|
<br />26
<br />
<br />RIVER Ml!ANDERlNO
<br />
<br />These are only landscape details, but careful evaluation of an
<br />adjusting stream is required because aggradation may follow the
<br />expected degradation, as a natural result of increased sediment move~
<br />ment (33), and episodic behavior (1) may be characteristic of high-
<br />ener9Y fluvial systems (Fig. 5b).
<br />
<br />SUMMARY
<br />
<br />The purpose of this paper has been to discuss some of the problems
<br />associated with prediction and to demonstrate why prediction in geo-
<br />morphology is of low resolution. Just as the engineer accommodates his
<br />uncertainty by the use of factors of safety so should the earth
<br />scientist use worst-case limits, as a basis for long-term prediction.
<br />
<br />This discussion of the seven problems is not an attempt to
<br />discourage prediction. Rather, it can be construed as encouragement.
<br />If problems are recognized then they can be solved. It is, however, an
<br />appeal for careful preliminary investigations at any s1te prior to the
<br />initiation of action of any type. Frequently critical information on
<br />past channel behavior and the present longitudinal channel profile
<br />variations are ignored. Perhaps most serious Is the tendency to Ignore
<br />the fact that a meander or a river reach is a part of a fluvial system,
<br />
<br />"
<br />
<br />
<br />...
<br />,..
<br />
<br />IIII'.O..C.._
<br />1.01--'-'0,1'.
<br />
<br />"
<br />
<br />l'
<br />t"
<br />
<br />"
<br />
<br />
<br />"
<br />
<br />1111..............,...
<br />...,_...",11,-
<br />
<br />,-
<br />
<br />..
<br />
<br />I'"
<br />
<br />t'-
<br />~"""
<br />t'-
<br />'f"'4
<br />
<br />Fig. 6.
<br />
<br />Terraces formed along Greenhorn Creek, a tributary to Bear
<br />River, Sierra Nevada, CA., as it incised 21 m. into
<br />gravels produced by hydraulic mining. Profile A was
<br />surveyed during the surrmer of 1979, and spring floods of
<br />1960 destroyed the lowest three terraces. Profile B is
<br />located About 1 km downstreom of Profile A (from Wildman,
<br />1961).
<br />
<br />~
<br />
<br />RIVER MORPHOLOGY AND BEHAVIOR
<br />
<br />27
<br />
<br />Therefore system behavior must also be anticipated. The purpose is to
<br />develop a better general understanding of a specific site in order to
<br />reduce the error inherent 1 n extrapo la t 1 on:
<br />
<br />The assumption is that the uncertainty can largely be removed by
<br />careful study. Although Simply being aware of the problems is of
<br />value, nevertheless 1 will attempt to suggest ways of minimizing their
<br />lmpact. For the problems of scale, and location, perhaps the most
<br />important consideration is that rivers have a history. They change
<br />through time and respond to climatic fluctuations and tectonic
<br />influences. Therefore any historical information on prior behavior is
<br />of value (Fig. 1), For very longterm prediction involving thousands
<br />of years, the Pleistocene and Holocene history must be understood
<br />because that history probably will be repeated. This history is
<br />preserved in the stratigraphic record and can be unlocked by careful
<br />sedimentologic investigations.
<br />
<br />For size. location. singularity. and sensitiVity an understanding
<br />of the river reach 1n the context of other rivers is ne!ded. VarfatiOM
<br />along the channel and the differences among channels in the same area
<br />must be rec09nized and explained, Hence, the system and its com-
<br />ponents must be understood and anomalies explained. For example,
<br />channel pattern differences must be understood in relation to the
<br />remainder of the drainage network (scale, location and Singularity)
<br />and the possible presence of thresholds (sensitivity),
<br />
<br />Convergence and divergence require that cause and effect relations
<br />and the processes and factors influencing the river be known. Com-
<br />plexity requires that the response of the river must be anticipated.
<br />The question must be asked, will the response of the channel be
<br />strai9htforward and prOgressive or will it be episodic as sediment
<br />loads fluctuate (Fig. 5)1
<br />
<br />General relations developed elsewhere should be applied to a
<br />specific river w;th caution. Therefore general principles and concepts
<br />developed from f;eld and exper;mental studies elsewhere should be
<br />brought to bear. but quantitative relations must be developed
<br />spec;f;cally for each river reach of concern.
<br />
<br />The solution proposed is a geomorphic approach using the method
<br />of multiple explanations (5, 15). As many explanations as can be
<br />developed for the phenomena at a site should be generated. Many of
<br />these chn be eliminated, but eventually the remaining hypothesis will
<br />provide an explanation of reach characteristics and behavior.
<br />Obviously this requires additional data collection, field work,
<br />surveying, sampling. and analyses. Consideration of the problems may,
<br />therefore, be time-consuming. difficult and expensiv!, but never as
<br />expensive as project failure. The development of the required under-
<br />stand,ng of river morphology and behavior will be intellectually
<br />demanding and rewarding. and it will be cost effective.
<br />Ackno~ledgments: This paper was prepared with support from both
<br />National Science Foundation and U, S, A~ Research Office. I thank
<br />H. O. Harvey, C. C, Watson, Harold 60yne 'Harold Halde and A N
<br />Strahler for caref~l rev1ews~ and ~ g~aJuate students A. eam bell
<br />R. Schumann. A. Pa,ne, J, Pltlick and S. Ouch! for useful cri~icism.
<br />
|