Integrating groundwater response function into the Jarvis-type model for Populus popularis transpiration simulations

Shallow groundwater is a critical water resource for sustaining vegetation growth in arid and semi-arid environments and affects stand transpiration ( T ) dynamics. However, it is still difficult to quantify the impact of groundwater on T . Here, we introduced a novel groundwater response function in the Jarvis-type model (referred to as MJS G ) and tested its performance using Populus popularis sapflow data over two main growing seasons (2018–2019). The results showed that the performance of the MJS G model depended on groundwater level. Specifically, when groundwater table depth was within 1.2–2.0 m, the precision of daily T simulation by the MJS G model was higher than that by the MJS model without groundwater response function over two years, with an increase in Nash-Sutcliffe Efficiency ( NSE ) from 0.787 to 0.825. Furthermore, in contrast to the MJS model, the MJS G model could better capture the diurnal course of T in 10:00–16:00, with a significant increase in NSE from 0.592 to 0.706. The improvement allows a more accurately estimate of tree water use under shallow groundwater fluctuations, which will help broaden the ecohydrological application of the Jarvis-type model to similar areas. • A new groundwater response function was proposed and tested for transpiration simulations in the Jarvis-type model. • Range of groundwater table depth ( GTD ) improving daily transpiration simulation was 1.2–2.0 m. • Positive influence of GTD on hourly transpiration simulation was in 10:00–16:00.