A High-Resolution, Random Forest Approach to Mapping Depth-to-Bedrock across Shallow Overburden and Post-Glacial Terrain

Abstract

Regolith, or unconsolidated materials overlying bedrock, exists as an active zone for many geological, geomorphological, hydrological and ecological processes. This zone and its processes are foundational to wide-ranging human needs and activities such as water supply, mineral exploration, forest harvesting, agriculture, and engineered structures. Regolith thickness, or depth-to-bedrock (DTB), is typically unavailable or restricted to finer scale assessments because of the technical and cost limitations of traditional drilling, seismic, and ground-penetrating radar surveys. The objective of this study was to derive a high-resolution (10 m2) DTB model for the province of New Brunswick, Canada as a case study. This was accomplished by developing a DTB database from publicly available soil profiles, boreholes, drill holes, well logs, and outcrop transects (n = 203,238). A Random Forest model was produced by modeling the relationships between DTB measurements in the database to gridded datasets derived from both a LiDAR-derived digital elevation model and photo-interpreted surficial geology delineations. In developing the Random Forest model, DTB measurements were split 70:30 for model development and validation, respectively. The DTB model produced an R2 = 92.8%, MAE = 0.18 m, and RMSE = 0.61 m for the training, and an R2 = 80.3%, MAE = 0.18 m, and RMSE = 0.66 m for the validation data. This model provides an unprecedented resolution of DTB variance at a landscape scale. Additionally, the presented framework provides a fundamental understanding of regolith thickness across a post-glacial terrain, with potential application at the global scale.

The waterscape continuum concept: Rethinking boundaries in ecosystems

“…Existing datasets can be collated to develop new products; Gleeson et al (2011), for instance, used existing geology maps and literature to develop planetary‐scale permeability maps. At the regional‐scale, Furze et al (2021) combined contemporary high resolution light detection and ranging (LiDAR) with historical borehole, trial pit and bedrock outcrop data to model depth‐to‐bedrock at 10 m resolution across the province of New Brunswick. At the local‐scale, the cost of geophysical equipment, such as ground penetrating radar (GPR) and electrical resistivity tomography (ERT), is becoming less prohibitive and is facilitating fine scales studies of ecohydrological processes (e.g., Briggs et al, 2017; Parsekian et al, 2015).…”

Terraced Landscapes as NBSs for Geo-Hydrological Hazard Mitigation: Towards a Methodology for Debris and Soil Volume Estimations through a LiDAR Survey

“…LUC is a routine developed in SAGA GIS: for every cell, it is defined as the local curvatures of the directly neighboring upslope contributing cells, where loca ture is defined as the sum of gradients to the neighboring cells [47,48]. High nega Finally, the volume trapped into terraces was estimated by computing the difference between the actual surface and the interpolated one.…”

“…LUC is a routine developed in SAGA GIS: for every cell, it is defined as the mean of local curvatures of the directly neighboring upslope contributing cells, where local curvature is defined as the sum of gradients to the neighboring cells [47,48]. High negative values of LUC identify the dry-stone wall's basis.…”

Evaluating spatially enabled machine learning approaches to depth to bedrock mapping, Alberta, Canada

“…Recently, information from digital elevation models (DEMs) and other remote sensing datasets has been used with statistical and machine learning models to understand spatial variations in the thickness and composition of surficial materials based on terrain and environmental characteristics. Studies using single-layer neural networks and ensemble tree methods have been used to predict DTB at regional [16,[20][21][22] and global scales [23]. However, most of these studies have achieved only moderate predictive accuracy (around 0.4-0.7 r 2 ) or have been applied in regions with relatively shallow DTB, where the correlation to land surface topography is expected to be high.…”