CVEN3501

Week 5
This week
How to estimate baseflow
Factors influencing flow generation
Rainfall Runoff Modelling principles
Linear and non-linear models
Unit hydrograph
Questions from last week?
Rainfall-runoff model
Estimate:
Effective/Excess rainfall (total rainfall minus losses)
Baseflow
Catchment properties
 Total runoff and direct runoff

Rainfall Catchment Runoff

Direct runoff and baseflow
Why separate baseflow?
Event/storm rainfall runoff models only consider the excess
rainfall and the runoff that results from this
To calibrate the model, we need to compare to observed direct
runoff
For design, we need to add the baseflow back
Baseflow sometimes (often?) ignored
Groundwater

Image Ghost Watershed Alliance Society

Straight-line method
Baseflow recharge
Rainfall-runoff model
Estimate:
Effective/Excess rainfall (total rainfall minus losses)
Baseflow
Catchment properties
 Total runoff and direct runoff

Rainfall Catchment Runoff

Starting questions
Is flow quicker along the ground or in the stream?
Is flow quicker on steep slopes or flatter slopes?
Starting questions
Effect of catchment shape
Effect of slope
Effect of drainage density
Effect of land use
How do we model all these processes?
Rainfall Runoff routing
Refers to process where the rainfall excess is “routed” through
catchment storage
We “picture” the catchment like a dam with storage – models are
therefore called “reservoir models”
Reservoir modelling
Rainfall  Catchment (storage)  hydrograph
Conservation of mass

I = inflow – excess rainfall or upstream flow

S = volume of water
Q = outflow
(remember units – hectares, mm/hr, m3/s)
Reservoir modelling
Conceptual storage: a hypothetical reservoir that is specified to
produce the same routing effect on a flood hydrograph as occurs
in a real stream reach
Effects of storage
𝑑𝑆
𝐼 − 𝑄=
𝑑𝑡
No storage  no change in storage  I = Q
Large storage  big differences between inflow and outflow
Estimate streamflow
Given Inflow (I) and Storage S = f(Q)
Calculate change in storage:

ΔS = S2 – S1
= [(I1+I2)/2 – (Q1 + Q2)/2] ΔT

S2 + Q2 ΔT/2 = S1 + [(I1 + I2) – Q1] ΔT/2

Use relationship
Right Hand Side (RHS): all the known terms
between S and Q
Algorithm
1. Choose ΔT – time step for calculations
2. All initial values = 0 (S, I, Q)
3. For each time step (i), calculate RHS:
Si-1 + [(Ii-1 + Ii) – Qi-1] ΔT/2
4. Solve for Qi:
Si + QiΔT/2 = RHS by using Si = f(Qi)
5. Calculate Si
6. Repeat steps 3 to 5 for each time step until S i = 0
Relationship between Q and S
Linear case - simplest:
• Outflow is proportional to inflow
• Storage is related to outflow
• S = Kq
q here could be inflow or outflow (since they are proportional) –
generally it will be outflow e.g. S = 14102Q
Units for K = time (called the lag time)
S~Q relationship for the Hacking River
K
Example
Given a linear model S = 14102Q, estimate the hydrograph for a
catchment area of 24km2 and baseflow of 10m3s-1 for the
following storm:
Excess
Time Rainfall
(hours) (mm)
0 0
1 10
2 20
3 5
4 0
 Excess Excess
Rainfall rainfall in Direct
Time hyetograph flow units Baseflow Storage Runoff Runoff
(hours) (mm) (m3s -1) (m3s -1) (m3) RHS (m3s -1) (m3s -1)
0 0 0.0 10 0 0.0 10.0
1 10 66.7 10 106417 120000 7.5 17.5
2 20 133.3 10 401576 452834 28.5 38.5
3 5 33.3 10 576706 650318 40.9 50.9
4 0 0.0 10 499356 563095 35.4 45.4
5 10 386309 435618 27.4 37.4
6 10 298854 337000 21.2 31.2
7 10 231197 260707 16.4 26.4
8 10 178857 201687 12.7 22.7
9 10 138366 156028 9.8 19.8
10 10 107042 120705 7.6 17.6
11 10 82809 93379 5.9 15.9
12 10 64062 72239 4.5 14.5
13 10 49559 55885 3.5 13.5
14 10 38340 43234 2.7 12.7
15 10 29660 33446 2.1 12.1
16 10 22946 25874 1.6 11.6
17 10 17751 20017 1.3 11.3
18 10 13732 15485 1.0 11.0
19 10 10624 11980 0.8 10.8
20 10 8219 9268 0.6 10.6
21 10 6358 7170 0.5 10.5
22 10 4919 5546 0.3 10.3
23 10 3805 4291 0.3 10.3
24 10 2944 3319 0.2 10.2
25 10 2277 2568 0.2 10.2
26 10 1762 1987 0.1 10.1
27 10 1363 1537 0.1 10.1
28 10 1054 1189 0.1 10.1
29 10 816 920 0.1 10.1
30 10 631 712 0.0 10.0
 60.0

