Water scarcity footprint of dairy milk production in New Zealand - A comparison of methods and spatio-temporal resolution

Water scarcity footprinting now has a consensual life cycle impact assessment indicator recommended by the UNEP/SETAC Life Cycle Initiative called AWaRe. It was used in this study to calculate the water scarcity footprint of NewZealand (NZ) milk produced in two contrasting regions; "non-irrigated moderate rainfall" (Waikato) and "irrigated low rainfall" (Canterbury). Two different spatial and temporal resolutions for the inventory flows and characterisation factors (CFs) were tested and compared: country and annual vs. regional and monthly resolution. An inventory of all the consumed water flows was carried out from cradle to farm-gate, i.e. from the production of dairy farm inputs to the milk and meat leaving the dairy farm, including all water uses on-farm such as irrigation water, cow drinking water and cleaning water. The results clearly showed the potential overestimation of a water scarcity footprint when using aggregated CFs. Impacts decreased by 74% (Waikato) and 33% (Canterbury) when regional and monthly CFs were used instead of country and annual CFs. The water scarcity footprint calculated at the regional and monthly resolution was 22 L-world eq/kg FPCM (Fat Protein Corrected Milk) for Waikato milk, and 1118 L-world eq/kg FPCM for Canterbury milk. The contribution of background processes dominated for milk from non-irrigated pasture, but was negligible for milk from irrigated pasture, where irrigation dominated the impacts. Results were also compared with the previously widely-used Pfister method (Pfister et al., 2009) and showed very similar ranking in terms of contribution analysis. An endpoint indicator was evaluated and showed damages to human health of 7.66 x 10(-5) DALY/kg FPCM for Waikato and 2.05 x 10(-3) DALY/kg FPCM for Canterbury, but the relevance of this indicator for food production needs reviewing. To conclude, this study highlighted the importance of using high-resolution CFs rather than aggregated CFs. (c) 2018 Elsevier B.V. All rights reserved.