Dissertation/Thesis Abstract

Predicting Soil Health and Function of the Barnes Catena Using Evapotranspiration, Vegetative, Geologic, and Terrain Attributes in the Eastern Glaciated Plains of North Dakota
by Bohn, Meyer Patrick, M.S., North Dakota State University, 2018, 465; 10790604
Abstract (Summary)

The benchmark Barnes soil series is an extensive northern Great Plains upland Hapludoll that is vital to the region. Accelerated erosion has degraded Barnes agricultural soil quality, but with unknown extent or severity. Samples from three extensive Barnes soil map units, stratified by evapotranspiration values, were collected to 50 cm and analyzed for chemical, morphologic, and physical properties germane to edaphic function. Multi-scale terrain attributes and remote-sensed soil proxies, and geologic covariates were implemented with Cubist to model soil properties. Best models included SOC, EC, pH, SOC-IC, and sand content. Pedons were classified with a clustering algorithm into six classes. Linear discriminant analysis of covariates and subsequent prediction of landscape grouped classes had moderate to nearly substantial agreement with field observations; only fair agreement was attained for all classes. Detailed morphologic observations confirmed extensive topsoil erosion for some landscape positions that merit investigation of soil function and potential state change.

Indexing (document details)
Advisor: Hopkins, David G.
Commitee: DeSutter, Thomas, Gasch, Caley K., Steele, Dean D.
School: North Dakota State University
Department: Soil Science
School Location: United States -- North Dakota
Source: MAI 57/06M(E), Masters Abstracts International
Subjects: Soil sciences
Keywords: Digital soil mapping, Erosion, Glacial till, Machine learning, Pedology, Remote sensing
Publication Number: 10790604
ISBN: 978-0-355-98796-6
Copyright © 2021 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy