|
<br />
<br />factors were assumed constant throughout the sensitivity anal-
<br />ysis, other factors were varied (see Table 1). The variation
<br />in the value of the factors that were studied was assumed to
<br />be due to a change in existing conditions, such as for the land-
<br />cover distribution, or a change in the historical observations
<br />such as for the isohyetal pattern, the location of the stonn
<br />center, the stonn orientation, and the temporal rainfall dis-
<br />tribution. A variation in landcover distribution reflects the
<br />potential for urbanization, Although the HMR52 points out,,,
<br />that the mo.ximum precipitation results from drainage cen~
<br />tering the isohyetal pattern (Le., the isohyetal and drainage
<br />patterns ,have approximately the same center and axial ori-
<br />entation), chree values for L, and two (extreme) values for
<br />4> were examined in the sensitivity analysis. ,HMR52 recom-
<br />mends an idealized elliptical isohyetal pattern with a ratio of
<br />the major to minor axis of 2.5 to 1 for distribution of the six
<br />1 ~hr increments of precipitation over the drainage area. In
<br />addition to a PI of 2.5 to 1, ratios of 4,0 to 1 and 1.0 to 1
<br />were examined in the sensitivity analysis. 'IJle optimum ,stonn-
<br />area siie is selected by the HMR52 program. HMR52 indi- '
<br />cates that certain temporal rainfall distributions result in more
<br />critical flow for the PMF than other temporal distributions,
<br />But the detennination is left to the analyst. Three temporal
<br />distributions were examined in t~e sensitivity an;1lysis. As for
<br />the remaining factors (i.e" those held constant in value) the
<br />reasons and justifications for their assumed value are. ex-
<br />plained in the following.
<br />
<br />Size of Drainage Area
<br />
<br />Although PMP values are availabie for drainage areas rang-
<br />ing from 10 to 200,000 sq mi (25.9 to 518; 420km2), a drainage
<br />area of 60 sq mi (155,5 km2) was assumed for the purpose of
<br />the sensitivity analysis, A 60 sq mi (155.5 km2) drainage basin
<br />is small enough to represent nonlinear basin behavior so as
<br />tooptimiie on tbe sensitivity of the PMI' to its contributing
<br />factors. It is also small enough to provide ease in the hye
<br />drologic computations, yet large enough for the sensitivity
<br />analysis results to ,be applicable in many practical design sit-
<br />uationse' Furthermore, the selected size of the drainage area
<br />pennits Incorporating in the sensitivity analyses the effects of
<br />variation in the spatiallandcover distribution as well as sub-
<br />dividing the watershed and flood routing.
<br />
<br />Shape of Drainage Area and Drainage Subdivision'
<br />
<br />Although watersheds usually have an irregnlar shape; the
<br />selection of a regular watershed with well-defined geometric
<br />properties greatly reduces the computational eff~rts required
<br />in the analysis. Thus an ellipse was select."J for the analysis
<br />due to its flexibility to represent a circle as well'as an elongated
<br />ellipse. Furthennore, the choice of an ellipse allows the use
<br />of its axes and foci to be used as coordinate references. To
<br />select a realistic shape ratio (the ratio between the major and,
<br />minor axes), a set oflO basins in the Washington, D.C.,area
<br />were analyzed (see Shalaby (1986).Appendix 4-A)]. The av-
<br />erage shape ratio was found to be 2.4. Because the difference
<br />between the recommended elliptical isohyetal shape ratio of
<br />2.5 and the aforementioned empirically determined elliptical
<br />drainage basin shape ratio of 2.4 is not significant, the rec-
<br />ommended shape ratio of2.5 was assumed as the basic drain-
<br />age 'shape, parameter. In the case of the sensitivity analyses
<br />herein a delineation of the subbasin boundaries' was neces~
<br />sary t; detennine the affects both of the location of the stonn
<br />center and the spatial variation in the landcover distrib:ut.ion.
<br />The delineation of the subbasin boundaries was guided by
<br />the spatial variations intended for the location of the stonn
<br />center and the land-cover distribution (see Table 1 for 05-
<br />, sumed values for L, and Lv), Fig, 1 "epicts the drainage
<br />
<br />
<br />Unbranched
<br />Prindpal
<br />Stream.
<br />
<br />
<br />
<br />Subbasin num.~er
<br />
<br />Drainage
<br />Basin
<br />Divide
<br />
<br />) )
<br />
<br />
<br />4
<br />
<br />-....-.
<br />
<br />I s
<br />
<br />
<br />Outlet
<br />
<br />AG. 1. ,A$sumedDralnage Subdlvlslon'and Stream Network
<br />
<br />. subdivision and stream, networx asSumed for the sensitivity
<br />analyses. The60sq mi (155-5 km2)elliptical drainage basin
<br />was subdivided into six 10 sq mi (25,9 km,2} subbasins [see
<br />Shalaby (1986)Appendix 4-B fordetails}. '
<br />
<br />, ,
<br />Antecedent Soil Moisture and Infiltration Rate
<br />
<br />The U,S:Anny Corps of Engineers and thiU.S. Bureau
<br />of Reclamation (USBR) h'ave recommended,that minimum
<br />constant infiltration rates be used in the PMF estimation pro-
<br />cedure. The assumption of a- constant, minimu.m rate of in-
<br />filtration 8fter saturation is modeled with the <I>-index method.
<br />Based on the infiltrometer stridies conducted by "he USBR
<br />(1973) a value for the 4>-index ohero in,lhr (0 mmlhr) was
<br />assumed for the sensitivity analysis for fully' developed land
<br />cover, while a value of 1 in.lhr (25.4 mmlhr) was assumed
<br />for undev!'loped hindeover, It should he noted that because
<br />the. antecedent soil moisture is assumed to be constant and
<br />saturated, it is not modeled directly but rather through the
<br />value of the 4>.index,
<br />
<br />Shape of Unit Hydrograph
<br />
<br />The two-parameter gamma UH was used in order to main-
<br />tain maximum flexibility with regard to the shape and scale
<br />of the VH, Furthennore in order to provide the maximum
<br />peak rate factor to model the PMF, a value approaching 0,5
<br />was assumed for p, the percent in the rising side of the UH,
<br />
<br />JOURNAL OF IRRIGATION AND DRAINAGE ENGINEERING / SEPTEMBER/OCTOBER 1995/329
<br />
|