Characterization of waterborne nitrogen emissions for marine eutrophication modelling in life cycle impact assessment at the damage level and global scale

Abstract

Current life cycle impact assessment (LCIA) methods lack a consistent and globally applicable characterization model relating nitrogen (N, as dissolved inorganic nitrogen, DIN) enrichment of coastal waters to the marine eutrophication impacts at the endpoint level. This paper introduces a method to calculate spatially explicit characterization factors (CFs) at endpoint and damage to ecosystems levels, for waterborne nitrogen emissions, reflecting their hypoxia-related marine eutrophication impacts, modelled for 5772 river basins of the world. The proposed method combines environmental fate factors (FFs) integrating (i) DIN removal processes in soils and rivers, based on the NEWS 2-DIN model, and in coastal waters, based on water residence time; (ii) coastal ecosystem exposure (XF) to N enrichment, based on biological cycling processes; and (ii) effect factors (EFs) based on species sensitivity to hypoxia. Three emission routes are discriminated as N from soil and N in emissions to river and to coastal waters. Damage factors (DFs) are also estimated, based on endpoint metric conversion from potentially affected to potentially disappeared fractions of species (i.e. potentially affected fraction (PAF) to PDF m3 year kg N−1) and harmonization across coastal ecosystems based on spatially explicit density of demersal species, to further express CFs as species year per kilogram N−1. Endpoint CFs show 6 orders of magnitude (o.m.) spatial differentiation amongst the river basins for the soil emission route, 4 for the river and 2 for emissions to coastal waters. Damage CFs vary 7, 5 and 3 o.m. for the same routes. After aggregation at the level of continents, maximum CFs and DFs are consistently found in Europe, but the aggregation reduces spatial differentiation to 1 o.m. for each route in both factors. The FFNsoil and species density terms are responsible for most of the spatial differentiation of the damage model. Uncertainty is higher for the residence time term used in the FF model, due to scarcity and inconsistency of data sources, the assumptions of representativeness of DIN persistence and removal rates. Major contributions to the current state-of-the-art of marine eutrophication characterization modelling are (i) full pathway coverage, thus reaching damage level; (ii) significant increase in geographic coverage; (iii) mechanistic modelling of exposure and effect factors; and (iv) application of spatially explicit damage to ecosystems factors based on species densities. Application of the developed CFs in life cycle impact assessment is recommended at a river basin scale, provided that emission location is known.