Observing System Simulation Experiment for designing a monitoring network in urban areas using the GRAMM/GRAL model

Urban areas and cities play a crucial role in mitigating climate change. First, a large share of greenhouse gases (about 60%) is emitted in urban areas. Second, city networks have formed to implement climate change mitigation measures at the local level. Therefore, cities have great potential to significantly reduce greenhouse gas emissions.

Realizing this potential requires solid knowledge of local greenhouse gas sources, which can be obtained through robust measurements of greenhouse gases in an urban network. These measurements can then serve as a starting point for quantifying and thus verifying local emissions. In order to optimize the investment in a measurement network and maximize the knowledge gained from these measurements, several parameters need to be considered, such as the number and location of nodes, the uncertainty of the measurements, and the co-measured species.

These parameters can be evaluated and optimized in an Observing System Simulation Experiment (OSSE). In our study we perform a high-resolution OSSE using the atmospheric transport model GRAMM/GRAL. We first feed a high-resolution anthropogenic emission inventory into the model and simulate CO₂ concentration in the urban atmosphere on 10 m resolution in a 12 km x 12 km domain. Next, we approximate CO₂ fluxes on neighborhood scale using an inverse framework. We test different configurations of possible measurement networks to assess under which circumstances and how well we can estimate CO₂ fluxes. We find that the accuracy of the estimated fluxes increases with node number and precision, reaching a mean error reduction of about 50 % for 16 nodes and a precision of 1.0 ppm in the best configurations. Sources emitted on ground level can be successfully estimated on hourly resolution, but the measurement stations are not sensitive enough to detect sources emitted from tall emission stacks in the vicinity of the measurement nodes. We further discuss the advantages of using CO as an additional tracer in the inversion with respect to measurement precision and sectoral source disaggregation. The framework developed allows for the planning of an optimal measurement network and can thereafter be used to derive fluxes and associated uncertainties in urban areas. This allows for verification of emissions and targeted monitoring of mitigation measures at the local scale.