Hi Kaed,
A useful reference for your question on representing lake–aquifer interaction is my paper published in the Journal of Hydrology:
Yihdego, Y. & Becht, R. (2013). Simulation of lake–aquifer interaction at Lake Naivasha, Kenya using a three‑dimensional flow model with the high‑conductivity technique and a DEM with bathymetry. Journal of Hydrology, 503, 111–122.
Link:
https://www.researchgate.net/publication/258199753_Simulation_of_lake-aquifer_interaction_at_Lake_Naivasha_Kenya_using_a_three-dimensional_flow_model_with_the_high_conductivity_technique_and_a_DEM_with_bathymetryThe study demonstrates a practical and physically consistent way to represent lake boundaries in MODFLOW using the high‑conductivity (HC) technique, which effectively treats the lake as a dynamic head boundary while allowing groundwater–lake exchange to be governed by hydraulic gradients and bed conductance. This approach avoids the oversimplification of a fixed CHD boundary and provides more realistic feedback between lake stage and groundwater heads.
A few points from that work that may help your simplified model setup:
• Boundary representation:
The HC technique behaves similarly to a lake package but is easier to implement in a conceptual model. It allows the lake to act as a head‑controlled boundary without forcing an artificially rigid constant head.
• Conductance matters:
Lakebed hydraulic conductivity and thickness strongly influence the magnitude and direction of exchange. Even in simplified models, representing this conductance term is essential.
• External boundaries:
For regional systems, GHB boundaries are generally more appropriate than fixed heads unless you have strong evidence for constant‑head conditions. They allow the system to respond to the introduction of the lake without over‑constraining the model.
• Processes not to neglect:
Recharge, lakebed leakage, and lateral inflows/outflows are the minimum set of processes needed to capture the first‑order effect of a new lake on groundwater levels.
Although the Naivasha system is larger and more complex, the modelling principles are directly transferable to the type of conceptual problem you described.
Happy to clarify any part of the method if useful.
Best regards,