Evaluation of atmospheric nitrogen deposition model performance in the context of U.S. critical load assessments. (February 2017)

• Record Type: Journal Article
• Authors: Williams, Jason J.; Chung, Serena H.; Johansen, Anne M.; Lamb, Brian K.; Vaughan, Joseph K.; Beutel, Marc
• Is Part Of: Atmospheric environment. Volume 150(2017)
• Page Start: 244
• Page End: 255
• Publication Date: 2017-02
• Subjects: Atmospheric deposition -- Nitrogen -- Model performance -- Critical load -- Exceedance; Air -- Pollution -- Periodicals; Air -- Pollution -- Meteorological aspects -- Periodicals; 551.51
• DOI: 10.1016/j.atmosenv.2016.11.051
• Languages: English
• ISSNs: 1352-2310

Abstract

Abstract: Air quality models are widely used to estimate pollutant deposition rates and thereby calculate critical loads and critical load exceedances (model deposition > critical load). However, model operational performance is not always quantified specifically to inform these applications. We developed a performance assessment approach designed to inform critical load and exceedance calculations, and applied it to the Pacific Northwest region of the U.S. We quantified wet inorganic N deposition performance of several widely-used air quality models, including five different Community Multiscale Air Quality Model (CMAQ) simulations, the Tdep model, and 'PRISM x NTN' model. Modeled wet inorganic N deposition estimates were compared to wet inorganic N deposition measurements at 16 National Trends Network (NTN) monitoring sites, and to annual bulk inorganic N deposition measurements at Mount Rainier National Park. Model bias (model – observed) and error (|model – observed|) were expressed as a percentage of regional critical load values for diatoms and lichens. This novel approach demonstrated that wet inorganic N deposition bias in the Pacific Northwest approached or exceeded 100% of regional diatom and lichen critical load values at several individual monitoring sites, and approached or exceeded 50% of critical loads when averaged regionally. Even models that adjusted deposition estimates based on deposition measurements to reduce bias or that spatially-interpolated measurement data, had bias that approached or exceeded critical loads at some locations. While wet inorganic N deposition model bias is only one source of uncertainty that can affect critical load and exceedance calculations, results demonstrate expressing bias as a percentage of critical loads at a spatial scale consistent with calculations may be a useful exercise for those performing calculations. It may help decide if model performance is adequate for a particular calculation, help assess confidence in calculation results, and highlight cases where a non-deterministic approach may be needed. Highlights: A novel performance assessment approach was developed to inform critical load applications. Model estimates were compared to wet and bulk inorganic N deposition measurements. Model bias and error were expressed as a percentage of regional critical load values. Bias was large relative to or even exceeded critical loads in some cases. This approach may help assess confidence in critical load and exceedance calculations.