Implementation of plume rise and its impacts on emissions and air quality modelling

Abstract

This work analyses the impact of implementing hourly plume rise calculations over Spain in terms of: i) vertical emission allocations and ii) modelled air quality concentrations. Two air quality simulations (4 km × 4 km, 1 h) were performed for February and June 2009, using the CALIOPE-AQFS system (WRF-ARW/HERMESv2.0/CMAQ/BSC-DREAM8b) differing only by the vertical allocation of point source emissions: i) using fixed vertical profiles based on the stack height of each facility and ii) using an hourly bottom-up calculations of effective emission heights. When using plume rise calculations, emissions are generally allocated to lower altitudes than when using the fixed vertical profiles, showing significant differences depending on source sector and air pollutant (up to 75% between estimated average effective emission heights). In terms of air quality, it is shown that hourly plume rise calculations lead to improved simulation of industrial SO2 concentrations, thus increasing modelled concentrations (1.4 μg m−3 increase in February, 1.5 μg m−3 increase in June) and reducing the model biases for both months (31.1% in February, 73.7% in June). The increase of SO2 concentrations leads to an increase of SO4−2 surface levels that varies according to the season and location (4.3% in February and 0.4% in June, on average). On the other hand, the impact on NO2 and PM10 concentrations is less significant, leading to average changes of a few μg·m3 at most (0.4 μg m−3 for NO2 and 0.2 μg m−3 for PM10). In order to maximize the precision of plume rise calculations, the use of stack parameters based on real-world data is mandatory.