Using Seismic Refraction Data to Estimate a Relationship Between Landscape Curvature and Deep Critical Zone Structure in the South Carolina Piedmont, USA

P‐wave velocity profiles from seismic refraction reveal deep critical zone (CZ) architecture along profiles hundreds of meters long. However, extrapolating local velocity measurements to infer CZ architecture at regional scales (1–20 km 2 ) remains challenging. Here, we present a strategy that transforms seismic observations from individual profiles into maps of CZ architecture spanning tens of square kilometers. Data from 15 seismic refraction profiles (approximately 6.6 km total length) collected in weathered crystalline rocks of the South Carolina Piedmont, USA, revealed approximately 400,000 m 2 of deep CZ architecture. Using casing depths from four boreholes, we show that the boundary dividing saprolite and fractured rock corresponds to a velocity of 1,870 m/s. Using velocity measurements from an outcrop within the survey area, we identify the bedrock velocity as 4,550 m/s. These velocities define a three‐layer CZ structure comprising soil and saprolite, fractured bedrock, and unweathered bedrock. We developed an empirical relationship between CZ structure and minimum and maximum principal curvatures, enabling prediction of CZ architecture over approximately 17 km 2 . The correlation between seismically inferred CZ structure and principal curvatures at our study site suggests that curvature metrics can be used to predict CZ structure at larger scales in crystalline terrains under subtropical climates. However, the empirical relationship struggled to predict CZ structure where landscape curvatures were near zero, suggesting that other variables likely contribute to local heterogeneity. Given that curvature is an important variable for erosion and groundwater flow, our results suggest it could be a promising metric for predicting CZ structure.