In this study, we focused on developing a new equation for predicting the initial wave amplitude induced by submarine landslides, based on a smoothed particle hydrodynamics (SPH) model that considers fluid–rigid body interactions. Existing empirical equations were primarily derived from wave flume experiments conducted under limited landslide parameter conditions, which limit their applicability to extended scenarios. To overcome these limitations, a range of landslide parameters was systematically established and replicated through numerical modeling. The water surface response within the simulation was quantified using a numerical sensing approach. The simulated water surface elevations were compared with laboratory wave-gauge measurements with high accuracy, demonstrating sufficient predictive accuracy to complement laboratory experiments. Furthermore, the relationship between major landslide parameters and initial wave amplitude was nondimensionalized and a new equation was derived through log-linear regression analysis. The proposed equation was able to predict initial wave amplitudes observed in both laboratory and field settings, confirming its applicability in the field. These results suggest that the developed equations can be applied to early tsunami risk assessment and early warning systems in coastal areas vulnerable to submarine landslide tsunamis.