Thermal engineering calculations are a fundamental stage in the design in areas with the spread of multi-frozen soils (MFS). Today, this kind of calculations is well studied and implemented by various software complexes based on numerical methods. In the design of facilities of the fuel and energy complex, there is a mandatory requirement for this type of calculations, consisting in the calculation for the whole life of the facility. Despite the abundance of developed methods and approaches in solving these problems, there is some uncertainty in setting the boundary condition on the surface of soil heat exchange with the air medium. This uncertainty is caused both by the variety of factors that directly influence heat exchange on the soil surface and by the complexity of their accounting, due to the cyclicity of these factors during the period of operation of the structure. In research literature, one can find various versions of setting the boundary condition on the ground surface. A question arises on the correctness of applying a certain method, because setting the upper boundary condition by various methods leads to obtaining different results in further modeling. At the same time, insolvency of any of the methods of setting the upper boundary condition seems impossible. This article examines some of the methods of setting the upper boundary condition from various authors found in the research literature. Analyzing their results has led to a complex method of setting boundary conditions on the surface, as well as the procedure of determining heat exchange parameters on the ground surface using local microclimatic characteristics for drawing up physically justified boundary conditions in numerical modeling of geocriological processes.