Simulation of mesoscale and diurnal variability of atmospheric CO2 concentrations in a region including complex terrain and densely populated areas

Abstract:

For an accurate simulation of the atmospheric state in the planetary boundary layer (PBL) biosphere – atmosphere interactions are of particular importance. Measurements indicate distinct regional scale spatio-temporal patterns in the atmospheric CO₂ distribution. The aim of our study is to understand which processes and environmental conditions (e.g. atmospheric transport, land use, orography) generate these patterns and how the variable CO₂ concentrations influence the stomatal control of transpiration and photosynthesis. For that, we use the regional scale terrestrial model system TerrSysMP-CO₂ that couples the atmospheric model COSMO (developed from the German Meteorological Service) to the Community Land Model (CLM). TerrSysMP-CO₂ includes a two-way coupling of CO₂, i.e. the actual CO₂ mixing ratio is used to calculate the biogenic CO₂ fluxes (photosynthesis, autotrophic/heterotrophic respiration) with CLM and, in turn, these fluxes prognostically change the atmospheric CO₂ content. High-resolution anthropogenic emissions complete the CO₂ budget in TerrSysMP-CO₂.

We will present final results simulated with TerrSysMP-CO₂ over a model domain including a low mountain range and the densely populated Rhine valley in the western part of Germany. Our results show a distinct diurnal cycle of CO₂ in the PBL with the highest values occurring in the early morning caused by near surface CO₂ accumulation due to soil respiration. With the onset of photosynthesis a strong decrease of atmospheric CO₂ is simulated as well as the turbulent vertical transport within the PBL. Downstream of densely populated regions significant higher CO₂ concentrations can be seen. Moreover, a strong horizontal heterogeneity arises between narrow valleys and mountain ridges caused by mountain-valley circulations. Compared with model simulations without CO₂ dynamics we see changes in the simulated temperature and moisture distribution in the PBL. These can be attributed to the response of simulated leaf stomatal resistances on the variable CO₂ which leads to modified transpiration fluxes. A comprehensive comparison with eddy-covariance flux measurements as well as with vertical CO₂ profiles measured at a 120 m tower show a good consistency of our model simulations.