Laserfiche WebLink
<br />25. The next step wall therefore to comp:Lle a "factor complex map." <br />It will be recalled that each factor (or factor set) varied independently. <br />Thus, for example, a section of gap might be subdiv:Lded into two segments: <br />at one point on the basis of cone index values, and at quite another place <br />on the basis of bank heights. The problem, then, is to combine all of the <br />factor maps into as:Lnglemap on which is identified each gap segment that <br />displays a unique combination of factor values. In effect, this process <br />identifies what might be called "gap types," a gap type being defined as <br />a segment of gap exhibiting a specific combination of factor value classes <br />throughout its length. The compilation of the factor complex map (or the <br />gap type map) was achieved by successively overlaying the factor maps on a <br />suitable base map (see fig. 4), and accumulating the data. When all the <br />factor maps had been overlain in this fashion, the resulting factor complex <br />map included a delineation of every gap type found in the area. The com- <br />pleted factor complex maps (fig. 5) represent the completion of the acqui- <br />sition of all available descriptive data. This provides a concise data base <br />that allows estimates of stream environment conditions that vary in both <br />time and space for a sizeable area in the real world. <br /> <br />Cornmen ts <br /> <br />26. It should be specifically noted that the European Waterways Study <br />concluded with a procedure that is entirely descriptive in nature; it does <br />not predict conditions except in the almost trivial sense of "predicting" <br />what the current velocities should be in a given channel at a given water <br />level. The point is that the entire study was devoted to Step 2 (data <br />acquisition) of the WES terrain analysis concept. It has been used as a <br />demonstration of the fact that acquiring quantitative terrain data is not <br />a trivial problem, but it ~ be done. <br /> <br />27. One obvious conclusion which may be drawn from the foregoing dis- <br />cussion is that a deterministic hydrologic model that would predict stream <br />flow as a function of basic watershed parameters is very badly needed, since <br />it would allow the generation of information similar to that developed in <br />the European Waterways Study, at the same time circumventing the necessity <br />for the elaborate statistical "fitting" procedures used therein to obtain <br />hydro graph and related data. Hydrologists have sought an analytical synthe- <br />sis of the full hydrologic cycle during the last decade and several "models" <br />have been developed. Probably the most complete model is that developed by <br />Crawford and Linsley. 3 Other important models include the Kansas Model, <br />developed by Lumb,4 and a model developed by the Agricultural Research <br />Service (Holtan and Lopez).5 These models have all been used to reconstruct <br />hydrographs that match well with the measured flows. However, their use <br />requires considerable "juggling" of input parameters to obtain the desired <br />fit. Further, the input parameters are usually seldom explicit measurable <br />environmental factors. This seriously restricts the use of the model in <br />ungaged watersheds. Also, it negates the model for use in studies dealing <br />with the effects of various land modification and resource management con- <br />cepts. Holtan, in describing his model, states, "Our model is currently a <br /> <br />20 <br />