0.16 



0.14 



"g 0.12 



o 



0.10 



< 0.08 



o 



0.06 



0.04 - 



0.02 - 



X TOTAL FLOW 

 O INTERFLOW 

 • BASE FLOW 



T 



16.0 



18.0 



24.0 



20.0 22.0 

 HOURS 



Figure 3.— Flow components of a complex hydrograph recorded at Hastings, Nebraska. 



very small watersheds should make every ef- 

 fort to select points that reproduce as ac- 

 curately as possible the true shape of the 

 hydrograph. 



Equation (2) includes two angle measure- 

 ments (A and B). Accordingly, it is apparent 

 that the results of any analysis will depend on 

 the units of measurement, as shown in table 2. 



Normally, the analysis would be performed 

 on data expressed in a unit suitable for the 

 watershed: c.f.s. for low water -yielding water- 

 sheds, and inch/hour for high water- 

 yielding drainages. Certainly, a great deal of 

 accuracy is lost when calculations are per- 

 formed on angles measured in minute frac- 

 tions of a radian. For example: suppose 

 streamflow from a 580-acre (2.346-km.^) 

 watershed is expressed in inches. If the in- 

 crease is 0.0005 inch in 0.1 hour, angle A is 

 approximately 0.005 radian; if expressed in 

 c.f.s., the angle is 1.24 radians. The effect of 

 loss in accuracy as well as the unit of measure- 

 ment is fully reflected in the last column of 

 table 2, which shows figures for surface run- 

 off based on data expressed in inch/hour. 



Equation (4) is used to partition stormflow 

 into surface runoff (including channel inter- 

 ception) and interflow. The equation contains 



a constant (K) to which we assigned a value 

 of 3.0. The exact value of K can be deter- 

 mined only if precise data from controlled 

 laboratory or field experiments are available, 

 but we do not have such data. 



A change in the value of K results in sig- 

 nificant changes in the analytical results (table 

 3). Increasing the value of K yields (a) larger 

 computed values for surface runoff; (b) a cor- 

 responding decrease in interflow; and (c) a 

 larger computed value for duration of surface 

 runoff. Our analyses for large forested water- 

 sheds show that the computed period for sur- 

 face runoff duration is extended by 1-3 hours 

 if the value of K is increased from 3.0 to 4.0. 



We believe that the value of K is related to 

 the proportion of the watershed represented 

 by the main channel and its tributaries. For 

 any particular watershed, the value of K 

 might well change wdth either or both the sea- 

 son and the characteristics of the storm. This 

 concept of K agrees with the concept of a 

 variable source area.'* 



'*John D. Hewlett and Alden R. Hibbert. Factors 

 affecting the response of small watersheds to precipi- 

 tation in humid areas. P. 275-290, in: Forest Hy- 

 drology, edited by William E. Sopper and Howard W. 

 Lull. New York: Pergamon Press, Inc. 1967. 



5 



