Investigating the potential of forested lands for natural flood management in Wales

Abstract

Natural flood management (NFM) is the use of natural processes and environments to mitigate flood risk by reducing peak flood flows. A key aspect of this concept is the management of forested land used for commercial purposes. Large tracts of land in upland Wales are used in this way, and the planning and management of planting and harvesting regimes are influenced by several factors such as market needs, growth rates of plantations and, increasingly, the effects of harvesting large areas of mature forest on groundwater levels and flooding. The development of a method to predict the differing impacts of plantation land scenarios is the subject of this thesis. A comprehensive literature review of the current state of NFM literature finds there are opportunities for deeper insight into the effects of forested land on NFM.
A study site was chosen in upland Wales, which has managed and unmanaged forest areas at differing stages of maturity and instrumented to allow for the collection of a rich dataset of rainfall, throughfall and other meteorological variables. The location of the site and the requirements of the study dictated that a novel method of recording and relaying this data in real-time to a data repository had to be designed and implemented; this was completed successfully, albeit with the usual setbacks and complications experienced when developing and deploying a novel system. Analysis of the data collected is described and presented to demonstrate the collection system and the data collected. The percentage of rainfall measured to be intercepted by each particular cluster during rainfall events is broadly comparable with values in the literature( ranging from 41-57%). The data is valid and robust enough to support the development of an interception model. A further requirement of the study is that the interception model should use remotely sensed data to provide values of Leaf Area Index (LAI) and Fractional Vegetation Cover (FVC). This will ensure that the model can be scaled up from the catchment scale to watershed, regional or even national coverage without the need for excessive and costly fieldwork. After investigation and research into current large-scale LAI products, a method was developed to retrieve LAI and FVC data at a much finer spatial scale than the current products offer. This gives a 10m pixel with a daily temporal resolution. Interception modelling using remotely sensed data has been carried out at disparate sites worldwide, and a model based on Gash’s well-known 1979 interception model was developed by Cui and Jia in China. This model has been modified to be more relevant to the forest conditions found in the UK and validated using data collected from the study site with a high degree of confidence in the results. Estimations of the average interception values for each cluster achieve Kling-Gupta Efficiency (KGE) values between 0.68 and 0.89 across all monitored forest clusters when compared with the measured values.. The interception model is then integrated into an existing hydrological model (EXP-HYDRO) and used to predict streamflow under various forest-type configurations and harvesting scenarios. The impact of clearcutting the entire catchment is shown to predict an increase in streamflow of 76%, while a more conservative approach (only clearcutting one watershed) is shown to only increase streamflow by 20-30%. The integrated model is shown to perform well when parameterised to represent the study area in its current configuration, which builds confidence that the streamflow outputs predicted for the scenarios of a homogenous forest and different harvesting scenarios are accurate. This project has successfully fulfilled the objectives set out in the original brief, and the integrated model has the capability to model streamflow based on remote sensed LAI and FVC values, making it a potentially valuable tool for aiding and informing forest management planning in the future.

Date of Award	17 Feb 2026
Original language	English
Awarding Institution	Bangor University
Sponsors	Knowledge Economy Skills Scholarships (KESS 2)
Supervisor	Sopan Patil (Supervisor), Morag McDonald (Supervisor) & Andy Smith (Supervisor)