ISSN (Print) : 0974-6846

Indian Journal of Science and Technology, Vol 7(12), 2096–2106, December 2014 ISSN (Online) : 0974-5645

Experimental Invesitigation and Modeling

of Rainfall Runoff Process
Ankit Chakravarti1* and M.K. Jain2
1
Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee–247667,
India; ankitchakravarti@gmail.com
Department of Hydrology, Indian Institute of Technology Roorkee, Roorkee–247667, India; jain.mkj@gmail.com
2

Abstract
Rainfall-runoff process of a watershed is important for water resources planning, flood forecasting, design flood estima-
tion, movement of pollutants transport through rainfall runoff process, and many other applications. For a mathematical
simulation of this phenomenon, controlled set of experiments were conducted to investigate this process through experi-
mentation on the catchment system placed over a rainfall simulator to obtain runoff hydrograph data. The experiments
were carried out over a non-cohesive sediment layer having sediment size of 0.5 mm to 1 mm placed over an impermeable plane
surface (smooth metal sheet), with a uniform rectangular cross section of dimension one meter wide and two meter long. Total
12 laboratory experiments were conducted to know behavior of this phenomenon using rainfall simulator. The generated
experimental data were simulated using a one-dimensional finite difference numerical model of kinematic wave equation
for overland flow to investigate the effects of variation in rainfall intensity and surface slope on the overland hydrograph.
The data was collected for catchment slope between 1% to 4 % and rainfall intensity varies from 30 to 90 mm/hr. The
comparison of observed and simulated runoff hydrograph reveals that the kinematic wave model simulates the rising,
equilibrium discharge and upper part of recession limb reasonably well in observed hydrograph with Nash–Sutcliffe effi-
ciency greater than 95%. However, in all the experiments, the lower portion of recession limb of the observed hydrographs
remained under predicted because of prolonged flow of water. The reason behind this variation may be attributed to the
retention of runoff of the recession limb because of surface water tension prevailing over the overland flow bed, channel
bed and even over the weir model used for flow measurement. The study further reveals that the resistance due to flow
decrease linearly with increase in catchment slope for a given rainfall intensity. It was also found that for a given rainfall
intensity, an increase in the catchment slope reduces the time to peak.

Keywords: Kinematic Wave, Laboratory Experiment, Rainfall-Runoff Modelling, Surface Roughness Simulation

1. Introduction determining flow depths and velocities and, hence, the

capacity of the flow to entrain and transport sediment
Rainfall- runoff processes plays an important role in and chemicals (Tayfur et al. 1993). Overland flow can be
hydrological cycle. The movement of water has mainly defined as a thin sheet of flow occurring before irregu-
arisen from the need to evaluate the amount of available larities cause a gathering of the runoff into discrete stream
runoff water at a particular location to meet local demand channels (Motha and Wigham 1995). It is generated as a
as well as risk of flooding due to excess water. Rainfall- result of rainfall in excess of the saturated hydraulic con-
runoff estimation from a watershed is important for safe ductivity of the soil or by saturation of the soil surfaces
and economical design of hydrological projects for the Woolhiser14.
purpose of water resources planning, watershed manage- A number of models such as physically based and
ment, water resources economics, flood forecasting etc. conceptual models have been used to simulate the rain-
The hydraulics of surface runoff are very important in fall runoff process. However, due to its complexity and

