I have calculated two ecotoxicity CFs of ~2.6 and ~280 CTUe, and want to konw if you would consider this difference significant relative to model uncertainty?
Uncertainty associated with ecotoxicity characterization factors (CFs) in USEtox are typically 1-2 orders of magnitude - this will depend further on the emission scenario (e.g. emissions directly into freshwater will have a lower uncertainty as no multimedia fate needs to be considered, and hence there is much lower model uncertainty involved) and whether your results are already aggregated or not. Further details of the uncertainty associated with each underlying USEtox factor can be found in Rosenbaum et al. (2008).
Given these ranges, your CFs are most likely significantly different. To evaluate this further, you could look at your different product systems that you evaluated, exclude impacts associated with overlapping components between the different product systems and only look at the difference in ecotoxicty impacts for the remaining (different) components. This way, you focus on the distinct product systems components and related impacts and further reduce related uncertainty in resulting (aggregated) CFs.
Regarding the GSD estimates listed in Table 2 of the Rosenbaum et al (2008) piece: Are there similar estimtes for emissions to urban air, marine water, and natural soil?
Differences in uncertainty ranges for different emission compartments are mainly a function of compartment characteristics. This means that uncertainty ranges (as expressed via GSD² in Table 2 of Rosenbaum et al. 2008) for emissions to urban and rural air, for emissions to freshwater and coastal marine water, and for emissions to agricultural and natural soi, respectively, are similar.
This is an important question!
Uncertainty associated with ecotoxicity characterization factors (CFs) in USEtox are typically 1-2 orders of magnitude - this will depend further on the emission scenario (e.g. emissions directly into freshwater will have a lower uncertainty as no multimedia fate needs to be considered, and hence there is much lower model uncertainty involved) and whether your results are already aggregated or not. Further details of the uncertainty associated with each underlying USEtox factor can be found in Rosenbaum et al. (2008).
Given these ranges, your CFs are most likely significantly different. To evaluate this further, you could look at your different product systems that you evaluated, exclude impacts associated with overlapping components between the different product systems and only look at the difference in ecotoxicty impacts for the remaining (different) components. This way, you focus on the distinct product systems components and related impacts and further reduce related uncertainty in resulting (aggregated) CFs.
Thanks for your response!
Regarding the GSD estimates listed in Table 2 of the Rosenbaum et al (2008) piece: Are there similar estimtes for emissions to urban air, marine water, and natural soil?
Differences in uncertainty ranges for different emission compartments are mainly a function of compartment characteristics. This means that uncertainty ranges (as expressed via GSD² in Table 2 of Rosenbaum et al. 2008) for emissions to urban and rural air, for emissions to freshwater and coastal marine water, and for emissions to agricultural and natural soi, respectively, are similar.