Modelling the effects of vegetation change on stormflow hydrology
Davie, Timothy John Acton
PublisherUniversity of Bristol
MetadataShow full item record
A study into the effects of vegetation change (particularly afforestation and reforestation) on the hydrology of an area requires a predictive investigative method. This is because empirical studies require a long time period to collate the necessary data and the results cannot be transferred to a remote site with any confidence. Physically based, distributed hydrological modelling offers a predictive capability but the use of effective parameters and lack of verification and validation casts doubts on the ability of the current generation of these models to be used in applications. In this thesis a new modelling scheme is developed that focuses directly on simulating the effects of afforestation on the storm event hydrology of small catchments in humid temperate regions. A mixed conceptual/physically based model (VSAS4), is developed to act as the hydrological base model. In order to parameterise the change in vegetation for VSAS4 a separate, pre-processing, forest growth model is developed. This model is a distance dependent, individual tree based, forest growth simulator. The combination of VSAS4 and the forest growth model is given the name LUCAS (Land Use Change, Afforestation, Simulator). Testing of LUCAS is carried out with verification and validation. The primary aim of these was to assessth e worth of the schemea nd highlight areasf or future research. The difficulty of verifying a complex modelling scheme such as LUCAS has led to the use of a scaled down sensitivity analysis. This is designed to test the schemes robustness and the influence of forest growth on the VSAS4 simulations. The forest growth model is verified and validated separately as an independent predictor of forest growth. The lack of a United Kingdom hydrological data set that spans the period of a forest growth on a catchment has led to only a limited validation of the scheme. The Tanllwyth catchment in Mid Wales was adjudged to be the best available (from a choice of four sites). The scheme cannot be considered a valid predictor of the effects of vegetation change on storm hydrology in humid temperate regions due to the lack of a full data set but the results suggest it cannot be considered invalid. The assessment of LUCAS as a predictor of the effects of vegetation change on storm event hydrology indicates that it has considerable potential but it is not able to be used directly in applications without further development. The validation of LUCAS using hydrograph reproduction and hypothetical scenarios has highlighted several avenues for future research within the study of the effects of vegetation change on stormflow hydrology. These are: the role of canopy closure in vegetation change; the need for soil water flow equations that account for more than just soil matrix flow; and the development of a probabilistic framework so that modelling schemes such as LUCAS can be used in applications without full verification and validation.