Wind turbine wake models that estimate the evolution of the velocity deficit are essential for designing optimal wind farms. A key component of these models is the wake growth rate. We analyze the performance of a recently proposed analytical wake growth rate model and for the streamwise velocity deficit behind an isolated turbine. We consider conditions with different vertical velocity gradients (VVGs) while keeping the streamwise average velocity and ambient streamwise turbulence intensity at the turbine hub height the same across cases. This is done by changing the aerodynamic roughness, friction velocity, and turbine hub height. Comparisons between the predictions of the model and LES data show that the model performance is better for cases with lower VVG. It was also found that the model predicts lateral/spanwise wake widths better than the vertical wake widths.