*Author for correspondence

 K. Shivalkar Yadav and R. Prabha

s­ patio-temporal variation, few models can ­accurately ary conditions at soil surface. The model was tested by
­simulate this highly non-linear process. There are ­comparisons to data from laboratory experiment and a
number of climatic and physiographic factors that sig- field plot. The study found good agreement between mea-
nificantly affect the transformation of rainfall into runoff sured and predicted hydrographs, but differences in the
in a catchment. The rain falling on the catchment under- recession limb were noted. The influence of slope was not
goes number of transformations under the influence of examined in this study. de Lima4 conducted laboratory
these factors before it emerges as runoff at the catchment experiments to emphasize the importance of spatial rain-
­outlet. fall intensity patterns of moving rainstorms on overland
Rainfall simulation has become a very effective tech- flow. A simple numerical model, based on the non-linear
nique for modeling overland flows, soil erosion and kinematic wave, was used for comparing the results for
rainfall runoff1,3,16,17. Rainfall simulation makes it pos- hypothetical storms moving up and down an impervious
sible to control the spatial and temporal characteristics plane surface. de Lima5 conducted laboratory experi-
of precipitation. It employs a watershed experimentation ments on an impermeable smooth plane surface with a
system as a tool for research of the behavior of the sys- movable sprinkling-type rainfall simulator to simulate
tem under varying conditions. The benefits of using the a moving storm. The results indicate considerable dif-
rainfall simulation approach are well documented, with ferences in runoff volumes and peaks and in overland
versatility being the foremost advantage16 (Meyer 1965). flow hydrograph shapes, for storms moving upstream
It gives ability to repeat experiments for predefined set of and downstream at differing velocities. Bronstert and
conditions. In addition, developing numerical models Bardossy (2003) reported a case study using data from an
that can predict accurately surface runoff is of interest to a experimental hill slope in a loess catchment.
wide variety of users, including city planners and irrigation Flood wave propagation in overland flow may be
practitioners. These models can also be easily extended described by the complete equations of motion for
to predict the fate and transport of contaminants along unsteady non uniform flow, known as the dynamic wave
the surface. It is widely accepted that a significant amount equations, first proposed by St. Venant in 1871. These
of contaminants originates from nonpoint sources and is equations are highly nonlinear and therefore do not
carried across the landscape by overland flow. have analytical solutions. Under a different set of sim-
plifying assumptions, more practical kinematic-wave is
presented18. These models constitute a relatively accurate
2. Review of Literature physical representation of the flow19, both approaches
For the study of rainfall runoff process, the factor that allow distributed overland flow and channel flow routing.
affects topographic and hydrological factors for runoff Kinematic wave model substitutes a steady uniform flow
generation has been investigated by number of inves- (stage-discharge) relationship for the momentum equa-
tigators on international literature in the past but still tion. However, unsteady flow is preserved through the
continue to be the object of research1 (e.g., Dunne, 1978; continuity equation.
Meyer, 1981; de Lima, 1988; Auzet et al., 1995; Basic et al., Henderson6 noted that kinematic waves behave closely
2001; Larue, 2001; Huang et al., 2002). However, runoff, to observed natural flood waves in steep rivers (slopes >
infiltration and soil erosion have been extensively studied 0.002). The overland flow component of the hydrological
in the field and laboratory most of the studies using simu- cycle is probably the most suited to such representation.
lated rainfall have applied rainfall at a constant rate. This For overland flow, the kinematic wave approximation to
contrasts with natural rainfall, which is highly variable in the Sant Venant equations of unsteady gradually-varied
both time and space20 (e.g., Huff, 1967; Eagleson, 1978; free surface flow is considered appropriate to most practi-
Sharon, 1980; de Lima, 1998). cal occurrences. The technique has been widely applied in
Willems20 studied that the rainfall simulation pro- practice, and has been investigate in a number of experi-
vides control of the spatial and temporal characteristics mental studies.
of precipitation, at both field and laboratory scale. Smith Lighthill10 developed the kinematic wave theory. The
and Woolhiser14 used a kinematic wave approximation theory became an accepted tool for modeling not only
to model the unsteady overland flow and were linked surface runoff but also for subsurface flow, soil moisture
dynamically to an infiltration model by using bound- movement, macro pore flow, snowmelt runoff, and soil

Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology 2097
 Extraction of Pigments from Rhodotorula Species of Dairy Environment

erosion. Smith14 focused on overland flow over sloping slope adjustment, continuous-spray, nozzles, supporting
infiltrating surfaces. The kinematic wave approxima- the nozzles; and the connections to the pumping system
tion was used to describe unsteady overland flow and and a constant head tank, and one electric motor for
was linked dynamically to an infiltration model by using artificial rainfall over the catchment. The pumping sys-
boundary conditions at soil surface. Singh13 conducted tem gives a stable pressure to avoid variations in rainfall
comparative study of kinematic wave model and Nash intensity during the simulated rainfall events. Pressure
model for a converging surface. The interdependent gauges monitored the pressure at the pump and nozzle.
behavior of kinematic wave parameters is shown and Although the rainfall simulator permits the use of several
based on this observation an experimental justification is rows of spray nozzles, the experiments described in this
given for retaining only one parameter in the kinematic study used eight nozzles, at a fixed height.
wave model.
The experimental investigations could be performed 2.1.2 The Flume
to understand surface, sub-surface or ground water runoff The impermeable plane surface (smooth painted metal
generation mechanism. Runoff is generated by rainstorms sheet) had a uniform rectangular cross-section having
and its occurrence and quantity are dependent on the 1.25 m wide and 5 m long, and that can be adjusted up
characteristics of the rainfall event, i.e. intensity, duration to 0 to 5% slope. The flume is filled with sediment having
and distribution. The mechanism of surface runoff gen- particle size ranges between 0.5 to 1.0 mm.
eration is however the focus of the present investigation.
The main objective of the present study was to quantify 2.1.3 The Runoff Recording System
the rainfall runoff process and its theoretical validation
using developed model to simulate observed and com- The purpose of runoff recording system is to determine the
puted data. Experiments were carried out using rainfall overland flow hydrographs in a particular time, the runoff
simulator. Several bed slopes and rainfall intensity were generated by each rainfall event is collected in a container
used for carrying out the laboratory experiments. placed at the bottom end of the flume, for the determi-
nation of the overland flow hydrographs. The runoff
recording system used a high sensitivity pressure trans-
2.1 Laboratory Set-up
ducer (depth sensor) connected to a data logger which
2.1.1 Rainfall Simulator records the amount of runoff water passing through a col-
The experiments described in this research work were lecting tank in time and the water level in the container.
carried out using a soil flume and a rainfall simulator. The
laboratory set-up for runoff generation was carried out in 2.1.4 Rainfall Intensity Pattern
the present investigation known as Advanced Hydrologic A recent study has emphasized the importance of spatial
System. Figure (1) shows a photographic representation rainfall intensity patterns of a storm on the shape of the
of the experimental setup. Advanced Hydrologic System
enables to demonstrate some of the major physical pro-
cesses found in hydrology and fluvial geomorphology,
including rainfall-runoff hydrographs for catchment
areas of varying permeability; the abstraction of ground
water by wells, with surface recharge and without surface
recharge from rainfall; the formation of river features
and effects of sediment transport. Realistic results can be
obtained from this small scale, floor standing apparatus,
which can be conveniently located in a laboratory and
requires no special services. Thus the apparatus is useful
to study the part of hydrological cycle bounded by the
arrival of net rainfall on ground surface and catchment
runoff either by surface streams or well abstraction. The Figure 1. S-12 MKII-50 Advanced Hydrologic Systems
basic components of the simulator are tilting flume for (Rainfall Simulator).

