Developing a harmonised framework to model nutrient emissions from land uses: a case study from Ireland

Nutrient loss from land use is a major cause of water-quality problems worldwide. High inputs of nitrogen (N) and phosphorus (P) to rivers, lakes, and coastal waters drive eutrophication, reduce aquatic biodiversity, and damage ecosystem services. In many regions, policy frameworks such as the European Union Water Framework Directive require reliable estimates of nutrient pressures to support effective water-quality management. However, measured load data are often limited or inconsistent, particularly at large spatial scales. As a result, export-coefficient (EC) approaches are widely used to estimate nutrient losses from land use. A continuing challenge is ensuring that the coefficients applied are relevant to the environmental context and selected in a consistent and transparent way, especially when values are transferred from international studies.

This study develops a harmonised framework for selecting and applying nutrient export coefficients based on international literature. A structured decision-tree approach is used to systematically assess whether published export coefficients are suitable for application under different climatic and environmental conditions. Each coefficient is screened against six practical criteria: compatibility with local land-use systems, similarity of soil types, relevance of climatic setting, comparability of dominant hydrological pathways, suitability of reporting format for load calculation, and study reliability—evaluated based on the quality of methods, length of monitoring, and peer-review status.

The framework is demonstrated using Ireland as a case study, and the analysis also compares how different land-cover datasets influence national nutrient export estimates. Three datasets are used to explore the effect of spatial representation: the Irish National Land Cover Map 2018 and the CORINE 2012 and 2018 Land Cover maps. National nutrient exports are calculated by multiplying harmonised export coefficients by mapped land-use areas and compared with a national benchmark study, the Source Load Apportionment Model (SLAM).

Across the different land-cover datasets, nutrient export coefficients derived from the framework show strong agreement with SLAM in estimating national nitrogen loads. This suggests that the decision-tree framework supports the selection of export coefficients for dominant agricultural systems, which are the main drivers of nitrogen loss. In contrast, larger differences between SLAM and framework-based estimates are observed for phosphorus, particularly in peatlands, wetlands, and forestry areas. This reflects both the higher sensitivity of phosphorus to coefficient choice and the influence of land-cover representation. Differences between land-cover datasets lead to significant changes in the mapped extent of organic soils and semi-natural land uses, resulting in notable variation in national phosphorus estimates. These findings show that even when identical export coefficients are applied, national nutrient totals can vary substantially depending on the structure and resolution of the underlying spatial data.

Overall, this study demonstrates that combining harmonised export coefficients with high-resolution land-cover data provides a robust and adaptable basis for national-scale nutrient modelling. The close agreement with SLAM for nitrogen supports the validity of the approach for dominant agricultural pressures, while the divergence observed for phosphorus highlights the need for improved representation of peatlands, wetlands, and forestry in national frameworks.