The recent availability of 1-m laser imaging, detection, and ranging (LiDAR) data in Pennsylvania provides a high resolution digital elevation model (DEM) which could improve on existing USDA-NRCS Order 2 soil survey mapping. The ability of LiDAR derived terrain indices to predict hydric soil presence was evaluated across the Northern Appalachians. We developed a logistic regression model to predict hydric soil presence using a dataset of 1153-field data points and several terrain indices derived from LiDAR DEMs. The best performing regression model included slope derived from a 1-m LiDAR DEM, depressions derived from a 5-m LiDAR DEM, and physiographic region. This model was able to successfully predict 67% of hydric soils and 73% of non-hydric soils from a validation dataset. The model performed better at predicting non-hydric soils compared with hydric soils and was not as effective in low slope areas. This suggests that the 1-m LiDAR hydrologic variables used in the study cannot completely account for soil hydric status.
All Science Journal Classification (ASJC) codes
- Soil Science