2098 Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology
 K. Shivalkar Yadav and R. Prabha

runoff hydrographs, times to peak and peak discharges ­ ifferential equation of the kinematic wave model which
d
(e.g. de Lima4). However, in this study, the simulated may be written in the combined form as follows:
rainfall pattern is used having three different pattern of ∂A ∂A
rainfall intensity i.e. 30 mm/hr. to 90 mm/hr. + amAm−1 = ql (3)
∂t ∂x 
Assuming an overland strip of unit width, the water area
3. Methodology
A can be replaced with the overland flow depth h, and
The experiments is carried out on a sand layer placed over lateral flow by rainfall excess intensity (i.e. q1=r-f  ). Thus,
an impermeable plane surface (smooth metal sheet), with the resulting kinematic wave equation for overland flows
a uniform rectangular cross section 1 m wide and 2 m is as follows:
long, as shown in figure 1, the sediment particle diam- ∂h ∂h
eter of 0.5 to 1.0 mm is required to fill over the flume. + ao mo hmo −1 = r −f  (4)
∂t ∂x
There is a slop adjustment device for adjusting the slope
in the range from 0 % to 5 %. Similar arrangement for qo = ao hmo  (5)
rainfall flow rate with different intensity ranging from 30
mm/hr to 90 mm/hr is possible with the help of 1/2 HP Where, r is the rainfall intensity in m/s, f is the infiltra-
pump. Water is reaches in the pipe line which is arranged tion rate in m/s, and ao = s1o/ 2 no and mo = 5 / 3. The most
above the sand tank and through the spray nozzle artifi- commonly used initial condition for overland flow is a
cial rainfall is spread over the catchment area having 2 m dry surface. For overland flow, the initial and boundary
length and 1m width in the flume filled with sand. During conditions are prescribed keeping in mind the fact that
rainfall water flows over surface of sand and reach to the the overland flow in small watersheds has negligible base
outlet and that runoff water flows through collecting tank flow. Considering these conditions the initial and bound-
filled with rectangular weir to measure the runoff water ary conditions have been defined as follow:
passing over it. The Depth sensor is attached below the
collecting tank to give the height over weir. hoi = 0 for every x at t = 0 (6a)

3.1 Mathematical Modeling qoio = 0 for every x at t = 0 (6b)

3.1.1 Kinematic Wave Formulation
The Kinematic wave (KW) theory was developed by hoo = 0 for every t at x=0 (6c)
Lighthill10, ever since then KW theory has been widely used
for modeling overland and stream flow. KW formulation qoj0 = 0 for every t at x=0 (6d)
uses physiographic parameters such Overland roughness, The calculation is made in a Cartesian coordinate rect-
slope, drainage area and length, and soil characteristics angular grid with the x-axis representing the distance
for computation of overland flow. The KW approximation in direction i, and having a step length Δx, and time on
of Saint-Venant’s equation can be written as the y axis with notation j, and having the time step of Δt.
Equation (4) and (5) can be solved using analytic method
∂A ∂(vA)
+ = ql  (1) for simple configurations Chow et. al (1988), however, in
∂t ∂x
present study, the governing differential equation were
Q = aAm  (2) solved numerically using explicit numerical method.
From equation (1) and (2) using implicit finite difference
Where n is the flow velocity in m/s, A is the water flow
method,
area in m2, Q is the discharge in m3/s, q1 is lateral inflow
hit++11 −hit+ 1 qit++11 −qit + 1 t+1
rate in m2/s, t is time in second, x is the distance measured + = ( r − f ) i+ 1  (7)
positive in the direction of flow in meter, and a and m ∆t ∆x
are parameters of the kinematic wave parameter which qit++11 = a(hit++11 )m  (8)
are closely related to the characteristics of the flow (Singh
1996). Equation (1) and (2) are the governing partial qit + 1 = a(hit + 1 )m  (9)

Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology 2099
 Extraction of Pigments from Rhodotorula Species of Dairy Environment

After simplifying the these equation, Where, Y0 = Runoff volume observed and Yc = Runoff
∆t ∆t ­volume computed.
f ( hit++11 ) = a ( hit++11 ) − a ht + 1 + hit+ 1 + ∆ t ( r − f ) i+ 1  + hit++11 
 ∆x ( i )
m m t+1

