Table 3.

Uncertainty analysis input data for estimating the health burden induced by increased air emissions of PCBs, and primary and secondary ${\mathrm{PM}}_{2.5}$.

Input parameter	GSD2	Sensitivity	% contribution
Inhalation of PCBs
Annual average air PCB concentrations (${\mathrm{C}}_{\text{air},\text{PCB},\mathrm{s}}$)a	1.1	$<0.01$	$<0.01$
Interspecies conversion factorb	19	$<0.01$	$<0.01$
Cancer dose–response factor (${\text{DR}}_{{\text{ADD}}_{\mathrm{p},\text{air}\_\text{PCB}},\mathrm{e}}$)c	1.3	$<0.01$	$<0.01$
Cancer severity factor (${\text{SF}}_{\text{PCB}}$)d	1.01	$<0.01$	$<0.01$
Noncancer dose–response factor (${\text{DR}}_{{\text{ADD}}_{\mathrm{p},\text{air}\_\text{PCB}},\mathrm{e}}$)e	See Figure S2	$<0.01$	$<0.01$
Noncancer severity factor (${\text{SF}}_{\text{PCB}}$)f	13.0	$<0.01$	$<0.01$
Inhalation of ${\mathrm{PM}}_{2.5}$, regional and U.S. populations
Heavy equipment primary ${\mathrm{PM}}_{2.5}$ emission factorg	1.3	$<0.01$	$<0.01$
Heavy equipment ${\mathrm{NO}}_{\mathrm{x}}$ emission factorg	1.3	$<0.01$	$<0.01$
Barge traffic primary ${\mathrm{PM}}_{2.5}$ emission factorh	1.7	$<0.01$	$<0.01$
Barge traffic ${\mathrm{NO}}_{\mathrm{x}}$ emission factorh	1.4	$<0.01$	$<0.01$
Rail transport primary ${\mathrm{PM}}_{2.5}$ emission factori	1.7	0.12	0.13
Rail transport ${\mathrm{NO}}_{\mathrm{x}}$ emission factori	1.7	0.87	6.70
Site-specific ${\mathrm{PM}}_{2.5}$ intake fractions (${\text{iF}}_{\mathrm{i}}$)j	4.6	0.05	0.21
Railroad ${\mathrm{PM}}_{2.5}$ intake fractions (${\text{iF}}_{\mathrm{i}}$)j	4.6	0.92	67.91
Dose–response factor (${\text{DR}}_{{\mathrm{PM}}_{2.5}}$)k	2.2	1.00	21.20
Severity factor (${\text{SF}}_{{\mathrm{PM}}_{2.5}}$)k	1.4	1.00	3.86
Inhalation of ${\mathrm{PM}}_{2.5}$, project workers
Personal exposure concentration (${\mathrm{C}}_{{\mathrm{PM}}_{2.5},\mathrm{p}}$)l	3.2	$<0.01$	0.01

Note: Presented values include geometric standard deviations (GSDs), sensitivity coefficients, and percent contributions to total output uncertainty. For these input parameters, we adapted an approach from MacLeod et al. (2002), assuming independent lognormal probability distributions (e.g., see Slob 1994). Following their approach, variance in each model output (cumulative health burden, DALYs) was calculated as a weighted sum of variances contributed by each input parameter with sensitivities as weights. Sensitivities were based on a 10% change in each input parameter relative to the total induced health burden in DALYs.

^aBased on the average variability within dredging seasons in the site-specific ambient air PCB monitoring results used for this study (Anchor QEA and Environmental Standards, Inc. 2009; Ecology and Environment 2004, 2017).

^bAccounts for uncertainty in the extrapolation of rodent data to humans as calculated by Huijbregts et al. (2005).

^cAccounts for experimental uncertainty (sample size), based on the ratio of upper bound and central estimate cancer slope factors (U.S. EPA 1996).

^dBased on the greatest 95th uncertainty interval for the corresponding DALY and incidence data as calculated by the Institute for Health Metrics and Evaluation (IHME 2017a). This assumes that the relative fractions of incidence for the three cancer types in these exposed populations are similar to those for the greater U.S. population (age- and sex-adjusted). Assuming these fractions are unknown would result in a maximum ${\text{GSD}}^2$ of 1.3 for this parameter. This would have a negligible (1%) effect on the total uncertainty in cancer health risk, since this uncertainty is driven by uncertainty in the interspecies conversion factor.

^eTotal uncertainty is displayed on Figure S2. Separate uncertainty distribution was applied in allometric scaling by body weight, accounting for chemical-specific interspecies differences. Interindividual variability was addressed by assuming a lognormal distribution for human variation, with an additional uncertainty distribution for the GSD of human variation. No subchronic uncertainty factor was applied, because the duration of the study by Tryphonas et al. (1991) was 55 months.

^fBased on Huijbregts et al. (2005) with considerably greater uncertainty than for cancer arising from use of an average severity factor, in DALY per case, across 49 diverse, noncommunicable diseases.

^gAccounts for uncertainty in the use of emission factors from Cao et al. (2016), based on variability across equipment types deemed to be most representative for this study. Furthermore, a separate uncertainty distribution was programmed in the Monte Carlo simulation to assign equal likelihood of Tiers 3 and 4 equipment.

^hReflects variability across dredging seasons based on the range of reported emission factors between 2013–2014 from the U.S. EPA SmartWay Carrier Performance database (U.S. EPA 2016). Furthermore, a separate uncertainty distribution was programmed in the Monte Carlo simulation to assign equal likelihood of each barge company in the database.

ⁱUncertainty distribution calculated from ranges of g per ton-mile emission factors summarized in a publication by the American Association of Railroads as provided by C. Crimmel (personal communication). Data were digitized using Plot Digitizer (version 2.6.8, Joe’s Java Programs).

^jUncertainty distribution based on the variability of intake fractions among models as calculated by Humbert et al. (2011).

^kUncertainty distribution as calculated by Gronlund et al. (2015).

^lBased on the variability of exposure levels reported by Lewné et al. (2007) for “construction machine operators” and “other outdoor workers exposed to diesel exhaust.” Worker impacts were considered as part of a separate sensitivity analysis.