Parameterization in Life Cycle Assessment inventory data: review of current use and the representation of uncertainty
- 871 Downloads
Parameterization refers to the practice of presenting Life Cycle Assessment (LCA) data using raw data and formulas instead of computed numbers in unit process datasets within databases. This paper reviews parameterization methods in the European Reference Life Cycle Data System (ELCD), ecoinvent v3, and the US Department of Agriculture's Digital Commons with the intent of providing a basis for continued methodological and coding advances.
Parameterized data are reviewed and categorized with respect to the type (raw data and formulas) and what is being represented (e.g., consumption and emission rates and factors, physical or thermodynamic properties, process efficiencies, etc.). Parameterization of engineering relationships and uncertainty distributions using Smirnov transforms (a.k.a. inverse transform sampling), and ensuring uncertain individual fractions (e.g., market shares) sum to the total value of interest are presented.
Seventeen categories of parameters (raw data and formulas) are identified. Thirteen ELCD unit process datasets use 975 parameters in 12 categories, with 124 as raw data points and 851 as formulas, and emission factors as the most common category of parameter. Five additional parameter categories are identified in the Digital Commons for the presentation and analysis of data with uncertainty information, through 146 parameters, of which 53 represent raw data and 93 are formulas with most being uncertainty parameters, percentages, and consumption parameters.
Parameterization is a powerful way to ensure transparency, usability, and transferability of LCI data. Its use is expected to increase in frequency, the categories of parameters used, and the types of computational methods employed.
KeywordsData Databases LCA Parameterization Uncertainty
This research was funded by the United States Department of Agriculture National Agricultural Library (agreement number 58-8201-0-149).
- Gaver DP, Kafadar K (1984) A retrievable recipe for inverse t. Am Stat 38:308–311Google Scholar
- Heijungs RR Frischknecht (2005) representing statistical distributions for uncertain parameters in LCA. Int J LCA 10(4):248–254Google Scholar
- Joint Research Center (2010) International Reference Life Cycle Data System (ILCD). Documentation of LCA data sets. Ispra, Italy: European Commission. Retrieved from lct.jrc.ec.europa.eu/Google Scholar
- Kim CS, Hallahan C, Lindamood W, Schaible G, Payne J (2004) A note on the reliability tests of estimates from ARMS data. Agr Resource Econ Rev 33(2):293–297Google Scholar
- Koehler KJ (1983) A simple approximation for the percentiles of the t distribution. Technometrics 25(1):103–105Google Scholar
- Shaw WT (2006) Sampling Student's T distribution—use of the inverse cumulative distribution function. J Comput Finance 9(4):37–73Google Scholar
- Sommer JE, Hoppe RA, Green RC, Korb PJ (1998) Structural and financial characteristics of US farms, 1995: 20th Annual Family Farm Report to Congress. Retrieved from http://www.ers.usda.gov/publications/aib746/
- Spiegel MR, Schiller JJ, Srinivasan RA, Alu R (2009) Schaum's outlines—probability and statistics, 3rd edn. McGraw-Hill, New York, NYGoogle Scholar
- Weidema BP, Bauer C, Hischier R, Mutel C, Nemecek T, Vadenbo CO, Wernet G (2011) Overview and methodology: data quality guideline for the ecoinvent database version 3 (final draft_revision 1). Retrieved from http://www.ecoinvent.org/fileadmin/documents/en/ecoinvent_v3_elements/01_DataQualityGuideline_FinalDraft_rev1.pdf
- Winitzki S (2003) Uniform approximations for transcendental functions. Proceedings of the ICCSA—2003, LNCS 2667 (p. 962). Presented at the International Conference on Computational Science and Its Applications—2003Google Scholar
- Winitzki S (2008) A handy approximation for the error function and its inverse. Retrieved from http://homepages.physik.uni-muenchen.de/~Winitzki/erf-approx.pdf