Geology Reference
In-Depth Information
In uniformly resistant bed materials, the
originally steep knickpoint evolved through
slope replacement into a reach of nearly uni-
form slope without a clear knickpoint (Fig. 8.9B).
On the basis of his experiments, Gardner (1983)
concluded that, in homogeneous bedrock,
parallel knickpoint retreat would occur only if
the rock were pervasively jointed (Box 7.1).
Alternatively, he observed that parallel retreat
could occur in layered rocks of variable
resistance when a highly resistant layer overlies
less coherent rocks (Fig. 8.9A). In this situation,
undercutting of the lip and its collapse would
cause the knickpoint to migrate upstream with
little modification to its shape. Even in the
stream-table experiments in which slope
replacement caused the knickpoint to be trans-
formed into a uniform slope, this transformation
took many days during which a clear knickpoint
migrated upstream through the channel.
When channel networks, rather than single
channels, are the experimental focus, some
stream tables utilize sprinkler systems that “rain”
on a rectangular, unchannelized, gently inclined,
sediment surface. The resulting discharge even-
tually forms a self-organized channel network.
To mimic base-level lowering with such a set-up,
the outlet of a network can be artificially lowered
in order to create a knickpoint (Parker, 1977).
Following base-level lowering, the knickpoint
propagates up the main stem and, as the junction
with each tributary is encountered, a new knick-
point tends to form at the mouth of the tributary
and then begins to migrate up it. In such experi-
ments, the rate of knickpoint migration has been
found to be proportional to the discharge (or
upstream drainage area), such that, as the head-
waters are approached, the rate of migration dra-
matically slows (Fig. 8.10A).
As a knickpoint migrates through a reach, one
can readily envision that the base of the
upstream channel is lowered by erosion and
that the new, lowered channel may be graded
to its downstream continuation. These stream-
table experiments showed, however, that this
geometry is only the first stage in a
complex
response
of the fluvial system that is modulated
by sediment fluxes (Fig. 8.10B). Imagine what
happens to the river bottom at a particular
point. Prior to the migration of the knickpoint,
it “feels” no difference in its base level of ero-
sion and maintains its former geometry. As the
knickpoint migrates through, however, the reach
is steepened, stream power increases, more
sediment is eroded than is deposited, and the
bed is lowered until it is approximately graded
to the adjacent downstream reach (although it
may not have attained the base level of erosion
at this time). As the knickpoint migrates farther
upstream, the sediment flux to that downstream
site increases due to the enhanced upstream
erosion. Without a corresponding increase in
discharge or slope, the increase in sediment
load can cause deposition within the previously
scoured channel, so the base of the channel
rises. Ultimately, as the knickpoints near tribu-
tary headwaters, sediment production decreases,
and the stream incises into its bed once again
(Fig. 8.10B). The multiple events of incision and
aggradation highlight the multi-faceted response
to simple base-level lowering. The fact that
similar responses may be recorded by Holocene
fluvial strata suggests that considerable caution
may be needed when interpreting phases of
aggradation and degradation.
In other stream-table experiments, the
sediment yield was measured as the stream
network adjusted to multiple events of base-
level lowering (Schumm
et al
., 1987). Following
each lowering, a striking initial increase in
sediment yield subsequently decayed exponen-
tially. This decay was not smooth, however,
but was punctuated by secondary peaks and
troughs (Fig. 8.10C) and by high variance. The
secondary peaks in sediment yield were
interpreted to result from sporadic collapses of
valley walls that suddenly shunt extra sediment
into the channel. In recent years, experimen-
tal knickpoint formation, rate of migration,
and role of knickpoints in inducing collapse
of adjacent hillslopes have been further docu-
mented with stereo digital cameras and video
on stream tables with circular sides (Bigi
et al
.,
2006; Hasbargen and Paola, 2000). These
experiments show that knickpoints form even
when base level is steadily lowered. The
cumulative height of successively formed
knickpoints suggests that their upstream
