Professor of Civil and Environmental Engineering San Diego State University San Diego, California
1. INTRODUCTION A unit hydrograph (UH) is a hydrograph for a unit depth of effective rainfall, applicable to a given basin and for a given duration (Sherman, 1932). In practice, the unit hydrograph is convoluted with the effective storm hyetograph to obtain the composite flood hydrograph. For proper convolution, the UH duration should be the same as the unit interval of the storm hyetograph. If this is not the case, the S-hydrograph [see online version] can be used to change the UH duration to match the unit interval (Ponce, 1989). The shape of the unit hydrograph, as characterized by its peak and time lag, is a measure of the amount of runoff diffusion prevailing in the basin (Ponce, 1989). Steeper basins will have little runoff diffusion, while milder basins may have a significant amount. Diffusion acts to spread the flows in time and space. Less diffusion means sharp hydrographs with little attenuation; more diffusion means substantially attenuated hydrographs. Thus, less/more diffusion means a shorter/longer time lag.
2. LINEAR RESERVOIRS A linear reservoir provides a certain amount of runoff diffusion. The amount of diffusion is characterized by the Courant number C, defined as follows (Ponce, 1989):
in which Δt (hr) is the unit interval and K (hr) is the reservoir storage constant. In unit hydrograph theory, for proper convolution, the unit interval Δt (hr) has to match the UH duration t_{r} (hr). Thus:
For runoff diffusion to be positive in a linear reservoir (Ponce, 1989):
i.e.,
Larger values of K, compared to t_{r}, provide more runoff diffusion. Typical values of C are in the range 0.1 ≤ C ≤ 2.
3. CASCADE OF LINEAR RESERVOIRS The cascade of linear reservoirs (CLR) is a hydrologic model that uses a number of linear reservoirs in series in order to provide a wide range of possibilities for runoff diffusion and associated unit hydrograph peak flow attenuation and time lag effects. In a CLR, the outflow from the first reservoir is the inflow to the second; the outflow from the second is the inflow to the third; and so on (Ponce, 1989). Typical values of the number of reservoirs N are in the range 1 ≤ N ≤ 10. Once C and N are defined, the CLR method provides a unique amount of runoff diffusion, i.e., a unique unit hydrograph peak and associated time lag. An online version of the CLR UH is given in online_uh_cascade.
4. DIMENSIONLESS UNIT HYDROGRAPH The dimensionless unit hydrograph (DUH) has dimensionless time t_{*} in its abscissa and dimensionless discharge Q_{*} in its ordinate. The dimensionless time is:
in which t = time (hr). The dimensionless discharge is:
in which Q = discharge, in m^{3}/s; and Q_{max} = maximum discharge, ie., the discharge attained in the absence of runoff diffusion, in m^{3}/s. According to the runoff concentration principle (Ponce, 1989):
in which i = effective rainfall intensity, in m/s; and A = basin drainage area, in m^{2}. Then:
For 1 cm of rainfall (in SI units):
Thus:
in which t_{r} is in hr and A in in km^{2}. An online version of the CLR DUH is given in online_dimensionless_uh_cascade.
5. GENERAL DIMENSIONLESS UNIT HYDROGRAPH A general dimensionless unit hydrograph (GDUH) can be generated using the CLR method for a basin of drainage area A and unit hydrograph duration t_{r} (or storm hyetograph's unit interval) . The resulting set of Q_{*} vs t_{*} unit hydrograph values can be shown to be solely a function of C and N, and to be independent of either A or t_{r}. Thus, for a given set of C and N, there is a unique GDUH, of global applicability. In practice, a set of C and N are chosen such that the runoff diffusion properties of the basin are properly represented in the GDUH. This requires careful consideration and the analysis of measured rainfall-runoff events. Steeper basins will require a large C and a small N; conversely, milder basins will require a small C and a large N. The recommended practical range of parameters is the following: 0.1 ≤ C ≤ 2; 1 ≤ N ≤ 10. Within this range, C= 2 and N= 1 provides zero diffusion, while C= 0.1 and N= 10 provides a very significant amount of diffusion. The case of zero diffusion is equivalent to the assumption of runoff concentration only, which is inherent in the rational method (Ponce, 1989). Once the GDUH is chosen, the ordinates of the unit hydrograph can be calculated from Eq. 10:
and the abscissa from Eq. 5:
The unit hydrograph thus calculated can be convoluted with the effective storm hyetograph to determine the composite flood hydrograph (Ponce, 1989). An online version of the GDUH as a function of C and N is given in online_general_uh_cascade. An online version of a GDUH series as a function of C in the range 0.1 ≤ C ≤ 2.0 and N in the range 1 ≤ N ≤ 10 is given in online_series_uh_cascade. An online version of a GDUH series for C equal to 2, 1.5, 1.0, 0.5, 0.2, and 0.1, and N in the range 1 ≤ N ≤ 10 is given in online_all_series_uh_cascade.
6. PROCEDURE The procedure to derive a unit hydrograph based on the GDUH is the following:
Either the convolution or the CLR methods can be used to calculate the composite flood hydrograph.
The composite flood hydrographs calculated by both methods are shown to be the same.
7. ADVANTAGES / DISADVANTAGES The GDUH described herein has the following significant advantages:
A disadvantage is that while the parameters C and N can be readily related to the basin's general topographic relief, considerable experience and/or measured rainfall/runoff data is required for a proper estimation of these parameters.
8. CONCLUSIONS A general dimensionless unit hydrograph (GDUH) based on the cascade of linear reservoirs is formulated and calculated online. The GDUH is shown to be solely a function of the Courant number and the number of linear reservoirs. Since the GDUH is independent of the basin drainage area and the unit hydrograph duration, it is applicable on a global basis. Each GDUH is related only to the basin's prevailing runoff diffusion properties. The latter are a function of the general topographic relief, i.e., more relief, less diffusion, and less relief, more diffusion. The model's two-parameter feature provides increased flexibility for simulating a wide range of diffusion effects.
9. RESEARCH OUTLOOK The theory and procedures presented herein are consistent with established hydrologic practice. The two-parameter model provides increased flexibility for simulating a wide range of runoff diffusion effects, while remaining within a hydrologic routing framework. Research into the estimation of C and N on the basis of basin relief, geomorphology and ecology will enable a better prediction of composite flood hydrographs, for both design and forecasting applications.
REFERENCES
NOTATION The following symbols are used in this publication: A = basin drainage area (m^{2} or km^{2}); C = Courant number, dimensionless, Eq. 1 or Eq. 2; CN = (NRCS runoff) curve number; i = effective rainfall intensity (m/s); K = (linear) reservoir storage constant (hr); N = number of linear reservoirs in series; Q = unit hydrograph discharge (m^{3}/s); Q_{max} = maximum discharge, in the absence of runoff diffusion (m^{3}/s), Eq. 7; Q_{*} = dimensionless discharge, Eq. 6 or Eq. 8; t = time (hr); t_{r} = unit hydrograph duration (hr); t_{*} = dimensionless time, Eq. 5; and Δt = unit interval of the storm hyetograph (hr). |
190618 |