4. Conclusions
At temperate locations around the world, high wet-canopy evaporation losses have been observed from forests using canopy water balance methods during large and extreme rainfall events and are associated with significant variability. Wet-canopy evaporation of up to approximately 40 mm d-1 have been recorded for large rainfall events (>50 mm d-1) and across all events range between approximately 2 and 38% of gross rainfall. Taken at “face value” these evaporation losses are qualitatively significant in the context of flood mitigation resulting from tree planting. Theoretical wet-canopy evaporation estimates made using the Penman-Monteith model for large and extreme events in mountainous regions of the UK suggest consistency with these high observed losses but uncertainties associated with the estimation of, in particular, aerodynamic exchange are so large that this test of consistency remains weak. During 4 major flood events in the Cumbrian Mountains, UK, meteorological conditions were favourable for high rates of wet-canopy evaporation: high windspeeds prevailed and surprisingly low relative humidity was observed at some locations. Thus the disparity regarding the significance of wet-canopy evaporation for flood mitigation between conclusions based upon results from catchments studies of forest cover effects and results from forest plot studies remains.
Our results suggest that it is possible for high rates of Ewcover forest to occur during large flood events in mountainous regions of the UK but not in all locations and not for all events. To be able to determine the potential of tree planting scenarios on flood hydrographs using hydrological models, estimates of the spatial and temporal patterns of wet-canopy evaporation through sequences of rainfall events are needed. Appropriate estimates require simulation of the control imposed by meteorological variables on wet-canopy evaporation to be made necessitating interpolation and extrapolation from (normally) sparse meteorological observation sites. This is difficult to implement with any accuracy and the uncertainties associated with this step are compounded with the large uncertainties associated with the estimation of aerodynamic exchange with forest canopies such that wet-canopy evaporation estimates may be represented best as scenarios. Simulations must also be able to represent the limitation on evaporation imposed by storage on the surfaces of different vegetation canopies (e.g. between foliated and unfoliated deciduous trees) meaning that scenarios of parameterisations must be incorporated into the simulations. Simulation scenarios must be associated with a confidence weighting that can be propagated to simulation results as expressions of modelling uncertainty.
If the considerable uncertainties associated with estimating meteorological conditions and Ewc across large areas such are to be constrained, collection and analysis of a larger number of well-placed and well-distributed meteorological observations is required, combined with concurrent wet-canopy evaporation observations.