Inverse problem for retrieving greenhouse gas fluxes at the non-uniform underlying surface from measurements of their concentrations at several levels

Abstract

The study focuses on the formulation, analysis, and solution of the remote sensing inverse problem to retrieve surface carbon dioxide (CO2) fluxes from measurements of CO2 concentrations at different levels within the atmospheric boundary layer. A three-dimensional hydrodynamic model of turbulent greenhouse gas (GHG) transport was used as a forward model to link the surface GHG fluxes to the drone observations of GHG concentrations. The 3D model provides a GHG concentration distribution by solving the diffusion-advection equation using information on wind speed, its direction, and turbulent exchange coefficients. The surface GHG fluxes are considered as a boundary condition. The spatial distributions of wind speed and turbulence coefficient “for a moment in time” are computed from the relaxation problem for the averaged Navier-Stokes and continuity equations, using a 1.5 order closure scheme (E-ω model). The inverse problem is to retrieve a surface GHG flux by minimizing the difference between the measured and modelled concentrations at several levels. The algorithm was applied to estimate CO2 fluxes over a non-uniform forest canopy at the Roshny-Chu experimental site in the foothills of the Greater Caucasus (Chechen Republic). To test the forward numerical problem, data on surface topography, vegetation height and density, spatial distribution of photosynthetically active solar radiation, as well as data on plant photosynthesis and soil CO2 fluxes were used.