Streamflow (m3s-1)
50.0

40.0

30.0

20.0

10.0

0.0
0 5 10 15 20 25 30 35

Time (hours)
Example
Excess rainfall as a flow (m3s-1) = excess rainfall multiplied by
catchment area
LHS of equation = S+Q ΔT/2
But S = KQ  LHS = (KQ+Q ΔT/2)
LHS = RHS  Q = RHS/(K + ΔT/2)
Excess rainfall event lasted 4 hours.
Storage took 30 hours to reduce to zero
Relationship between Q and S
Non linear case – more realistic:
Lag time is related to outflow S i.e.= K(q) x q
Common relationship is power-law
S = kqm
k and m are parameters that need to be calibrated from observed
streamflow
Note difference between lag time (K) and lag parameter (k) –
confusing! But standard!
How to get this function?
Plotting all S versus Q points and fitting a function on it (linear or
non-linear).
Implications of non-linearity
Linear case
inflow and outflow are proportional
Double inflow  double outflow

Non Linear Case

Double Inflow  Outflow more than doubles
Non linearity parameter
S = k.qm

m < 1 means K is smaller for larger flows

i.e. larger floods are faster (K is lag time)
Bigger outflow from “storage”
This week
How to estimate baseflow
Factors influencing flow generation
Rainfall Runoff Modelling principles
Linear and non-linear models
Unit hydrograph
Unit hydrograph learning objectives
What are Unit Hydrographs?
How does the UH relate to its catchment?
How do we use the UH?
How do we create UHs of different durations?
Unit hydrograph: definition
A unit hydrograph is defined as “the characteristic response of a
given watershed to a unit volume of effective water input, applied
at a constant rate over a specified amount of time”.
Unit hydrograph: definition
It is a way of characterizing the shape of a hydrograph resulting
from effective rainfall (i.e. determine the peak and time to peak).
It can be used in conjunction with any effective rainfall
approaches to determine the whole catchment response.
It is a simple linear model that can be used to derive the
hydrograph resulting from any amount of excess rainfall.
Unit hydrograph: definition
Hydrograph that results from response from 1mm of uniform
rainfall during a storm of duration D
Unit hydrograph: assumptions
1. The excess rainfall has an intensity that is constant in time and throughout the
catchment.
2. The base time (duration) of the direct runoff hydrograph resulting from rainfall
of a given duration is constant, regardless of the rainfall intensity.
3. Values of direct runoff for a storm of given duration are directly proportional to
rainfall excess volumes (ie twice rainfall amount produces twice runoff
amount).
4. The hydrograph is assumed independent of antecedent conditions

Also- remember we are estimating direct

runoff only: no baseflow!
Using the UH
Linearity – you can multiply the UH by the recorded rainfall
amount
Linearity – you can add together multiple UHs
Using the UH
From the 1 mm UH, we can obtain the X mm UH by multiplying
by X.
If we have rain at different time periods we can add multiple UHs
together
This gives the final streamflow hydrograph.
Using the UH
Unit hydrograph
Unit Hydrograph (P=1)
1.0

0.8

0.6

P 0.4
0.2

0.0

15.0

10.0

Q 5.0

0.0
0 3 6 9 12

T
P=4mm at t=1
Streamflow Hydrograph (P1=4)
4

P2
1

60

40

Q 20
0
0 3 6 9 12
T
P=2mm at t=2
Streamflow Hydrograph (P1=0, P2=2)

2.0

1.5

P 1.0

0.5

0.0

30.0

20.0

Q 10.0
0.0
0 3 6 9 12
T
P=3mm at t=3
Streamflow Hydrograph (P1=P2=0, P3=3)
3.0
2.5
2.0

P 1.5
1.0
0.5
0.0

50.0

40.0

30.0

Q 20.0
10.0

0.0
0 3 6 9 12
T
Hydrograph for t=1 to 3
Streamflow Hydrograph (P1=4, P2=2, P3=3)

