• I 

 I I 



ti-l ti 



Figure 1. — Parameters used to locate points 

 along the base flow line. 



Each successive point on the hydrograph is 

 the basis for a new calculation of P/. The 

 event ends w^hen the value of a point on the 

 base flow line equals a rate value on the 

 hydrograph trace; i.e., Pj=i?j. The next 

 hydrograph event begins when the hydro- 

 graph steirts to rise again. At that point, 



Interflow Computations 



The curve that separates interflow from 

 surface flow (which includes channel intercep- 

 tion) is similarly computed from point to 

 point. Any point Sj on the interflow curve is 

 equal to the value at a prior time fS/.^ ) plus 

 an incremental vEilue (Wi). 



Thus, Si=Si.^ + Wi (3) 



where 



Wi = [iAt^ +ASM''^'xAS]/(Af + KxAS) (4) 

 where 



M = ti - ti.^ 

 AS = i?j -Sj.j 

 K = 3.0 



Surface runoff ends when the curve inter- 

 sects the hydrograph; i.e., = R[. 



DISCUSSION OF TECHNIQUE 



Table 1 shows the results of analyses of 

 hydrographs of representative storms in sever- 

 al regions of the United States. The analyzed 

 hydrographs include both simple zind complex 

 hydrographs. Source for the basic hydrograph 

 data is Hobbs and Burford.^ In all cases, the 

 analyses were performed on data expressed in 

 cubic feet per second (c.f.s.). Data originally 

 expressed in inches per hour were converted 

 to c.f.s. Figures 2 and 3 show the separation, 

 respectively, of a simple and complex hydro- 

 graph into its components. The figures show 

 the total flow hydrograph, and the computed 

 interflow and base flow rates. 



Normally, on large watersheds, the time in- 

 terval between sampled points on the hydro- 

 graphs is one or more hours. Therefore, the 

 analytical results will be independent of the 



^Harold W. Hobbs and James B. Burford. Hydro- 

 logic data for experimental agricultural watersheds in 

 the United States, 1963. USDA Agr. Res. Serv., Misc. 

 Pub. 1164. 1970. 



hydrologist. On very small watersheds, the 

 hydrologist tries to reproduce the shape of 

 the hydrograph by sampling at small and vari- 

 able increments of time because the smaller 

 the time interval used the more accurately 

 can he reproduce the hydrograph 's originad 

 shape. Thus, on very small watersheds the 

 hydrologist can influence the results by in- 

 creasing or decreasing the number of sampling 

 points used to reproduce the shape of the 

 hydrograph. 



Our analyses show that when we reduced 

 the number of sampling points 75 percent, we 

 changed the computed results by significant 

 amounts. The maximum change was observed 

 on a small watershed (0.07 km.^ ) from which 

 all runoff ceased after 1.75 hours: computed 

 total flow was increased 0.457 cm. (28 per- 

 cent); stormflow increased 0.508 cm. (28 per- 

 cent); and stormflow runoff duration was in- 

 creased 6.6 minutes. Consequently, the 

 hydrologist who analyzes the hydrographs of 



3 



