• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • No language data
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Advancing Digital Soil Mapping and Assessment in Arid Landscapes

Brungard, Colbe W. 01 May 2014 (has links)
There is a need to understand the spatial distribution of soil taxonomic classes, the spatial distribution of potential biological soil crust, and soil properties related to wind erosion to address land use and management decisions in arid and semi-arid areas of the western USA. Digital soil mapping (DSM) can provide this information. Chapter 2 compared multiple DSM functions and environmental covariate sets at three geographically distinct semi-arid study areas to identify combinations that would best predict soil taxonomic classes. No single model or type of model was consistently the most accurate classifier for all three areas. The use of the “most important” variables consistently resulted in the highest model accuracies for all study areas. Overall classification accuracy was largely dependent upon the number of taxonomic classes and the distribution of pedons between taxonomic classes. Individual class accuracy was dependent upon the distribution of pedons in each class. Model accuracy could be increased by increasing the number of pedon observations or decreasing the number of taxonomic classes. Potential biological soil crust level of development (LOD) classes were predicted over a large area surrounding Canyonlands National Park in Chapter 3. The moderate LOD class was modeled with reasonable accuracy. The low and high LOD classes were modeled with poor accuracy. Prediction accuracy could likely be improved through the use of additional covariates. Spatial predictions of LOD classes may be useful for assessing the impact of past land uses on biological soil crusts. Threshold friction velocity (TFV) was measured and then correlated with other, easier-to-measure soil properties in Chapter 4. Only soils with alluvial surficial rocks or weak physical crusts reached TFV in undisturbed conditions. All soil surfaces reached TFV after disturbance. Soils with weak physical crusts produced the most sediment. Future work on wind erosion in the eastern Great Basin should focus on non-crusted/weakly crusted soils and soils formed in alluvium overlying lacustrine materials. Soils with other crust types are likely not susceptible to wind erosion. Threshold friction velocity in undisturbed soils with weak physical crusts and undisturbed soils with surficial rocks was predicted using a combination of penetrometer, rock cover, and silt measurements.

Page generated in 0.0314 seconds