Modelling the impact of dams and exotic vegetation in New Zealand braided rivers

River planform results from the complex interaction between flow, sediment transport and vegetation, and can evolve following a change in these controls. Disentangling this complex causation path as a preliminary measure to devising restoration measures is not straightforward. We propose a modelling approach that can be used as tool for analysis of observed trajectories and to forecast future behaviours in dam- and vegetation- impacted braided rivers.

We focus two iconic braided river cases in New Zealand’s South Island: the Lower Waitaki River and the Waimakariri River. The Waitaki is impacted by the combined effects of exotic vegetation and a hydropower scheme that has altered the flow regime. As the Waitaki River is unable to clear vegetation efficiently, vegetation encroachment has promoted a shift towards a single-thread morphology. In contrast, the more active Waimakariri River, despite having been subjected to similar vegetation, retains a largely unvegetated channel due to its ability to naturally clear vegetation.

A two-dimensional physics-based numerical model capable of accounting for the evolution of morphology and vegetation in braided reaches is constructed and applied to the two rivers.

Calibration and validation of the vegetation parameter settings, which is critical to obtaining realistic planform styles, is carried out in applications to the two test cases by selecting the parameter values that allow the model to predict vegetation encroachment in the Waitaki and efficient vegetation clearing in the Waimakariri. The model responds sensibly to changes in parameters, showing that more aggressive vegetation types cause a sharper reduction of braiding.

The calibrated model is applied to reconstruct planform changes in the Lower Waitaki under a reconstructed natural flow regime, showing that, even in the absence of the hydropower scheme, the river would have suffered from vegetation encroachment due to its naturally steady hydrology.

Finally, summary metrics that represent vegetation presence in each model are computed and their dependence on the flood frequency is analysed. We find that vegetation presence across rivers and flow regimes can be explained as a function of the duration of periods of vegetation growth, intervening between floods that cause vegetation removal.