A multi-model approach to monitor emissions of CO<sub>2</sub> and CO from an urban-industrial complex

Abstract

<p><strong>Abstract.</strong> Monitoring urban-industrial emissions is often challenging, because observations are scarce and regional atmospheric transport models are too coarse to represent the high spatiotemporal variability in the resulting concentrations. In this paper we apply a new combination of a Eulerian model (WRF with chemistry) and a Gaussian plume model (OPS). The modelled mixing ratios are compared to observed CO<sub>2</sub> and CO mole fractions at four sites along a transect from an urban-industrial complex (Rotterdam, Netherlands) towards rural conditions for October–December 2014. Urban plumes are well-mixed at our semi-urban location, making this location suited for an integrated emission estimate over the whole study area. The signals at our urban measurement site (with average enhancements of 11 ppm CO<sub>2</sub> and 40 ppb CO over the baseline) are highly variable due to the presence of distinct source areas dominated by road traffic/residential heating emissions or industrial activities. This causes different emission signatures that are translated into a large variability in observed ΔCO : ΔCO<sub>2</sub> ratios, which can be used to identify dominant source types. We find that WRF-Chem is able to represent synoptic variability in CO<sub>2</sub> and CO (e.g. the median CO<sub>2</sub> mixing ratio is 9.7 ppm (observed) against 8.7 ppm (modelled)) , but it fails to reproduce the hourly variability of daytime urban plumes at the urban site (R<sup>2</sup> up to 0.05). For the urban site, a plume model should be added to the model framework to adequately represent plume transport especially from stack emissions. The explained variance in hourly, daytime CO<sub>2</sub> enhancements from point source emissions increases from 30 % with WRF-Chem to 52 % with WRF-Chem in combination with the most detailed OPS simulation. The simulated variability in ΔCO : ΔCO<sub>2</sub> ratios decreases drastically from 1.5 to 0.6 ppb ppm<sup>−1</sup> which agrees better with the observed standard deviation of 0.4 ppb ppm<sup>−1</sup>. This is partly due to improved wind fields (increase in R<sup>2</sup> of 0.10), but also due to improved point source representation (increase in R<sup>2</sup> of 0.05) and dilution (increase in R<sup>2</sup> of 0.07). Based on our analysis we conclude that a plume model with detailed and accurate dispersion parameters is inevitable for top-down monitoring of greenhouse gas emissions in urban environments with large point source contributions within a ~ 10 km radius from the observation sites.</p>

Applying Machine-Learning Methods Based on Causality Analysis to Determine Air Quality in China

“…Toja et al [1] demonstrated that the average height of the buildings showed a clear influence on the vertical profile of carbonic oxide concentration. Super et al [2] proposed a multi-model approach based on the combination of Eulerian model and Gaussian plume model to monitor emissions of CO from the urban-industrial complex. For statistical characteristics-based methods, time series analysis and significance test are often used to evaluate air quality.…”