ИСТИНА

The need to obtain accurate estimates of greenhouse gas (GHG) fluxes at regional and global scales has led to the development of innovative mathematical models that allow the retrieval of surface fluxes using remote sensing data. The most promising way to study high-resolution fluxes is the use of unmanned aerial vehicles (UAVs). A significant difficulty in obtaining surface fluxes using remote sensing methods can arise over areas with complex topography and mosaic vegetation patterns. There are also significant difficulties in measuring surface GHG fluxes in such areas using various in-situ methods. Therefore, there is still a need to improve existing technologies and to develop new methodologies for determining fluxes and describing their variability. In our study, we present a modified and improved inverse algorithm for the retrieval of greenhouse gas fluxes over a complex terrain. The main differences from the previous version of the algorithm are related to the computational grid and the numerical scheme for calculating the three-dimensional GHG distribution, as well as the measurement of concentrations and estimation of fluxes on horizontal planes above the canopy instead of on surfaces, bypassing the irregularities of the surface topography. As a result, the accuracy of flux estimation was significantly improved compared to the previous version of the inverse model algorithm. Two experimental sites were chosen for our modeling study, which differ in the heterogeneity of the vegetation and the topography of the underlying surface. In the first case, the swampy and forested areas of the carbon supersite "Mukhrino" (Khanty-Mansiysk Autonomous Okrug, Russia, 60°53'20" N, 68°42'10" E) were considered. The areas were characterized by a slight difference in surface elevation and a mosaic vegetation structure. In the second case, the areas of the Roshni-Chu mountain forest site, which is part of the "Way Carbon" supersite (Chechen Republic, Russia, 43°2'59" N, 45°25'32" E), were considered. This experimental site was characterized by homogeneous forest vegetation, but very complex topography with surface elevation differences up to 200 m. In both cases, the "measured" concentrations used as input data in the inverse problem were modeled using detailed data on plant canopy structure, LAI distributions, measured soil CO2 fluxes, photosynthesis rates, and necessary data to calculate the wind velocity and turbulence coefficient fields. This approach made it possible to evaluate the accuracy of the inverse algorithm with different values of the concentration measurement error and with a different structure of the underlying surface.