|
<br />318
<br />
<br />BULLETIN OF THE ASSOCIATION OF ENGINEERING GEOLOGISTS
<br />
<br />,.
<br />
<br />a 14
<br />Z
<br />a
<br />u
<br />& 12
<br />~
<br />w
<br />~
<br />I! 10
<br />t;:
<br />~
<br />u ,
<br />~
<br />~
<br />u
<br />~ 6
<br />..
<br />~ ,
<br />x
<br />u
<br />~
<br />" 2
<br />
<br />
<br />
<br />o
<br />
<br />31
<br />
<br />30
<br />
<br />AUGUST
<br />
<br />JULY
<br />
<br />Figure 13. Hydrograph showing the effect of a debris flow for
<br />Crystal River above Avalanche Creek near Redstone, Coio..
<br />July-August 1959.
<br />
<br />PROBLEMS IN ESTIMATING THE WATER
<br />DISCHARGE OF SEDIMENT-BEARING
<br />FLOWS
<br />
<br />The presence of recently eroded channels and
<br />coarse deposits may lead some investigators to con-
<br />clude that a waterflood has occurred, An indirect
<br />discharge estimate then is obtained using common
<br />techniques, such as slope-area methods. For ex-
<br />ample, the slope-area estimates of discharge along
<br />the East River tributary, assuming a water-flood
<br />flow for the event on July 24, 1977, are as much as
<br />112 to 286 m"/s (3,950-10,000 ft"/s) (Mears, 1979,
<br />table I), About 200 m (650 ft) farther downstream,
<br />at the toe of the debris fan, 100 m (325 ft) in front
<br />of the terminal lobe, our estimate made from high-
<br />water marks indicates the peak water flow here
<br />was less than 8 m"ls (280 ft"/s). It is unlikely that
<br />peak attenuation of such magnitude would have
<br />taken place in such a short distance,
<br />Downstream of the tributary junction, the flow of
<br />the East River was approximately 6 m'ls (200 ft3/s)
<br />(well below bankfull) on the day the area was vis-
<br />ited (June 21, 1978), A discharge in excess of this
<br />amount would have left high-water marks in the
<br />channel or on the floodplain below the debris fan,
<br />Yet no high-water marks for a stage greater than 6
<br />m"ls (200 ft3/s) were observed on the East River
<br />immediately below the debris fan, Therefore, the
<br />water runoff from this tributary was most likely less
<br />than 8 m'ls (280 ft"/s) , much less than originally re-
<br />ported (Mears, 1979).
<br />
<br />When such conventional slope-area methods that
<br />use Manning's and Bernoulli's equations are mis-
<br />takenly applied to debris flows, excessive water-
<br />flood flow estimates usually result, Several impor-
<br />tant assumptions that are invalid when applying the
<br />slope-area method to debris flows are:
<br />
<br />I. That the water-flood passage is uniform with
<br />steady flow or gradually varied flow,
<br />2. That the vertical-velocity distribution is loga-
<br />rithmic.
<br />3, That the high-water marks (defined by seed
<br />lines, drift, and water lines) remaining after the
<br />passage of the flood reflect the height of the
<br />water flow. For a waterflood, the marks should
<br />not indicate any ponding (backwater) resulting
<br />from a damming effect common at the front of
<br />debris flows,
<br />4. That channel erosion should be minimal.
<br />5. That Mannings' roughness coefficient "n,"
<br />which has been determined for water-flow dis-
<br />charges, should be used in the computations
<br />(Benson and Dalrymple, ]967; Da]rymple and
<br />Benson, 1967),
<br />
<br />There are hydraulic relations available for use
<br />with debris flows, Gol'din and Lyubashevskiy
<br />(1966) discuss some formulas proposed by Russian
<br />hydrologists to compute the average velocity of
<br />mudflows in Crimean river channels.
<br />Waterfloods and debris flows are end members of
<br />a continuum of a wide range of sediment-water mix-
<br />tures in channels. Hooke (1967) believes the differ-
<br />ence between waterfloods and debris flows is that
<br />(a) the sediment load in waterfloods is varied by
<br />deposition or erosion, whereas only the coarsest
<br />fragments are deposited by debris flows, and (b)
<br />streams (waterfloods) continue to flow so long as a
<br />slope exists, but debris flows stop on a finite slope.
<br />Thus, a debris flow cannot be converted into a wa-
<br />terflood by deposition. A problem exists where wa-
<br />terfloods transport large volumes of sediment, or
<br />where a debris flow is followed by a more fluid wa-
<br />terflood, disrupting the evidence of the debris flow,
<br />The problem of making slope-area discharge esti-
<br />mates for small mountain streams transporting a
<br />large but unknown percentage of sediment is illus-
<br />trated in Figure 14 which is a plot of drainage area
<br />versus discharge. The curve is an estimate of the
<br />500-year discharge for the Rocky Mountains in Col-
<br />orado from McCain and Jarrett (1976). Debris-flow
<br />slope-area discharge estimates (circles in Figure 14)
<br />are discussed in the following section, The triangles
<br />
<br />"
<br />
|