WSP05833 - Laserfiche WebLink

Home Browse Search

.- RIVER MORPHOLOGY AND BEHAVIOR: PROBLEMS OF EXTRAPOLATION S. A, Schumm M.ASCEI Abstract Rivers are complex landfonms; therefore, e~planation of their behavior and especially the prediction (extrapolation) of future behavior involves considerable risk. The risk can be reduced and the accuracy of prediction improved if consideration is given to seven prOblems associated with the prediction of river response and behavior. The seven problems are: scale, location, convergence. divergence. singu)arity, sensit1vity and comp1e~ity. Consideration of these problems requires that more attention be paid to the river reach of concern from the geomorphic perspective. Therefore, information on the history of the reach. a complete morpho- logic description of the reach and an understanding of the fluvial system of which it is a component, is necessary. Introduction There are three aspects of the investigation of river morphology and behavior. These are: l} description. the description and explan~ tlon of river characteristics; 2) prediction, the extrapolation of present conditions to the future. and 3) postdiction, the extrapola- tion of present conditions to the past. The engineer is concerned primarily with explanation and prediction. The geomorphologist has been primarily concerned with explanation and postdict1on. although recently with increasing attention to environmental concerns. both short- and long-tenm prediction of river behavior has become an additional geomorphic 90al, Indeed, extrapolation for up to 10,000 years is required for evaluation of some hazardous-waste disposal site~ Geomorphologists are frequently chided for their unwillingness to pred'ct with assurance. lhis reluctance 1S understandable when con- sideration 1s g1ven to the complex1ty of the fluv1al system and to the problems of extrapolation for long periods of time. The purpose of this discussion is to consider some of these problems and to suggest ways that their effects can be minimized when there is a need to predict river behavior. The geomorphologist uses three methods to postdict and to predict river change and response. and each method can be related to different spans of time. As the concern here 1s with the future. only prediction will be considered. For the short term (1-10 years) old maps and c-J, .--l r- .-4 Professor of GeOlogy. Department of Earth Resources. Colorado State University. Fort Collins, CO 80523. 16 , RIVER MORPHOLOOY ANI> BEHAVIOR 17 aerial photographs may be available to provide a goad historical record that can be used to predict for decades. For intermediate sp'ans.of time (1-100 years) empirical relations can be used. These are generally statistical correlations between channel morphology and some independent var1abl~ such as dis~haTg~. The equation that describes the relation can be used to predlct rlver response to altered conditions for decades and centuries. For very long periOds of time (1-1000 years) all geologic and geomorphic infonmation must be brought to bear. and the evolutionary development of landforms must be considered. Frequently this can be accomplished by using ergod1c\ty or a space-for~time substitut\oo. Ergodicity assumes that a long record at one location has the same statistical characteristics as a record obta'\n~d: from a number of locations at one time (11), For example, the different morphol09ic characteristics of an incised channel (channelized stream. gully) from head to mouth can reveal the evolutionary changes that will take place at one location throu9h time (25). All three methods can be used to predict for millen1a, but obviously confidence in the prediction decre.ses with incre.sing time span. The three methods provide valuable information, but the bother- some question is always. how far can modern data ar trends be extra~ polated in space and time. when the engineer and geomorphologist are required to make predictions concerning specific locations within a complex physical system? The answer depends largely on t~e under~t~nd- ing of the system, its s1mp11city, and the uniformity of ,ts cond,t,ons, but there are at least seven problems that should be considered when extrapolation of fluvial and indeed. any geomorphic relations are attempted. Problems of Extrapolation The seven prOblems produce an insufficient understanding of present and past conditions. They are as follows: 1) scale, the dimensions of both time and space, 2) location. the relation of a site to its surroundings, 3} convergence, the production of similar results from different causeS or processes. 4) divergence, the production af different results from similar causes or processes. 5) singularity. the distinctive differences and responses of similar land~ forms, 6) sens1t1vity, the susceptibility of landforms to small external changes, and 7) complexity, the complex and episodic behavior of a geomorphic system to altered conditions. In order to explain and to extrapolate. considerat1on must be given to each of these prOblems. Although 1t is true that most situations will not involve all problems, e.ch should be considered a possibility in every tase, .Scale: Scale prOblems involve both time and space (13, 30), and the shorter the t1me span. the smaller the space. and the moTe rapid and un1form the process the more spec1fic can be explanation and extra- polation. l'