Home > Research > Publications & Outputs > Reduction of rainfall reaching the ground by de...

Electronic data

Text available via DOI:

View graph of relations

Reduction of rainfall reaching the ground by deciduous trees on the edge of stands - a study into incremental wet canopy evaporation through storm events

Research output: ThesisMaster's Thesis

Publication date13/01/2021
Number of pages189
QualificationMasters by Research
Awarding Institution
Thesis sponsors
  • Lune Rivers Trust
  • The Woodland Trust
Award date13/01/2021
  • Lancaster University
<mark>Original language</mark>English


Can deciduous trees be used to reduce hydrograph peaks? This is an intriguing question to scientists, charities, and the public alike and the answer is yes. It is beneficial to understand how hydrological processes are influenced by edge effects and differs from other studies that are 100+m into woodlands, due to the planting of narrow belts for natural flood management.

This research quantifies the WCE (Wet Canopy Evaporation) of a mature Chestnut tree through storms between May 2018 and April 2019, combined with manual weekly volumetric data from three Beech and two Oak trees in a narrow woodland on the Lancaster University campus. From the six trees studied, the WCE% during the measurement period was 41.97% of gross-rainfall, while the remainder of the gross-rainfall was partitioned into throughfall (54.39%) and stemflow (3.65%). The WCE% of the tree species (excluding the chestnut tree) are significantly higher than that found in other studies where measurements are taken 100s of metres into woodlands away from edge effects; but this data agrees with Herbst et al.’s (2006) findings relating to WCE of hedges. In part this is due to the greater ventilation of the canopy and stems. Although lower than during the leafed periods, WCE remained high from leafless branches and stems. It was also found that the WCE% decreases as storm size increases. The Frumau horizontal and vertical rain gauges (Frumau et al., 2011. Hydrological Processes 25: 499-509) found that horizontal rainfall caused under-estimation of gross rainfall collected by the tree, which was influenced by wind speed and direction. Negative WCE (i.e. larger throughfall than rainfall) was seen as the tree collected rainfall from a larger area. If corrected rainfall was known the WCE would be larger.

The Penman equation showed a poor fit, overestimating evaporation as it shows potential WCE. With little improvement in the utilisation of stores, the Rutter original underestimates evaporation that occurs. The Rutter Sparse model provided the best fit, but was still poor, underestimating evaporation. The Rutter Sparse parameters were altered showing the best fit altering the aerodynamic resistance to 5s/m rather than converting Hazelrigg weather stations wind speed. Alternatively, the canopy capacity was increased and throughfall coefficient decreased to produce a good fit, however these were calculated using the data collected suggesting the best alterations to the model account for the edge effect better by altering the aerodynamic resistance. The research highlights how model parameters representative of conditions at the centre of large woodland blocks should not be used to estimate WCE for narrow belts of trees. Indeed, narrow tree belts could be considered as potential ‘hot spots’ of evaporation requiring more direct measurements to understand their significance as a tool for removing net-rainfall from catchment systems during flood peaks.