Table 3.
The types of downstream uncertainties in recent health effects of mitigation modeling studies.a
| Sector | Parametric uncertainties | Structural uncertainties |
|---|---|---|
| Household energy | ||
| Specification of mitigation scenarios | Average value of reduction in GHG emissions due to insulation improvements | Feasible transitions from household fossil fuel combustion to electricity |
| Estimating exposures | Values of the parameters of building physics model | Occupant behavior and increased consumption of resources given higher end-user efficiency |
| Estimating health impacts | Values of the pollutants’ relative risk coefficients | Pollutants to consider in the assessment |
| Urban land transport | ||
| Specification of mitigation scenarios | Percentage increase in the level of active travel (walking and cycling) | Nonlinear “safety in numbers” effect of increase in proportion of cyclists on rates of cyclist injuries; different future “active travel visions” |
| Estimating exposures | The values of the parameters of the emission–dispersion air pollution model | Reduction of emissions from transport in London are representative for other European cities; reduction in transport emissions results in proportional reduction in particulate matter |
| Estimating health impacts | The values of the physical activity–disease relative risk coefficients | Diseases affected by physical activity; linear versus nonlinear relationships between physical activity and health outcomes |
| Food and agriculture | ||
| Specification of mitigation scenarios | Percentage reduction in livestock production by 2030 | Contribution of different livestock to greenhouse emissions and different assumptions about feedstocks |
| Estimating exposures | Percentage reduction in intake of saturated fat | Full replacement of saturated fats with unsaturated fats |
| Estimating health impacts | Saturated fat-ischemic heart disease mortality relative risk coefficient | Exposure–health outcome pathways |
| aFriel et al. (2009); Maizlish et al. (2013); Wilkinson et al. (2009); Woodcock et al. (2013); these uncertainties are naturally not unique to co-benefits modeling. | ||