A study investigated whether machine learning could improve the Energy Exascale Earth System Model (E3SM) Land Model's simulation of groundwater table depth (GWTD). Researchers developed a machine learning-supported framework, starting with a stand-in computer model trained on land model simulations, then combined the surrogate model with global observations of GWTD to adjust key parameters related to water runoff in E3SM. This approach improved how well E3SM modeled GWTD compared to simulations using default parameters, particularly within the 1 to 5 meter depth range, where groundwater can influence surface processes. However, this representation of GWTD did not consistently improve how E3SM represented other water-related variables, including soil moisture, runoff, and evapotranspiration. The study demonstrated that machine learning can efficiently calibrate model parameters that would otherwise require computationally expensive global simulations, but it also revealed the limits of calibration strategies relying on a single variable. The results showed that improving GWTD alone does not improve the model's representation of related hydrologic variables, and that calibration based on land simulations driven by atmospheric forcing produces different responses to changes in model parameters in simulations where land and atmosphere are interactive. The findings suggest scientists may need calibration approaches that optimize multiple hydrologic variables and may need to run land and atmosphere models together rather than land models alone.