High resolution simulations of methane concentrations in the Netherlands using WRF-Chem

Abstract

Enhanced methane CH4 concentrations account for 23% of the anthropogenic radiative forcing. This study seeks to contribute to the insights in the methane emissions in the Netherlands. In this study the WRF-Chem model is used to conduct simulations of CH4 concentrations in the Netherlands. The aim of this study is to test the quality of high resolution simulations of methane concentration. An emission inventory with a resolution of 1 km x 1 km is used as input in the WRF-Chem model. High resolution simulations of the methane concentration in 2017 are conducted with a grid size of 3 km x 3 km, on the domain of the Netherlands. The emissions from the inventory were used as input for 13 individual tracers, one for every emission subgroup in the inventory. To test the quality of these simulations, the results are compared with in situ measurements of methane concentration from measurement sites at Lutjewad and Cabauw. The analysis showed an underestimation of the simulated CH4 concentration compared to the observed CH4 concentration during peak events. This underestimation is probably caused by using the average CH4 emissions over the volume of a 3 km x 3 km grid as input. For this reason local plume emissions are less dense in the simulations. The observed diurnal range in the CH4 concentrations is 3 to 4 times larger at Lutjewad and Cabauw than the diurnal range in the simulated CH4 concentration during summer. The opposite would be expected from the daytime nighttime ratio of the boundary layer height, which is larger in the simulations than in the observations. Furthermore, implementing diurnal variability in several smaller emission sources did not improve the diurnal range in the simulated methane concentration. The WRF-Chem simulations correctly reproduce the wind direction for which the excess CH4 concentration is highest in Lutjewad. The relative magnitude of the excess CH4 concentration for this wind direction is underestimated in the simulations. This can be caused by the poor representation of the diffusion of plume emission in the simulation. Another cause could be underestimated methane emissions around the measurement sites. The WRF-Chem simulations showed underestimations in the magnitude of the variability of methane concentration on several timescales. In order to increase the accuracy of the WRF-Chem simulations of CH4 concentrations, further research is need to gain insight into the causes of the differences in the variability of the simulated and observed CH4 concentrations.