Canterbury Plains of South Island. It leaves the mountains through a bedrock gorge. Above
the gorge, the valley of the Rangitata is braided, and it appears that the Rangitata should be
a braided stream below the gorge, as are all the other rivers which cross the Canterbury
Plain. However, below the gorge, the Rangitata is meandering. A few miles farther
downstream, the river cuts into high Pleistocene outwash terraces, and it abruptly converts
from a meandering to a braided stream. The braided pattern persists to the sea. If the
Rangitata could be isolated from the gravel terraces, it probably could be converted to a
single-thalweg sinuous channel, because the Rangitata is obviously a river near the pattern
There are other New Zealand rivers that are near the pattern threshold. and therefore, they
are susceptible to pattern change. In fact, New Zealand engineers are attempting to
accomplish this pattern change in order to produce 'single-thread' channels which will cause
less flood damage, have greater stability at bridge crossings, and be less likely to acquire
large sediment loads from their banks and terraces. For example, the engineers have had
success in converting the Wairau River, a major braided stream, from its uncontrolled
braided mode to that of a slightly sinuous, single-thalweg, relatively more stable channel.
The increase in sinuosity is only from 1.0 to 1.05, and this was accomplished by the
construction of curved training banks. On Figure 5.24 the Wairau River plots close to the
threshold line, and with the reduction of sediment load produced by bank stabilization, it
appears that the pattern threshold can be crossed successfully.
Farther to the south near Kaikoura, the Kowhai River is being modified in the same manner
as was the Wairau. Whereas much of the sediment load in the Wairau River is derived from
bank and terrace erosion which can be controlled, high sediment loads are delivered to the
Kowhai River directly from steep and unstable mountain slopes. On Figure 5.24 the Kowhai
River plots well above the threshold line, and without a major reduction in upstream
sediment, it may be difficult to maintain a single-thalweg channel at this location.
The variability of the Rangitata River pattern indicates that braided to single-thalweg
conversions should be possible for the Chippewa and Wairau Rivers. However, not all
braided rivers can be so readily modified, as this depends on their position with regard to the
line defining the pattern thresholds on Figures 5.24 and 5.25.
Perhaps the simplest way to determine the most likely pattern for a river is to determine its
pattern from the earliest maps and aerial photographs. If the river was braiding in the past, it
seems unlikely that an attempt to convert to a sinuous channel will be successful. However,
if the historic river was meandering and it is now braided, although there have been no major
erosional or hydrologic changes in the drainage basin, then a conversion back to a
meandering pattern is appropriate. Another approach is to determine the characteristics of
other nearby rivers. If the subject river is very different it may be converted to the local
character if neither the sediment load or hydrologic character are significantly different.
5.6 MODELING OF RIVER SYSTEMS
The necessity for quantitative prediction of river channel response is increasing. The accuracy
of such a prediction depends on the quality of the data. There are generally two ways of
predicting response. One is the mathematical model and the other is the physical model.
Mathematical models utilize a number of mathematical equations governing the motion of
water and sediments in a channel. Regardless of the potential of mathematical models, to date
they have been best used to study channel response using 1-dimensional, or at most 2-
dimensional, approximations. For complex three-dimensional channel processes, it is very
difficult to accurately formulate mathematically what happens in a river. Studies of channel