MOESM6 of Assessing Lagrangian inverse modelling of urban anthropogenic CO2 fluxes using in situ aircraft and ground-based measurements in the Tokyo area Ignacio Pisso Prabir Patra Masayuki Takigawa Toshinobu Machida Hidekazu Matsueda Yousuke Sawa 10.6084/m9.figshare.8148338.v1 https://springernature.figshare.com/articles/journal_contribution/MOESM6_of_Assessing_Lagrangian_inverse_modelling_of_urban_anthropogenic_CO2_fluxes_using_in_situ_aircraft_and_ground-based_measurements_in_the_Tokyo_area/8148338 Additional file 7: Figure S6. Impact of changing the off diagonal terms on the prior error covariance matrix. S6a) Reducing the correlations to 10 km for all grid cells: the error reduction still follows roughly the prior fluxes distribution due to the diagonal terms proportional to the fluxes. S6b) The off diagonal terms are zero and the diagonal terms constant and set by the maximum gridcell value (1-sigma = max over the domain). The uncertainty is reduced mainly around the location of the observations and the error reduction follows the flow of the Lagrangian trajectories driven by the meteorological winds. 2019-05-17 05:00:00 Lagrangian error covariance matrix anthropogenic CO 2 fluxes term MOESM error reduction