Modeling groundwater flow and solute transport in the active layer of hillslope system in permafrost environments

Subsurface hydrology in regions dominated by permafrost is expected to change as a response to global climate change. Groundwater transports energy as well as dissolved solutes such as contaminants and carbon. To investigate the changes in advected energy as well as potential implications for solute transport, we created a permafrost hillslope modeling study that simulates current day active layer hydrology as well as future conditions based on climate projections.

Simulations are conducted with a state-of-the-art physically based numerical model (ATS) and combine a generic modeling approach with site-specific boundary conditions representative of the Adventdalen valley in Svalbard. We find that in the current climate, the subsurface hydrothermal state of the active layer along the hillslope transect is affected by lateral groundwater flow through differences in moisture distribution up- and downhill. Although lateral heat advection along the transect was found to be negligible, we show that the moisture distribution by gravitationally-driven seepage flow along the hillslope leads to unexpected temperature differences between the uphill and downhill parts of the transect. A non-negligible warming effect is observed uphill, resulting in deeper active layer depths than downhill.

Additionally, preliminary results based on transport modeling indicate that solute migration is mostly longitudinal and slow due to low liquid saturation of the active layer in summer. Under warmer conditions (increased air temperatures), lateral heat advection is expected to increase with more available energy, but solute migration may be partially counteracted by a greater volume of unfrozen soil in summer caused by less saturated conditions closer to the surface.

Furthermore, we discuss the potential implications this has for subsurface transport of solutes and dissolved constituents, and highlight challenges for numerical modeling of these systems.