|
<br /> ,-
<br />" ,
<br /> TABLE S. Summary of Effect of One-Factor, Two-factor, and Three-Factor Variation on PMF - -
<br /> Numbers of " HMR52 opti, Percent
<br /> factors varied mum storm.area Volume of change in
<br /> from baSe Storm-center Storm T empora\ rainfaU Isohyetal pattern size rain1all PMF PMF trom
<br />point location orientation distribution axes ratio (SQ mil (in,) (cts) base point
<br />(1) (2) (3) (4) , (5) (6) (7) (8) \...- (9)
<br /> (a) Results [or Seneca Oeek.
<br /> o (base point) (0,0) ~ by (29,1. 181' ~ by 234' center-Peaked 2.5 to 1 100 26,63 114.227,69 0
<br /> 5,7)
<br />I (0. - 3,83) $ by 181' ~ by 227' center-peaked 2,5 to 1 175 25.41 108.520,89 -4,99
<br /> (27,56.4,0) .
<br />2 (0. - 3,83) ~ by 181' ~ by 227', ea"rIy-peaked 2,5 to 1 175 25.41 97.662.58 -14,50
<br /> . (27,56,4,0)
<br />3 (0. - 3,83) ~ by 270' ~ by 317' early-peake~ 2.5 to 1 450 21.88 87,354,51 -23.53
<br /> , (27,56.4,0)
<br /> (b) Results for Hypothetical Drainage Basin
<br /> o (base point) (0,0) 1810 center-peaked 2.5 to 1 50 23,19 136.085 to 0
<br /> 173.742
<br />I (0. - 3,83) 181" center-peaked. 2.5 to 1 100 21.61 124,885 to -8,23
<br /> 159.953
<br />2 (0,,-3.83) 181" . . cady-peaked 2.5 to 1 100 21.61 111,310 to -19.25
<br /> . 138.111
<br />3 (0.--:3,83) ,270' I ' early-peaked 2.5 to 1 300 17,36 88.162 to -36,01
<br /> , ' 110,291
<br />
<br />Note: To convert sq rrii to km2. multiply by 2.59; to convert inches to mm, mulitply'by 25.4; and to convert efs to m3/s, multiply by 0.02832.
<br />,.. ~.
<br />
<br />in PMF, from the base point PMF, was conducted betwee~
<br />the hypothetical drainage basin and Seneca Creek, The per-
<br />cent change from the base, pointPMF resfllts from a one-,
<br />twO:", and. three-factor variation. Selection ,pf, the;, ,meteoro.,.
<br />logical fact()rs to vary is discussed in Shalaby (1986), Table
<br />5, part a. 'gives the values for th~"optimum storm,-area ~ize"
<br />'the volume of rainfall, and the PMF resulting from a one-
<br />factor. two~factor. and three-factor variation foi' Seneca Creek.
<br />Table 5, part b, gives the corresponding values for the hy-
<br />Pothetical 60 sq mi (155.5 km') drainage basin assumed in
<br />this research. A subjective comparison of the res~lt~',indic;ates
<br />Ihat the percent changes from 'the base point PMF, for the
<br />two drainage basins, for a one-factor, two.:.factor. and, three-
<br />factl?r variation are. not very different.. , Therefore, th,e, ad-
<br />justment factorS for anyone-factor, two-factor, or three-fac-,
<br />tor variation may be assumed valid for drainage basins similar
<br />to the ones studied herein, Furthermore, a comparison of the
<br />values between parts a and b of Table 5 reveals that the
<br />smaller the drainage area [i.e" the 60 sq mi (155.5 km')
<br />hypothelical drainage basin as compared to the 129 sq mi
<br />(334.4 km') Seneca Creek drainage basin], the more sensitive
<br />Ihe PMF is to the coniributing factors. But, however, it would
<br />be useful to conduct a more extensive verification of the ad-
<br />justment factors using both smaller and larger drainage ba-
<br />sins.
<br />
<br />CONCLUSIONS
<br />
<br />The major goal of this research is to ,obtain an accurate
<br />estimate of the PMF for high hazard hydrologic design prob-
<br />lems. A set of guidelines was developed, herein to assesS the
<br />effects of each factor on the PMF. Use of the guidelines
<br />provides greater assurance that the optimum PMF will be
<br />found and reduce the effort to find the optimum PMF, which
<br />is required for design_
<br />The sensitivity analysis conducted yielded important in-.-
<br />fonnation concerning the interactions between the factors
<br />that are needed to estimate a PMF. A significant interaction
<br />exists between the temporal rainfall distribution and both the
<br />storm-center location and the land-cover distribution. When
<br />variables have strong interdependence, it is }nadequate to try
<br />to estimate the design variable by varying each variable sep-
<br />
<br />arate1y, Awareness:of'the engineer of the importance of the
<br />interaction effects will provide greater assurance that the op-
<br />timum PMF will be found.
<br />Each of the factors contributing to the PMF may be as-
<br />sess"ed in teriiis of its contribution'to the volume, spatial dis-
<br />tribulion, and iemporal distribution of both the rainfall and
<br />the runoff and, thus, the resulting PMF, Understanding the
<br />na'tunb of the contribution of each factor in' estimating the
<br />PMF allows the designer to identify which factors must be
<br />carefully chosen, Relatively high accuracy in the values of
<br />insensitive inputs is not important to the overall accuracy of
<br />the computed PMF,: "
<br />, Jbe guidelines presented in this research allow the design
<br />enginee:r aud(or policymaker to obtain either a planning es-
<br />timale or a design estimate of the PMF, A planning estimate
<br />of the PMF is obtained by using a single HMR521HEC. I
<br />computer run and the adjustment factors based on the re-
<br />search results just presenled. When a design estimate of the
<br />PMF'is required, the designer must conduct a site-specific
<br />sensiIivity analysis, which can be guided by the research re-
<br />sults, presented herein. Obtaining a design value of the PMF
<br />would be guided by the infonnalion concerning the interac"
<br />tion between factors and the effects of each faClor on the
<br />PMF, in Qrder to identify those variables whose values must
<br />be varied and/or carefully selected in the PMF estimating
<br />procedure.
<br />A preliminary verification of the adjustment factors was
<br />conducted ,using Seneca Creek as a case study. A comparison
<br />of the resulIs between the hypothetical drainage basin of this
<br />research" and Seneca Creek for a one-. two-, and three-factor
<br />var~ation was conducted. A subjective comparison of the per-
<br />cent change in PMF from Ihe base poinl PMF between the
<br />hypothetical drainage basin and Seneca Creek, revealed Ihat
<br />the adjustment factors may be assumed val\d for recommen-
<br />dation for drainage basins similar to the ones used in this
<br />study. However, further study is needed for a validation in
<br />recommending the adjustment factors on a more general ba.
<br />sis.
<br />
<br />APPENDIX. REFERENCES
<br />CH2M Hin. (1983), "Seneca phase-II watershed study," Rep_ prepared
<br />for Maryland Nat, Capital Park and PIng. Commission, Md_
<br />
<br />3361 JOURNAL OF IRRIGATION AND DRAINAGE ENGINEERING 1 SEPTEM8ERIOCTOBER 1995
<br />
|