The Critical Zone (CZ) is the complex interaction of the hydrosphere, atmosphere, lithosphere, biosphere and pedosphere. It is in the CZ where most biological activity on earth can be found. At the centre of the CZ, the pedosphere is the medium in which all other regions of the CZ interact. The main objective of this study was to model two aspects of the CZ: the presence of exposed bedrock (EB) areas and the depth of the pedosphere (soil depth) in the Tulameen region of Southern British Columbia. Random Forest (RF) a classification tree method was used to predict the presence of EB. Prediction accuracy was found to be 88% with an independent validation dataset. The top three predictors of EB presence, which are a Landsat 7 PCA, Topographic Ruggedness Index (TRI), and a Normalized Difference Vegetation Index (NDVI) were further explored with modified partial dependence plots (PDPs) to determine the probability of EB presence. The depth of the pedosphere was predicted with a Generalized Linear Model (GLM), Random Forest (RF) and Residual Kriging (RK). Depth measurements came from the predicted EB layer which acted as a proxy for 0 m depth. In addition well water and soil pit information were used to define deeper depths for the region. GLM with RK was determined to produce the best model to measure depth, with an RMSE of 0.9 m in the 0 to 2 m range for depth measurements. EB proved to be a reliable and efficient proxy in addition to conventional soil depth measurements which are time consuming and costly to generate. The obtained results indicate that GLM with RK and the use of EB layers can aid in further studies of the CZ.
Copyright is held by the author.
This thesis may be printed or downloaded for non-commercial research and scholarly purposes.
Supervisor or Senior Supervisor
Thesis advisor: Schmidt, Margaret
Member of collection