P2
1

120

80

Q 40
0
0 3 6 9 12
T
Properties (1)
UH is a catchment property - not a storm property. It is a model
that convert the rainfall hyetograph into a corresponding runoff
hydrograph.
The UH period (or duration) can be defined for any storm
duration
Properties (2)
The UH assumes the length of the hydrograph is the same for
different storms with the same duration (known as ‘time
invariance’).
• Say we have a storm that lasts 1 hour and 10 mm of effective rainfall is
recorded, and the resulting surface hydrograph has a duration of 5 hours.
• If you have another storm of the same length (1 hour) where 1000 mm is
rainfall is recorded, according to unit hydrograph theory, the surface
hydrograph should again be of a 5 hour duration.
• Equivalent to an exponent m=1 in the Q~S relationship S=KQm
Properties (3)
The UH assumes the effective rainfall is distributed uniformly
within the unit period the unit hydrograph is form.
The effective rainfall is distributed uniformly over the entire
catchment area too.
Linking to catchment area
The amount of effective precipitation is equal to the volume that
is in the stream as surface runoff,
the area under the unit hydrograph (the volume of water due to 1
mm excess rainfall over the entire catchment) is equal to the total
area of the catchment.
Example
Given a 2-hour UH and the rainfall intensities for a catchment,
estimate the resulting hydrograph.
Example 1
A 2-hour UH for a given catchment is:
Time (hours) 0 2 4 6 8 10 12 14
UH ordinates
0 12 42 70 51 22 10 0
(m3s-1)

The rainfall totals in a particular storm on the above catchment

were:
Time (hours) 0-2 2-4 4-6

Rainfall (mm) 20 40 0
2 hour UH
80

70

60

50

40

30

20

10

0
0 2 4 6 8 10 12 14
UH X 20 UH X 40 (Shifted by 2
hrs)
 Example 1 (solution)
How to use UH to estimate the surface runoff:
Time (hours) 0 2 4 6 8 10 12 14
20 x 12
First rainfall period 20 x 42 20 x 70 20 x 51 20 x 22 20 x 10
= 240

Second rainfall period 0 0 4012 40 x 42 40 x 70 40 x 51 40 x 22 40 x 10

Third rainfall period 0 x 12 0 x 42 0 x 70 0 x 51 0 x 22 0 x 10

Total 0 240 1320 3080 3820 2480 1040 400 0

Example 1 extended
Given the 2-hour UH and the rainfall intensities for a catchment,
estimate the resulting hydrograph.
Assume the loss rate for the storm was constant and equal to 3
mm/hour. Also assume that the baseflow is constant and equal to
100 m3s-1.
Example 1 extended
A 2-hour UH for a given catchment is:
Time (hours) 0 2 4 6 8 10 12 14
UH ordinates
0 12 42 70 51 22 10 0
(m3s-1)

The rainfall totals in a particular storm on the above catchment

were:
Time (hours) 0-2 2-4 4-6
Rainfall Intensity
13 23 4
(mm/hr)
Example 1 extended (solution)
First find excess rainfall; infiltration minus losses

Time (hours) 0-2 2-4 4-6

Excess Rainfall (13-3)2 =

40 0
(mm) 20
UH X 20 UH X 40 (Shifted by 2
hrs)
 Example 1 extended (solution)
Now add the baseflow to the modelled flow
Time (hours) 0 2 4 6 8 10 12 14
20 x 12
First rainfall period 20 x 42 20 x 70 20 x 51 20 x 22 20 x 10
= 240

Second rainfall period 0 0 4012 40 x 42 40 x 70 40 x 51 40 x 22 40 x 10

Third rainfall period 0 x 12 0 x 42 0 x 70 0 x 51 0 x 22 0 x 10

Total 0 240 1320 3080 3820 2480 1040 400 0

Baseflow 100 340 1420 3180 3920 2580 1140 500 100
What’s the catchment area?
P*A = V => A = V/P
V = (240 * 3600 *2) + (1320 * 3600 *2) + (3080* 3600 *2) + (3820
* 3600 *2) +
(2480 * 3600 *2) + (1040 * 3600 *2) +
(400 * 3600 *2) = 89136000 m3
P = 20 + 40 = 60 mm = 0.06 m
A = 89136000 / 0.06 = 1.49 * 109 m2
= 1490 km2
Formalizing the mathematical
representation of UH
The convolution equation:

