Scholarlink

Purpose <p>Prospective LCAs sometimes neglect how potential future impacts could be affected by future shifts in environmental conditions. Despite expected future changes in water availability, prospective characterization factors (CFs) for the updated freshwater deprivation impact assessment method AWARE, AWARE2.0, do not exist yet. Additionally, interpreting AWARE requires an understanding of its uncertainty. This paper improves the understanding of AWARE2.0’s uncertainty by combining sets of alternative input values for the calculation of regular and of prospective AWARE2.0 CFs.</p> Methods <p>Data from the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) framework is used to estimate AWARE2.0’s uncertainty and to calculate prospective AWARE2.0 CFs. For the uncertainty assessment, stochastic AWARE2.0 is generated by Monte Carlo sampling from parametric distributions representing AWARE2.0’s input data. Two prospective versions of AWARE2.0 are calculated: A “climate-prospective” model ensemble consisting of five climate models produces prospective CFs under climate change for three scenarios. “Fully prospective” AWARE2.0 additionally integrates projected water demand changes for three shared socioeconomic pathways in a simplified manner.</p> Results and discussion <p>Stochastic AWARE2.0 allows a discussion of major challenges in modeling the uncertainty of AWARE2.0 and suggests that its water availability input can deviate from observations by more than 20%. “Climate prospective AWARE2.0” shows that for many areas the future change in CFs under climate change is primarily determined by the selected climate models instead of the climate scenarios. The ensemble average CFs in some cases still change by more than 50% up to 2049. With the future socioeconomic changes integrated in “fully prospective AWARE2.0”, CF increases are more frequent than when considering climate change only.</p> Conclusions <p>The input model choice for calculating AWARE CFs can considerably affect both regular and prospective CFs. With its prospective CF timeseries, this paper contributes to dynamic and prospective water deprivation LCIA while improving the understanding of input-data related uncertainty in AWARE2.0.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Harnessing model ensembles to assess uncertainty and provide prospective characterization factors for AWARE2.0

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • Georg Seitfudem,
  • Markus Berger,
  • Anne-Marie Boulay

摘要

Purpose

Prospective LCAs sometimes neglect how potential future impacts could be affected by future shifts in environmental conditions. Despite expected future changes in water availability, prospective characterization factors (CFs) for the updated freshwater deprivation impact assessment method AWARE, AWARE2.0, do not exist yet. Additionally, interpreting AWARE requires an understanding of its uncertainty. This paper improves the understanding of AWARE2.0’s uncertainty by combining sets of alternative input values for the calculation of regular and of prospective AWARE2.0 CFs.

Methods

Data from the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) framework is used to estimate AWARE2.0’s uncertainty and to calculate prospective AWARE2.0 CFs. For the uncertainty assessment, stochastic AWARE2.0 is generated by Monte Carlo sampling from parametric distributions representing AWARE2.0’s input data. Two prospective versions of AWARE2.0 are calculated: A “climate-prospective” model ensemble consisting of five climate models produces prospective CFs under climate change for three scenarios. “Fully prospective” AWARE2.0 additionally integrates projected water demand changes for three shared socioeconomic pathways in a simplified manner.

Results and discussion

Stochastic AWARE2.0 allows a discussion of major challenges in modeling the uncertainty of AWARE2.0 and suggests that its water availability input can deviate from observations by more than 20%. “Climate prospective AWARE2.0” shows that for many areas the future change in CFs under climate change is primarily determined by the selected climate models instead of the climate scenarios. The ensemble average CFs in some cases still change by more than 50% up to 2049. With the future socioeconomic changes integrated in “fully prospective AWARE2.0”, CF increases are more frequent than when considering climate change only.

Conclusions

The input model choice for calculating AWARE CFs can considerably affect both regular and prospective CFs. With its prospective CF timeseries, this paper contributes to dynamic and prospective water deprivation LCIA while improving the understanding of input-data related uncertainty in AWARE2.0.