Modeling the impact of spatiotemporal vegetation dynamics on groundwater recharge

Abstract Climate change affects the growth of vegetation and its physiological states such as leaf area index (LAI), which in turn can affect groundwater recharge because of changes in evapotranspiration (ET). Presently, most recharge modeling studies over-simplify transient vegetation conditions and the potential corresponding impact on recharge by using climatological values of vegetation parameters such as LAI. Our study uses the Community Land Model (CLMv4.5) to investigate the sensitivity of recharge to interannual varying vegetation in Minnesota (USA) across different climate, hydrogeology, and ecoregions at a 25 km spatial resolution and for the period of 2000–2015. The Ensemble Kalman Filter (EnKF) was used to calibrate soil and runoff parameters to statewide water table depth observations. Results of the study indicate that although year-to-year varying vegetation does not affect long-term climatological recharge estimates, it can drive disproportionately large variability in annual and seasonal recharge. Comparing simulations with dynamic and climatological vegetation inputs, the average magnitude difference (root mean square difference, RMSD) for recharge was 21.1% in response to only a 4% difference in LAI inputs. Regression analysis revealed that the combination of local hydrogeology and vegetation-type affects the magnitude of recharge response to LAI and ET changes. We also found cross-ecoregion dominance of temperature rather than precipitation controlling LAI anomalies and resulting recharge variability, with springtime temperature being the primary factor because of its impact on leaf-out conditions. Drier western Minnesota showed higher relative LAI differences as well as higher spring time and relative annual recharge compared to the wetter eastern part of the state, indicating higher vulnerability of the water-limited region to changing vegetation and climatic conditions. Our study shows that models can underestimate or overestimate annual and seasonal recharge if vegetation dynamics are neglected, demonstrating the need to incorporate transient vegetation conditions when assessing the impact of future climate change on recharge.