Notations:
• Qi represents the i'th ordinate of the surface runoff hydrograph
• Ui represents the i'th UH ordinate
• Pi represents the i'th excess precipitation amount
• j is the number of precipitation hyetograph ordinates
• k is the number of unit hydrograph ordinates
• n is the number of hydrograph ordinates
 Streamflow Hydrograph (P1=4, P2=2, P3=3)

P2
P 1

120
j
80 final
Q 40 hydrograph
U
U Q
0
H
0 3 6 9 12
T

k n
Q0 0
Q1  P1U 1
Q2  P1U 2  P2U 1
Q3  P1U 3  P2U 2  P3U 1
......
Qj  P1U j  P2U j 1  P3U j  2  ...  PjU 1
Q j 1  P1U j 1  P2U j  P3U j 1  ...  PjU 2
......
Qk  P1U k  P2U k 1  P3U k  2  ...  PjU k  j 1
Qk 1  P2U k  P3U k 1  ...  PjU k  j  2
......
Qn  PjU k
Matrix representation
The matrix representation is the basis for estimating a unit
hydrograph based on recorded excess precipitation and surface
runoff for a given storm event. This is a convenient
representation for computer based calculations.
Matrix representation of the UH
If the rainfall had three time
increments (j=3), the UH had
five time increments (k=5), then
the resulting matrix equations
would be as shown:
Limitations of UH theory
1. Space invariance of effective rainfall
The effective rainfall is assumed to occur uniformly over the
catchment under UH theory. The spatial variability of rainfall
increases as the catchment becomes larger and larger. Spatial
variability is definitely an important factor when the storm is a
convective one.
Limitations of UH theory
2. Time invariance of effective rainfall
The effective rainfall is assumed to be constant in the time period
of the UH. This is never the case with recorded data.
Assumption of "linearity"
The UH theory assumes that the ordinates of the surface runoff
hydrograph are directly proportional to the volume of a given unit
storm
This permits the addition and subtraction of a UH to obtain
hydrographs for longer than unit storm durations.
This is the major limiting assumption behind UH theory. This is
important since all catchments are nonlinear in nature, although
some are more nonlinear than others.
Time period
Only temporal heterogeneity is considered, at the resolution of
the defined time period of the UH.
Since the time period of a UH is the only varying factor, one will
have different UH's for the same catchment depending on the
storm duration. Terms such as a 1-hour UH, 6-hour UH, 12-hour
UH are commonly used to denote the unit hydrograph for
differing time periods.
Deriving different durations
Once the unit hydrograph for a particular duration has been
developed, it is relatively easy to derive unit hydrographs for
other durations.
Two cases
• Case 1: Longer duration and integer multiple of the base UH (e.g. a 6 hour
hydrograph from a 3 hour UH)
• Case 2: Shorter UH (e.g. a 1 hour hydrograph from a 2 hour UH)
Case 1: multiple durations of the original
Easiest case
Uses multiple shifted and scaled versions of the original UH
Scaling factor is ratio between base duration and new duration
E.g. from 1 hour UH to 3 hour UH – we scale by 1/3
Scaling ensures that the new UH has a rainfall total of 1mm
Case 1: multiple durations of the original
Example: If we have 1 hour unit
hydrograph and need a 2 hour
unit hydrograph.
Case 1: multiple durations of the original
1. Use two x 1 hour unit
hydrographs and scale by 0.5
2. Lag the second UH by one 1+2
hour (blue)
3. Add the hydrograph
2
ordinates (green)
1
Example
Find the 4-hour UH for the following 2-hour UH

Time (hours) 0 2 4 6 8 10 12 14

UH ordinates
0 12 42 70 51 22 10 0
(m3s-1)
Solution
We need to take the 2 x 2 hr UH
We need 1 mm rainfall but we would have 2 mm (2 x 1mm) so we
need to scale by ½
We shift the second UH by 2 hours.
Solution

Time (hours) 0-2 2-4

Excess Rainfall (mm) 0.5 0.5

Solution
Time (hours) 0 2 4 6 8 10 12 14

Original 2hr UH 0 12 42 70 51 22 10 0

0.5 x 2-hour
0 6 21 35 25.5 11 5 0
UH1 (m3s-1)
0.5 x 2-hour
0 6 21 35 25.5 11 5
UH2 (m3s-1)
4-hour UH
0 6 27 56 60.5 36.5 16 5
(m3s-1)
 0.5 x 2 hour unit hydrograph 0.5 x 2 hour unit hydrograph (shifted)

