Generic autogenic behaviour in fluvial systems: lessons from experimental studies - From Depositional Systems to Sedimentary Successions on the Norwegian Continental Margin - page 13

0 ≤β< 1 β =1

β> 1

F a

β

R s

Fig. 7. Relation between channel-belt stacking density and sedimentation rate for three possible regimes defined by the
exponent β in a power law relation between avulsion frequency and sedimentation rate. Dark grey indicates channel belt
sand bodies and yellow indicates flood plain fines (redrawn from Bryant et al . 1995). Experiments suggest that the relation-
ship pictured in the right hand side panel is most likely to occur in nature, which is an increase of avulsion rate with
aggradation rate causing channels to stack more densely resulting in higher connectivity. F a is the frequency of avulsions
and R s is the aggradation rate.

channel-belt stacking density and hence connect-
edness is directly correlated to lateral (horizontal)
changes in sedimentation rate. Leeder (1978) sug-
gested that reduction in subsidence rate with time
increases the stacking density by allowing chan-
nel belts more time to remove floodplain fines.
Bryant et al . (1995) examined various forms of
coupling between avulsion frequency and aggra-
dation rate by examining their exponential rela-
tionship. If F a is the frequency of avulsions and R s
is the aggradation rate then F a ≈ R s

findings of Hickson et al . (2005), who state that
only if subsidence is faster than aggradation rate,
will the river adjust and migrate to the topographic
low that is formed.

Reconstructions of generic avulsion behaviour

For reconstructions of generic avulsion behaviour,
detailed surface and subsurface mapping in com-
bination with good age control is needed. Much of
the hypothesis launched here still needs to be
tested by thorough fieldwork, which at present
gives ambiguous results. The cases dealt with
below are nothing more than examples that help
to demonstrate the frequency of avulsion and its
relation to aggradation rates and are not meant as
an exhaustive review.

β , where β is a
positive, real valued exponent. This leads to three
qualitatively different regimes (Fig. 7) with β = 0
resulting in a constant avulsion frequency, as
assumed in Leeder's (1978) model. For β = 1 the
stacking pattern is independent of aggradation
rate and for β > 1 the autogenic behaviour would
increase with aggradation rate; this case is evident
for all laboratory models presently known. This
means that maximal removal of floodplain fines
and greatest connectivity of channel bodies would
occur if aggradation rates are highest.
Hickson et al . (2005) conclude, on the basis of
their findings, that the two-dimensional variation
in alluvial architecture is controlled very strongly
by externally forced sedimentary facies migra-
tions such as changes in sediment supply, base
level or subsidence. However, the three variables
together control the aggradation rate, the basic
control on facies change. If the imposed variations
are slow then facies migrations are kept at a mini-
mum but if they are relatively fast (as in some of
Hickson et al ., 2005 runs), then they become a
dominant control on alluvial architecture. Leeder's
(1978) point about the effect of the lateral changes
in sedimentation rate (stating that avulsion rates
must be highest at subsidence maxima and lowest
at subsidence minima, while the overall lateral
stacking density of channel belts may remain
unchanged) agrees well with the experimental

Steep-gradient and moderately-gradient
systems

Scott and Erskine (1994) studied twelve similarly
sized Australian alluvial fans all subjected to the
same catastrophic, rain-triggered, floods. The fans
and catchment areas involved have similar sizes
and gradients and were all located in a zone which
received very similar rainfall intensities. Hence,
the fans were subject to similar but significant
flood discharges. Of the 12 fans, seven were
entrenched and five were not before the storm
event. The fans reacted in a different way to the
storm event. Effects ranged from no change at all
to trench incision or backfilling. Scott & Erskine
(1994) propose that each fan showed a different
stage of a similar autogenic cycle. The cycle
consists of: (i) aggradation of the fan; (ii) the initia-
tion of a fan-head trench due to exceeding the
threshold slope; (iii) coalescence of scour pools to
a continuous trench; and (iv) backfilling of the
trench due to its widening and slope reduction.