after increasing the complexity and including outdoor and transportation MEs, the percentage of reduction was also strongly influenced by the ambient PM2.5 concentration in each area, in addition to the infiltration factors.

**Figure A2.** Percentage (%) of exposure concentration reduction between Tier 1 and: (**a**) Tier 3 and (**b**) Tier 4.

#### *Appendix A.3. Spatial Distribution of the Predicted Avoided Mortality*

The annual mean predicted avoided mortality due to the utilization of Tiers 2, 3 and 4 is illustrated in Figure A3a–c.

Figure A3a–c show similar spatial distribution of the predicted avoided mortality. As anticipated, the number of predicted mortalities after using Tiers 2 and 3 was higher, because the concentrationresponse function used by BenMap-CE to calculate the mortality was affected by the reduction (or increase) of the exposure concentration. The absence of the time-activity data in Tier 2 and Tier 3 led to an underestimation of the total exposure concentration and as a result the exposure difference between those models and our baseline model (Tier 1) was higher.

**Figure A3.** Map of the Greater London Area showing the spatial distribution of the mean avoided mortality (in death cases): (**a**) Tier 1–2, (**b**) Tier 1–3, (**c**) Tier 1–4.
