Has your modelling of air emission impacts taken account of the local inversions in the Severn Valley and the frequency of local katabatic winds in stable conditions in winter?
The met data used in the model was for Shawbury, and Numerical Weather Prediction (NWP) data based on the location of the stack coordinates (see Section 2.12. of Technical Appendix 6-1). Consequently, to answer this question, we will focus on the NWP data only as this met data is based on the conditions at the location of the ERF, and specifically 2019 as this gave the highest process contributions for the majority of pollutants.
A Capping inversion can occur at the top of the boundary layer. The dispersion model models these effects by including extra terms in the plume concentration algorithms to allow for the reflection of material below the boundary layer top. An inversion will always be present in convective and neutral conditions, but only present in stable conditions if there is a temperature jump at the boundary layer. From the 2019 NWP data, the temperature jump occurs 2701 times (approximately 31% of the time) within the 2019 data set, thus capping inversions will have been modelled for convective, neutral and stable conditions.
Within the boundary layer, temperature inversions are expected to occur in during stable conditions. Stable conditions within the boundary layer are indicted when the model calculated boundary layer depth divided by the Monon-Obukhov length is greater than 1. Using the 2019 NWP data as an example, this occurs 3423 times (approximately 39% of the time), and the boundary layer height varies from 50m to 1036m. Note within the 2018 met data expected temperature inversions were more frequent (circa 56% of the time, however, 2019 gave higher predicted ground level pollutant concentrations).
To demonstrate that the plume is able to penetrate the inversion, the dispersion modelling software is also able to plot the fraction of the plume that penetrates the inversion layer. For example, on the 3rd October 2019 at 7am, the model predicts stable conditions, and a boundary layer height of 50m. The fraction of the plume penetrating the inversion layer can then be plotted as follows.
This demonstrates that the whole plume penetrates the inversion. However, in this case, the height of the inversion is lower than the height of the stack. Consequently, a boundary height of 90m was also considered – i.e. a height 20m above the height of the stack.
On the 5th January at 7am the model predicts stable conditions, and a boundary layer height of 90m. Please note that as the height of the stack is 70m, the plume height must also be considered. As shown in the plume height graph, the plume will not actually reach the temperature inversion until it is around 100m from the stack.