The goal of this project was to provide an analytical model on rainwater interception performance of a selection of common urban trees in the Metro Vancouver area, given a series of climatic and tree characteristics. Overall, the model performed with a reasonable capacity to simulate the interception loss and results mimicked actual observation, given limited data inputs and with stated assumptions. It should be remebered that the model summarizes general conditions. Minor disagreements between modelled and measured data are reasonable, as the measured data only represent the interception loss of one tree species for one rainfall event.
Despite the satisfactory outcomes, there are a few limitations identified in this study:
- E/R, S, and p were assumed to be constant over the whole rainfall event. This assumption could lead to the discrepancy between modelled interception loss and actual interception loss. In fact, both E and R should correspond to the period after tree canopy is completely saturated.
- Leaf area index (LAI) values were applied from a different study area. LAI varies in different environments, even for the same tree species. For example, various types of land use could lead to different LAI values for the same species, because it measures the total leaf surface area (one side) divided by land area (Nowak, et al., 2013). Deviation in LAI could cause biases in the estimation of both S and p in this model, as S and p were derived from their relationships to LAI.
- Variations in leaf phenology of different species were ignored. Leaf phenology determines the timing of the emergence of leaves; the growth of leaves and leaf fall (Rodriguez et al., 2014). This model simply assumed the leaf-on season is spring, and the leaf-off season is winter for all broadleaf species, which is not true in reality. Although no dramatic changes in interception pattern were expected for the four selected species if leaf phenology were specified, the cumulative amount of interception might vary in different species.
Sensitivity analysis of the model parameters and the feasibility of throughfall data measurements should be considered when implementing the rainfall interception model. Special attention should be paid to obtaining more precise values for key parameters, as they are the fundamental components to achieve satisfactory model performance. Some parameters require more time and effort than others to estimate. Recommendations to address the limitations are:
- Prioritizing efforts on obtaining E/R, if resources and time are constrained. Among the three key parameters, the model is the most sensitive to E/R. Optimizing could improve the model performance at a higher degree than the other two parameters.
- Estimation of R is critical. Special attention should be paid to the choice of method used to estimate R, because it is used to determine the amount of time the canopy is saturated and should correspond to the hours when rainfall equals or exceeds a given threshold. Although it was not possible to test the impact of R alone on the model in this study, sensitivity analyses suggested R tends to be more influential than E (see Sensitivity Analysis).
- Measure LAI of target tree species. Commercial instruments such as LAI-2000 and AccuPAR ceptometer are often used for LAI measurement, but these instruments can be expensive and are characterized by a low portability. Alternatively, a newly developed cell phone app – Pocket LAI is an innovative way to make LAI measurements. Many studies have tested the Pocket LAI on different species and reported that it is a suitable alternative to the other commercial tools for estimating LAI, especially when resources and portability are the key issues (See applied studies).
- Specify foliage months for each species. High-resolution remote sensing images could provide a relatively accurate match with the ground observation regarding the detection of green-up dates (Polgar and Primack, 2011); however, this approach could be limited by economic constraints. Leaf phenology data of some plants might be obtained at The Japanese Metrological Agency, which has been recording leaf phenology data in phonological gardens at over 100 weather stations since 1953 (Ibáñez et al., 2010). However, different growth environments could result in changes in foliage timing.
For more details, please download the Full Report