Use this URL to cite or link to this record in EThOS:
Title: Uncertainties in modelling agricultural phosphorus transfers
Author: Krueger, Tobias
ISNI:       0000 0004 2694 0447
Awarding Body: Lancaster University
Current Institution: Lancaster University
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Modelling phosphorus transfers from agricultural land to receiving waters is important as part of assessingth e risk of pollution and developingm itigation strategiest o meet the demands of international legislation such as the EU Water Framework Directive (2000/60/EC). Current phosphorus models, however, are afflicted by limited experimental support of their underlying assumptions, difficulties in measuring parameters and variables, and ambiguities in calibrating parameters against observations - in other words: uncertainties in data, model structures and parameters. These uncertainties were confronted explicitly in this thesis for two case studies at the grassland field and catchment scale. A pragmatic model learning framework was proposed, which approached the modelling problem from a downward perspective. Model structures, parameterisationsa nd input data were treated as hypothesesw ithin the GeneralisedL. ikelihood Uncertainty Estimation (GLUE) framework. Model input and evaluation data uncertainties were estimated and incorporated into the model diagnostic scheme while maintaining the rigour of hypothesist; e sting. Pragmatismh ad to be retained where statistical information to characteriseu ncertainties was absent. Data uncertainty had significant effects on prediction uncertainty. Model development was approached in the order hydrology, sediment, phosphorus. At the replicated field scale, hydrology was modelled by an ensemble of lumped conceptual store formulations. Hydrological variability between fields resulted in variability of model performance and rejection. Importantly, modelling revealed a 'leaking' of the fields, which helped revise the understanding of the study site. Sediment exhibited hysteretic behaviour, exhaustion effects and flushing effects in surface and sub-surface pathways. Moderate hysteresis could be modelled by a new hysteresis model using a different parameterisation for each event. Phosphorus behaved consistently in time, space and acrossp athways. A power-law phosphorus-sedimenrte lationship simulated the observed phosphorus enrichment at low sediment concentrations. At the catchment scale, hydrology was modelled using the semi-distributed Dynamic Topmodel. Behavioural simulations at the outlet were rejected againstn estedd ischargem easurementsM. odel failure was due to a misrepresentation of drainage pipes and quick-flows as well as uncertainties about the catchment topology. An improved hysteresis model could still not simulate all sediment dynamics, presumably due to quasi-random in-stream erosion and land manakement incidents. Event parametersc orrelatedw ith antecedenwt etness,w hich led to a conceptualm odel of a dynamic transport-/source-limited regime. A constant power-law captured again the dominant phosphorus behaviour. Understanding the sediment dynamics seemed crucial for robust prediction of phosphorus transfers in space and time. However, mechanistic models covering all relevant processes of source dynamics, mobilisation, transport and land management would be difficult to conceptualise, parameterise and test, given the current understanding of processes and measurement capabilities. Caution is advised when using phosphorus models in the regulatory arena without explicit consideration of uncertainties.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available