We investigate the importance of including in low-resolution atmospheric models the plume rise associated with the strong buoyancy of hot gases from vegetation fires. This sub-grid transport mechanism is simulated by embedding a 1D cloud resolving model, with appropriate lower boundary conditions, in each column of a 3D host model. Remote-sensing fire products are used in combination with a land use dataset for selection of appropriate fire properties. The host model provides the environmental conditions, and the plume rise is simulated explicitly. The final height of the plume is then used in the source emission field of the host model to determine the effective injection height, and the material emitted during the flaming phase is released at this height. Model results are compared with 500 hPa AIRS carbon monoxide (CO) data for September 2002 and with CO aircraft profiles from the SMOCC campaign, showing the huge impact on model performance.