4 hour unit hydrograph

2 hour unit hydrograph

Case 1
Note that the UH time increments do not need to be the same as
the time-period - as is the case above where they are at 2-hour
increments.
Case 2 – changing from a longer to
shorter period
More complicated because the new duration has to be an integral
multiple of the starting duration.
The method involves forming an S-hydrograph which is the
hydrograph from a unit rainfall per unit of time for an indefinite period
of time.
The S-hydrograph is formed by the cumulative effect of an infinite
number of T-hour UH's shifted by T hour time lags. This hydrograph
assumes a deformed S-shape and has a limiting value equal to the
unit rain amount times the catchment area per unit time (know why?)
Because a steady state has then been reached, and the storage has
no effect.
Case 2 – changing from a longer to
shorter period
More complicated because the new duration has to be an integral
multiple of the starting duration.
The method involves forming an S-hydrograph which is the
hydrograph from a unit rainfall per unit of time for an indefinite period
of time.
The S-hydrograph is formed by the cumulative effect of an infinite
number of T-hour UH's shifted by T hour time lags. This hydrograph
assumes a deformed S-shape and has a limiting value equal to the
unit rain amount times the catchment area per unit time (know why?)
Because a steady state has then been reached, and the storage has
no effect.
Estimation of a D-hour UH based on the
S-hydrograph of a T-hour UH
The difference between the two S-hydrographs, shifted by a lag D,
is the hydrograph for a unit rainfall occurring for the duration D.

Steps:
1. Construct the S-hydrograph from the scaled T-hour UH
2. Shift the position of the S-hydrograph by the duration D.
3. Compute the differences between the two S-hydrographs
4. Divide the estimated differences by D. This is the desired D-
hour UH.
Δt’ < Δt
 80
S-hydrograph

S-hydrograph ordinate (m3/s) Shifted S-hydrograph

60
40

2-hour UH x 2
=> 2 mm for 2 hours
20
0

0 5 10 15
Time (hours)
Example
Find the 2-hour UH from the following 4-hour UH

Time
0 2 4 6 8 10 12 14 16
(hours)
4-hour UH
ordinates 0 6 27 56 60.5 36.5 16 5 0
(m3s-1)
Step 1: Calculate the s-hydrograph
Find the S-hydrograph which is the hydrograph for a 1mm/hour
rainfall storm continuing indefinitely.
 Time
0-4 4-8 8-12 12-16 16-20
4-hour- (hrs) S-
Time
UH (m3s- Hydrograph
(hrs) Rain
1
) 4 4 4 4 4 (m3s-1)
(mm)

0 0 0 0
2 6 24 24
4 27 108 0 108
6 56 224 24 248
8 60.5 242 108 0 350
10 36.5 146 224 24 394
12 16 64 242 108 0 414
14 5 20 146 224 24 414
16 0 0 64 242 108 0 414
18 20 146 224 24 …
20 0 64 242 108 …
22 … … … …
24 … … … …
Step 2 - 4
Step 2: Shift the S-hydrograph by 2hours
Step 3: Find the difference between the 2 S-hydrographs
Step 4: Scale to find the 2-hour-UH.
 Difference (m3s-1) 2-hour-UH (m3s-1)
S- S-Hydrograph
Time (Hydrograph for 2 (Hydrograph for 1
Hydrograph (lagged by 2
(hrs) mm rain over 2 hour mm rain over 2
(m3s-1) hours) (m3s-1)
period) hour period)

0 0 0 0 0
2 24 0 24 12
4 108 24 84 42
6 248 108 140 70
8 350 248 102 51
10 394 350 44 22
12 414 394 20 10
14 414 414 0 0
16 414 414 0 0
Why should we study the UH approach?
It is a fundamental concept used in many catchment simulation
models.
It has been widely studied for use in ungauged catchments.
It provides a foundation for evaluating the function controlling
catchment runoff.
Summary
Main assumptions of the UH methodology:
• Catchments respond as linear systems, implying both the proportionality
principle and the superposition principle.
• The effective rainfall intensity is uniformly distributed over the entire river
basin.
• The rainfall excess is of constant intensity throughout the rainfall duration.
(Note IUH extensions).
• The duration of the direct runoff hydrograph, that is, its time base, depends
only on the effective rainfall duration.