(10)
∆x

By solving the above equation using Newton-Raphson 4. Analysis and Discussion

iterative method, of Results
f (ht + 1 ) Experimental data obtained from Advanced Hydrologic
( hit++11 ) k+1 = ( hit++11 ) k − f / (hi+t+11 )k  (11)
system (rainfall simulator) is analysied using one dimen-
i+ 1 k
tional kinematic wave overland flow simulation model.
Where, The one dimentional kinematic wave model for overland
∆t flow, was developed in Fortran programming language.
f / ( hit++11 ) = 1+ a m( hit++11 ) 
m−1
(12) The simulation was done using time step ∆t of 5 sec and
∆x
spatial grid size ∆x is taken as 1 cm. Experimental data
were simulated using developed model in order to study
3.2 Nash Sutcliffe Criterion the overland flow roughness due to variation of slope and
The closeness of reproduction of the benchmark solution rainfall intensity on the catchment. Data was observed for
for performance evaluation of kinematic wave model solu- catchment slope between 1 % to 4 % and rainfall intensity
tion were evaluated using the well known Nash-Sutcliffe varies between 30 to 90 mm/hr.
efficiency criterion (Nash-Sutcliffe, 1970). It is experssed Developed kinematic wave model was calibrated
as; for manning’s roughness coefficient “n” to fit into data
Nash-Sutcliffe Efficiency in (%), observed from experiment for 30 mm/hr rainfall intensity
for different slopes. Plote between observed and model
∑
n
 (Yo −Yc )2  computated hydrograph are shown in Figure 2 to 5 for
NES =  1 − i −1  × 100 (13) slope of 1%, 2%, 3%, and 4% respectively.
∑
n
 (Yo −Ym )2 
 i −1 As can be seen from these plots the kinematic wave

model could reproduce resonably well rising limb of
Where, Yo = Observed flow (Experimental rainfall flow) the observed hydrograph as well as steady state dis-
value at time t, Yc = Computed flow (kinemaic flow) value charge. In case of falling limb of the hydrograph, two
at time t, Ym = Mean of observed values. segments are clearly visible. On where only surface
runoff domainates which is reproduceed well by kine-
3.3 Error in Runoff Volume Computation matic wave model and second the release of water from
The error in runoff volume for the present study was sand bed, which the kinematic wave model has not
­estimated by, simulated because the mechanism of release of water
 Y  from sand bed was not incorporated in the kinematic
Volumetric error, in % =  1 − c  × 100 wave model.
 Y  o

Table 1. Pertinent characteristics of observed and computed hydrograph for 30 mm/hr intensity
of rainfall and Area = 2m2
S.No Slope Volume Time to Peak NSE
Observed Computed Error Observed Computed Error
(%) (lit) (lit) (%) (min) (min) (%) (%)
1 1 9.3 7.9 4.3 2.7 2.4 11.1 98
2 2 6.3 5.9 6.3 2.3 2.1 8.6 97.2
3 3 8.3 7.9 4.8 2.1 2.0 4.7 91
4 4 7.6 7.3 3.9 2.0 1.8 5 93.2

2100 Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology
 K. Shivalkar Yadav and R. Prabha

Table 2. Pertinent characteristics of observed and computed hydrograph for 60 mm/hr intensity
of rainfall and Area = 2m2
S.No Slope Volume Time to Peak NSE
Observed Computed Error Observed Computed Error
(%) (lit) (lit) (%) (min) (min) (%) (%)
1 1 12.5 12.2 2.4 3.9 3.5 10.2 97.8
2 2 10.4 10.1 2.8 3.2 2.9 9.3 96.6
3 3 16.2 15.8 2.4 2.8 2.6 7.1 97.0
4 4 12.5 12.1 3.2 2.5 2.4 4 92.6

Table 3. Pertinent characteristics of observed and computed hydrograph for 90 mm/hr intensity
of rainfall and Area = 2m2
Volume Time to Peak
Slope NSE
S.No Observed Computed Error Observed Computed Error
(%) (lit) (lit) (%) (min) (min) (%) (%)
1 1 16.1 15.7 2.4 3.6 3.3 8.3 96
2 2 18.3 17.9 2.1 3.1 2.9 6.4 97
3 3 15.4 14.9 3.2 3.2 3 6.2 96.1
4 4 15.1 14.7 2.6 2.6 2.5 3.8 96.7

Figure 2. Comparison of observed and computed Figure 3. Comparison of observed and computed
hydrograph for rain fall intensity 30 mm/hr at 1 % slope. hydrograph for rain fall intensity 30 mm/hr at 2 % slope.

Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology 2101
 Extraction of Pigments from Rhodotorula Species of Dairy Environment

Figure 4. Comparison of observed and computed Figure 7. Comparison of observed and computed
hydrograph for rainfall intensity 30mm/hr at 3% slope of the hydrograph for rainfall intensity 60mm/hr at 1% slope of the
plane. plane.

Figure 5. Comparison of observed and computed Figure 8. Comparison of observed and computed
hydrograph for rainfall intensity 30mm/hr at 4% slope of the hydrograph for rainfall intensity 60mm/hr at 2% slope of the
plane. plane.

Figure 6. Variation of Manning’s roughness with overland Figure 9. Comparison of observed and computed
flow plane slope for rainfall intensity 30 mm/hr. hydrograph for rainfall intensity 60mm/hr at 3% slope of the
plane.

2102 Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology
 K. Shivalkar Yadav and R. Prabha

Figure 10. Comparison of observed and computed Figure 13. Comparison of observed and computed
hydrograph for rain fall intensity at 4% slope of the plane. hydrograph for rainfall intensity 90mm/hr at 2% slope of the
plane.

Figure 11. Variation of Manning’s roughness with overland Figure 14. Comparison of observed and computed
flow plane slope for rainfall intensity 60 mm/hr. hydrograph for rainfall intensity 90mm/hr at 3% slope of
the plane.

Figure 12. Comparison of observed and computed Figure 15. Comparison of observed and computed
hydrograph for rainfall intensity 90mm/hr at 1% slope of the hydrograph for rainfall intensity 90 mm/hr at 4% slope of
plane. the plane.

Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology 2103
 Extraction of Pigments from Rhodotorula Species of Dairy Environment

Figure 16. Variation of Manning’s roughness with overland

flow plane slope for rainfall intensity 90 mm/hr. Figure 19. Variation of Manning’s roughness with rainfall
intensity for 3% slope of the plane.

Figure 17. Variation of Manning’s roughness with rainfall Figure 20. Variation of Manning’s roughness with rainfall
intensity for 1% slope of the plane. intensity for 4 % slope of the plane.

Figure 18. Variation of Manning’s roughness with rainfall

intensity for 2% slope of the plane.

2104 Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology
 K. Shivalkar Yadav and R. Prabha

The comparison between the volume of observed and the surface slope are the major parameters controlling the
simulated hydrograph, its volumetric error, time to peak characteristics of the outflow hydrograph.
and Nash-Sutcliffe criterion for performance evaluation It is observed from Table 1 to 3 that for a given rainfall
at different value of surface slope is shown in Table 1. The intensity, an increases in the overland plane slope, reduces
value of Manning’s n obtained through calibration for dif- the time to peak. This may be due to raipd draing of the
ferent overland flow plane slope was plotted and given in water due to increased surface slope. For a given rainfall
Figure 14 It can be seen that from Figure 17 the value of intensity, the observed data were simulated very well at
n decreases linearly with increase in overland flow plane higher values of surface slope as compaired to lower value
slope. of surface slope. However we can also say that the at a
Further, the rainfall intensity is increased to 60 mm/ given rainfallintensity the kinematic wave model give a
hr and the value of slope of the overland plane is varied better fit for a steeper value of surface slope. Simillarly,
from 1% to 4% as shown in Figure 6 to 9 The compari- for a given surface slope, the observed data showed good
son between the observed and computed volume, time agreement for higher values of rainfall intensity.
to peak and the volumetric error and Nash-Sutcliffe It was found that the nummerical model simulates
efficiency shown in Table 2. The value of Manning’s n the rising limb of the observed hydrograph very well,
obtained through calibration for different overland flow but the the falling limb of the hydrograph was underes-
plane slope was plotted and given in Figure 15 It can be timated. This may be due to fact that, when water moves
seen that from Figure 18 the value of n decreases linearly on the surface in response to slope, it is postulated that
with increase in overland flow plane slope. two phenomena will act in conjuction to effectively alter
Now, the rainfall flow rate is increased to 90 mm/hr the hydrodynamics of the soil water. Firstly, water moving
and again slope of the overland plane is varied from 1% on the soil surface will exert a small pull on pore water
to 4%. The observed overland flow hydrographs and sim- by applying an upward suction force that is proportional
ulated by the kinematic wave model is shown in Figure to V2/2g , where V is the average surface water velocity
10 to 13. A comparison between the observed and com- which is several orders of magnitude larger than the pore
puted volume, time to peak and the volumetric error, water velocity. Due to this suction force water from the
and Nesh-Sutcliffe efficiency shown in Table 3. The value pores are drained out and thus contirbute to the surface
of Manning’s n obtained through calibration for differ- runoff which is not estimated by the one-dimentional
ent overland flow plane slope was plotted and given in kinematic wave model. Secondly, while applying the
Figure 16 It can be seen that from Figure 19 the value of water on the catchment, after stoping the rainfallintensity,
n decreases linearly with increase in overland flow plane some amount of water is still stored in the overhead pipe.
slope. This volume of water is then released subsiquently from
From Figure 2 to 13 it can be seen that, after the start the nozzle due to the gravity. This amount of water, is not
of the rainfall on the catchment, the water flow rate at the included in the numerical model.
outlet increases raipdly with time, this portion of hydro-
graph is known as the rising limb. However at a certain
time the water flow rate at the outlet equals the rainfall
5. Conclusion
intensity, this time is known as time to peak. After the The overland runoff hydrographs presented in this study
occurrence of time to peak the rainfall flow rate almost was conducted in the laboratory with equipment consist-
become constant with time. However the minor flac- ing of a rainfall simulator (nozzle spray), an impermeable
tuation are still observed, these may be due to variety overland flow plane i.e. smooth metal sheet and a run-
of reasons including actual flow phenamina. When the off recording system. The results indicate that there is
rainfall intensity is stopped the water flow rate at the less difference in the amount of runoff volumes, peaks at
catchment outlet start redusing at a very slow rate, this is different rainfall intensities and slope of the catchments.
known as the falling limb of the hydrograph. The results But considerable difference in falling limb of hydrograph
indicate a considerable difference between the observed shapes. Future laboratory and field experiments will test
and computed runoff volumes, time to peak, was obtained. relationships for a wider range of conditions that will
Therefore, it can be inferred that the rainfall intensity and include the use of other rainfall intensity, different slopes
and infiltrating surfaces. Future work should also include

Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology 2105
 Extraction of Pigments from Rhodotorula Species of Dairy Environment

