Evaluating the Potential of Ground Penetrating RADAR to Model Soil Restrictive Layers.

Spatial information on the presence and depth to subsurface restrictive layers are needed to inform hydrologic studies; however, this information is frequently unavailable due to the high cost and labor involved in traditional surveying techniques. Compared to pedon descriptions produced by hand excavation or vehicles, ground penetrating radar is capable of generating valuable spatial information on soil restrictive layers over a greater extent more quickly. We generated, 86 GPR transects in two adjacent Pennsylvania Ridge and Valley Province watersheds to determine depth to, and presence or absence of, soil restrictive features. Horizon depth and morphological properties from eight soil pits and 86 soil cores were used as a basis for interpretation of reflection signatures on radar images. Correlation and logistic regression analysis was used to evaluate relationships between the depth and presence of a radar reflections identified using GPR and 1-m LiDAR derived landscape parameters, including aspect, slope, curvatures, specific contributing area (SCA), topographic position index (TPI), and topographic wetness index (TWI). Comparisons between pedon descriptions of restrictive layers and radar image signatures of the same restrictive layers indicate that GPR produces strong radar reflections for several types of soil features, including: fragipans; horizons with fragic properties; high clay subsoils; and BC and CB transitional horizons. However, differentiating between radar reflections without corresponding morphological information is difficult due to similar radar reflection patterns. Logistic regression analyses indicates that soil features exhibiting strong radar reflections have a higher probability of occurring in lower landscape positions on north aspects (FD-36) and in lower curvature positions on south aspects (Mattern). Our results suggest that GPR reflection signatures coupled with site specific pedon descriptions can greatly extend land manager knowledge of the distribution of restrictive layers and their potential to influence surface runoff (i.e. colluvium-derived restrictive layers) in these watersheds.