Does where you plant trees make a difference in hydrologic response?

Abstract

Riparian zones are considered to be a good way of reducing water flow and sediment losses to streams, but is planting trees further away from the stream bank just as effective? Here we have used a combination of analytical models and numerical models to estimate the likely effects of the positioning of trees in a catchment on the hydrologic response. An analytical model of a planar slope was used extended in a piecewise manner to determine the effect of varying roughness of a section of the slope on runoff depth, velocity and quantity. This was compared to a numerical solution of the full flow equation on a slope. Results show that the analytical solution predicts a larger runoff depth than the numerical solution, which is to be expected as it ignores some of the terms in the full solution. The numerical model shows the same abrupt transient in head (height of water on soil surface) at a change in roughness assumed in the analytical model. A uniform planar slope of length of 100 m was split into 4 equal quarters and the effect of slope, runoff rate and roughness on the discharge rate at each quarter and at the bottom of the slope was investigated with the analytical model. This showed that the discharge rate would change in quarter with different roughness but relax back to the original discharge rate in the next quarter of the slope, when the changed occurred in the upper 3 quarters of the slope. Only when the roughness change occurred in the last quarter of the slope was the discharge rate affected at the bottom of the slope. Slope angle was found to have the least effect on changing discharge rate at the bottom of the slope. The numerical solution though, could not produce a stable solution when the length of the slope length, runoff rate, roughness and slope angle were large, while the analytical solution was able to produce results in all cases considered. Neither the analytical or numerical solutions of flow down the sloping surface included the effect of prior soil conditions on the amount of runoff generated. In order to investigate soil and climate effects on runoff the problem was also solved using the THALES catchment model. Results with the catchment model THALES generally supported the analytical model but also allow the climate and soils (infiltration and evapotranspiration) when the vegetation was changed to be assessed. Three contrasting sites were chosen; Melbourne, Brisbane and Perth, along with three soil materials (clay(C), clay loam (CL) and sandy loam (SL)). The soil materials were used to created soil profiles with four 0.3 m layers (total depth 1.2 m); soil#1 SL for all four layers; soil#2 SL for top layer and CL for lower 3 layers; and soil#3 C for all 4 layers. Two slopes; A1-10°and A2-30°were used, and combined with three soils and 3 sites resulted in 18 scenarios. The results showed that the planting of trees at different positions of the slope had an effect for sandy loam soils and moderate slopes in a winter dominated rainfall climate like Melbourne. However, for a summer dominated rainfall like Brisbane the position of the trees did not significantly affect the amount of runoff generated. The position of where the trees were planted was only shown to have a significant effect in Melbourne, where the further down slope the trees were planted the more they intercepted subsurface flow and evapotranspired this, reducing runoff. Again slope was shown to have the smallest effect on runoff except for two of the scenarios investigated. Given the time requirement to run the analytical solution offers a quick and reasonably robust way to determine the effects of changes in soil, typographic, vegetation and runoff rate on runoff generated at the bottom of a planar slope.