A surface model for water and energy balance in cold regions accounting for vapor diffusion

• Main Authors: Dandar, Enkhbayar, Saaltink, Maarten W., Carrera, Jesús, Nemer, Buyankhishig
• Format: Text
• Language: English
• Published: 2018
• Subjects: permafrost; Subarctic
• Online Access: https://doi.org/10.5194/hess-2017-452; https://www.hydrol-earth-syst-sci-discuss.net/hess-2017-452/

Computation of recharge in subarctic climate regions is complicated by phase change and permafrost, causing conventional conceptual land surface models to be inaccurate. We conjecture that large vapor pressure gradients, driven by the large temperature difference between the soil surface and the thawing permafrost active layer, may cause a significant water and energy transfer during late spring and early summer. To analyze this conjecture, we develop a two-compartment water and energy balance model that accounts for freezing and melting and includes vapor diffusion as a water and energy transfer mechanism. It also accounts for the effect of slope orientation on radiation, which may be important for high latitude mountain areas. We apply this model to weather data from the Terelj station (Mongolia). We find that vapor diffusion plays an important quantitative role in the energy balance and a relevant qualitative role in the water balance. Except for snowmelt and a few large precipitation events, most of the continuous recharge is driven by vapor diffusion fluxes. Large vapor fluxes occur during spring and early summer, when surface temperatures are moderate, but the subsoil remains cold, creating large downwards vapor pressure gradients. Temperature gradients reverse in fall and early winter, but the vapor diffusion fluxes do not, because of the small vapor pressure differences at low temperature. The downwards latent heat flux associated to vapor diffusion is essential for the thawing of the active layer. On a yearly basis, it is largely compensated by heat conduction, which is much larger than in temperate regions and upwards on average. Furthermore, we find that total surface runoff is small and concentrated at the beginning of spring due to snowmelt. Recharge is relatively high and delayed with respect to snowmelt because a portion of it is associated to thawing at depth, which may occur much later.