|
<br />54 GEOMORPHlCALL Y EFFECTIVE FLOODS
<br />
<br />and measurements of the magnitude of channel and
<br />floodplain disruptions generated by a flood.
<br />Data in Table 2 are not exhaustive, but the information
<br />demonstrates the kinds of data required to compute the
<br />energy expended per unit area by a flood. The absolute
<br />value of flood energy expended per unit area, or the
<br />average flood power, may provide no clear differentiation
<br />of effective and ineffective floods. Floods with a relatively
<br />low average stream power and expended energy can
<br />produce catastrophic impacts on alluvial channels and
<br />floodplains, such as during the Plum Creek. Colorado
<br />flood in 1965 [Osterkamp and Costa, 1987]. Other floods
<br />like the Centralia, Washington, and Porter Hill, Oregon
<br />floods, with five times the peak stream power, and over
<br />ten times the average stream power of the Plum Creek
<br />flood, can cause only minimal changes. Likewise, long-
<br />duration floods on the Mississippi River are capable of
<br />generating large values of total energy, but minimal
<br />geomorphic changes, because the peak stream power per
<br />unit area is too low to exceed resistance thresholds of its
<br />channels and floodplain_ Apparently, effective floods
<br />require some optimal combination of stream power,
<br />duration, and energy expenditure. This optimal
<br />combination depends on the floodplain and channel
<br />resistance thresholds, and the hydrologic characteristics of
<br />a panicular fluvial system. More data like those in Table
<br />2 will help clarify this important problem. In the next
<br />section we propose a model to guid~ these investigations.
<br />
<br />4. EFFECTIVE FLUVIAL EVENTS: A MODEL TO
<br />INCLUDE FLOW DURATION
<br />
<br /><--.
<br />
<br />The ability to compute the distribution of stream power
<br />per unit area of a flood throughout the hydrograph,
<br />combined with consideration of potential -landsurface
<br />resistance thresholds, allows us to constroct a conceptual
<br />model of geomorphically effective floods (Figure 11)_
<br />Three hypothetical stream-power graphs are plotted in
<br />Figure I I. Curve A represents a flood of long duration but
<br />very low peak stream power. Total energy generated by
<br />the flood at a particular site, represented by the area under
<br />the stream-power graph, may be large. But in spite of a
<br />large total energy expenditure, and long flow duration,
<br />peak stream power never rises above the threshold required
<br />to significantly disrupt alluvial channels and floodplains.
<br />There has been some effort to identify minimum thresholds
<br />of critical stream power and boundary shear stress for
<br />alluvial systems, but far more work in a variety of
<br />environments is required [Magilligan, 1992; Prosser and
<br />Slade, 1994]. Great floods along large, low-gradient rivers
<br />such as the Mississippi River flood of 1927, which
<br />
<br />00
<br />~oo
<br />;:ffi
<br />ztii
<br />--:2
<br />O:w
<br />wo:
<br />;:..
<br />0::>
<br />"-0
<br />::;00
<br />,,0:
<br />ww
<br />0:"-
<br />t--
<br />CJ)
<br />
<br />
<br />_~~q~~q~ ~!q~i_o_f'!. !t!r.e_s_h_oJ~
<br />
<br />Energy available for
<br />geomorphic change
<br />
<br />Alluvial erosion threshold
<br />-. --- -- - -Minimal-erosion-
<br />
<br />TIME
<br />
<br />Fig_ II. Conceptual stream-power graphs used to document
<br />geomorphic effectiveness of different kinds of tloods.
<br />
<br />generated peak stream power per unit area of about 12
<br />W/mz, would be representative of curve A_
<br />Curve B represents large floods that generate high
<br />values of peak stream power per unit area, and have
<br />moderate to long duration_ Average flood stream power per
<br />unit area is high, and total energy expended by the flood is
<br />large_ Peak stream power per unit area can be great enough
<br />to generate processes capable of eroding some bedrock
<br />boundaries, such as cavitation or macroturbulance [Baker
<br />and Costa, 1987; O'Connor, 1993]. Tremendous changes
<br />in alluvial channels are possible, even total unraveling of
<br />floodplains, because of the large energy expenditure
<br />represented by the area under the stream-power graph
<br />above the alluvial threshold. Area above the bedrock
<br />threshold represents the amount of energy available to
<br />erode and effectively modify bedrock flood-channel
<br />boundaries. Floods represented by curve B are likely to be
<br />the most geomorphically effective fluvial events in any
<br />landscape, and would include exceptional floods like the
<br />Rubicon River and Teton River dam-failure floods, and
<br />colossal paleofloods like the Missoula and Bonneville'
<br />floods [O'Connor and Baker, 1992; O'Connor, 1993].
<br />Stream-power graph C represents floods that generate
<br />high values of instantaneous peak stream power per unit
<br />area, but are short-lived. The energy represented by the
<br />area under the stream-power graph above the alluvial
<br />threshold is small, and these floods are impotent to
<br />accomplish any significant amount of geomorphic change,
<br />even though instantaneous peak stream power per unit area
<br />may be among the highest values documented and well
<br />above landscape resistance thresholds_ Total energy
<br />represented by the area under the curve above the bedrock
<br />threshold is also small, and the flood engenders little or no
<br />
<br />-
<br />
|