the comparison of the experimental results with numeri- 3. Ibanez CAS, Calvo A. Design and operation of a small
cal results i.e. kinematic wave theory. and portable rainfall simulator for rugged terrain. Soil
Technology. 1997; 11:163–70.
4. De Lima JLMP, Singh VP. The influence of the pattern of
6. Acknowledgements moving rainstorms on overland flow. Advances on Water
Resources. 2002; 25(7):817–28.
The laboratory experiments described herein were con-
5. De Lima JLMP, Singh VP. Laboratory experiments on the
ducted in the Department of Hydrology, Indian Institute influence of storm movement on overland flow. Physics and
of Technology, Roorkee-247667 (India) Chemistry of the Earth. 2003; 28:277–82.
6. Henderson FM. Open Channel Flow. New York: The
List of notations MacMillan Company; 1966.
7. Ajay G. Modeling of Rainfall Runoff process of a small
A = Area (m2)
catchment. [M.E dissertation (Hydrology)]. University of
C = Courent number
Roorkee, India; 1979.
D = Hydraulic depth (m) 8. Hossain MM. Application of kinematic wave theory to
f = Infiltration rate (mm/hr) small watershed’s. [Ph .D. Thesis]. University of Roorkee
g = Acceleration due to gravity (m/s2) (India); 1989.
h = Depth of overland flow (m) 9. Das K, Ajit. Design and operation of a laboratory for rain-
i = Integer for space increment (mm) fall runoff simulation. [M.E dissertation (Hydrology)].
j = Integer for time increment (s) University of Roorkee, India; 1983.
K = Hydraulic conductivity (m/day) 10. Lighthill MJ, Whitham GB. On kinematic waves. I. Flood
k = Kinematic wave number movement in long rivers. Proc Royal Soc Lond. 1955; Series
L = Maximum length of flow(m) A 229(1178):281–316.
m = Kinematic wave parameter 11. Nash JE, Sutcliffe JV. River flow forecasting through con-
NES = Nash-Sutcliffe Efficency (%) ceptual model. Part 1-A discussion of principles: Journal of
Hydrology. 1970; 10:282-90.
n = Manning’s roughness coefficient
12. Singh VP. Kinematic wave modeling in water resources.
Q = Discharge in (m3/s)
Surface-water hydrology, New York: Wiley; 1996.
ql = Lateral inflow rate (m/s) 13. Singh VP. A laboratory investigation of surface runoff.
q0 = Overland flow(m3/s) Journal of hydrology. 1975; 25(3-4): 187–200.
r = Rainfall intensity (mm/hr) 14. Smith RE, Woolhiser DA. Overland flow on infiltrating sur-
S = Slope of the catchment (%) face. Water Resources Research. 1971; 7(4):899–913.
t = time increment (s) 15. Tekle, Alemayehu. The study of runoff mechanics of a natu-
Yc = Volume computed (%) ral watershed. [M.E dissertation (Hydrology)]. University
Yo = Volume observed (%) of Roorkee, India; 1998.
Ym =Mean of observed value (%) 16. Tossell RW, Dickinson WT, Rudra RP, Wall GJ. A portable
∆t =time step (s) rainfall simulation. Can Agric Eng. 1987; 29:155–62.
∆x = Space step (cm) 17. Turner AK. The simulation of rainfalls studies in overland
flow. Journal Inst Eng Austr. 1965; 39:9–15.
a = Kinematic wave parameter
18. Ponce VM. Engineering Hydrology: Principles and
ao = kinematic wave parameter for overland flow
Practices. Englewood Cliffs, NJ: Prentice-Hall; 1989.
19. Ponce VM, Simons DB. Applicability of kinematic and dif-
7. References fusion models. J Hydr Div. ASCE. 1978; 104(3):353–60.
20. Willems P. A spatial rainfall generator for small spatial
1. Bryan RB, Poesen J. Laboratory experiments on the influence scales. Journal of Hydrology. 2001; 252:126–44.
of slope length of runoff, percolation and rill development. 21. Yen BC, Chow VT. A study of surface runoff due to mov-
Earth Surface processes and Landforms. 1989; 14:211–31. ing rainstorms. Hydraulic Engineering Series No. 17,
2. Chow VT, Maidment DR, Mays LW. Applied Hydrology. Department of Civil Engineering, University of Illinois,
Singapore: McGraw-Hill, New York Book Co; 1988. p. 272– Urbana, USA. 1968.
309.

2106 Vol 7 (12) | December 2014 | www.indjst.org Indian Journal of Science and Technology