Use this URL to cite or link to this record in EThOS:
Title: Modelling water and solute transport within vegetated soils using a stochastic framework
Author: Jackson, Bethanna Marie
ISNI:       0000 0001 3587 9236
Awarding Body: Imperial College London (University of London)
Current Institution: Imperial College London
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Restricted access.
Access from Institution:
Models predicting the fate of water and dissolved chemicals in vegetated soils are required for a wide range of applications. Substantial uncertainty is present due to measurement errors, parametric uncertainty, and structural issues related to model con- ceptualisation. Due to the costs and intrusiveness of subsurface measurements there are limited datasets available to interrogate models against. Furthermore, the models are typically computationally intensive, making it di±cult to fully explore parametric and other uncertainty spaces. Hence there are two pressing needs which must be met to improve the utility of models: more data and constraints are required to quantify the impacts of uncertainty, and e±cient methodologies to explore sensitivities and uncer- tainties are also needed. This dissertation presents and applies a stochastic framework addressing the above concerns. Approaches and underlying assumptions to modelling water °ow and solute transport within soils and plants are examined, and two ex- isting models extended. The problem of uncertainty is investigated, and appropriate approaches suggested. Monte-Carlo techniques, including Markov chain Monte Carlo methods, are developed for application to the models, and tested using a comprehen- sive hydrological and radiological dataset from a plot-scale lysimeter experiment. The integrity of the experimental data is examined. Sensitivity analysis and calibration of the hydrological and radiological data sets is performed, with the ability of the model and framework to recover parameters interrogated. Structural uncertainty and e®ects of erroneous inputs are discussed. Results demonstrate the power of the methods to generate insights into process response and quantify uncertainties. The e±ciency of Markov Chain Monte Carlo techniques is demonstrated, but the advantages of retain- ing simple set search methodologies are also clear. Consideration of model structure also signi¯cantly reduces the uncertain parametric space. However, despite the unusu- ally comprehensive experimental dataset, major issues of uncertainty remain, of which data issues are a dominant component.
Supervisor: Butler, Adrian Sponsor: United Kingdom Nirex Limited
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available