Empowering model diagnosis with a water tracer model in WRF-Hydro: influence of lateral flow and groundwater on water transit times

While most traditional land models simplify the representation of terrestrial hydrology to the vertical processes, it is increasingly important to represent lateral water movement as model resolution increases. To understand the role of lateral processes in affecting the water transit times (WTTs) in watersheds, we incorporate a water tracer module in the WRF-Hydro model (WT-WRF-Hydro), which explicitly represents surface and subsurface lateral flows. Comparing with simulations that include only vertical flow, enabling the representation of lateral flow shortens the WTTs in a humid watershed due to the additional water pathways through lateral flow. In contrast, enabling lateral flow extends the WTTs in a dry watershed due to the re-infiltration of surface water during surface lateral flow, highlighting the different effects of lateral flow on WTTs in different watersheds.

Recently, WT-WRF-Hydro has been further applied to six National Ecological Observatory Network (NEON) sites to numerically tag monthly precipitation continuously over a seven-year period. Compared with water isotope measurements, WT-WRF-Hydro tends to underrepresent the seasonal variations of water tracer dynamics, with overestimation of WTTs in four basins and underestimation of WTTs in the rest. The overestimation of WTTs in the four basins contrasts with our general assumption of underestimation of WTTs by models and suggests an overestimation of groundwater storage and inadequacy in water mixing processes in WT-WRF-Hydro at those catchments. Using a combination of modeled and observed WTTs may help us understand and diagnose model deficiencies, highlighting the value of water tracers in hydrologic modeling.