Examining life cycle impacts of metropolitan food system scenarios: comparing top-down and bottom-up LCA modeling approaches

Food systems have evolved to supply the growing demands of urban populations, often without providing reliable supplies of nutritious, environmentally sustainable foods. In this study we assessed life cycle impacts (global warming potential (GWP), energy, water, and land use) for the Des Moines Metropolitan Statistical Area (DM-MSA) U.S. Midwest) food system, comparing two LCA modeling approaches. Metropolitan food system scenarios for the DM-MSA include a baseline in which about 95% of table food is imported, compared to an alternative where 50% of food is grown within the MSA (local scenarios). Food system environmental impacts were quantified using the Environmentally Extended Input Output (USEEIO) model (2020 to 2050) and a CleanMetrics model (2020). Based on outputs from both models the local food system scenarios produced consistently lower per capita environmental impacts. The local USEEIO scenario predicted 25% less water use and 15% less GWP produced and energy used compared to baseline scenarios. However, overall the USEEIO model reported more than twice the GWP produced and energy and water use per person in 2020 compared to the CleanMetrics model. Land use was calculated outside the two model frameworks and used data for regionally-specific yields in USEEIO which significantly reduced land area needed to achieve the target production. Increased local food production could substantially lower the environmental footprint of urban food systems in the U.S. Midwest. The re-structured version of the USEEIO model also provides opportunities for integration with other social, water, energy and climate models, enabling a more holistic perspective on food system production, consumption and impacts at critical decision-making scales.