Modeling Earth Systems and Environment

[Link]

REVIEW ARTICLE

Modeling and investigating the effect of the LID methods

on collection network of urban runoff using the SWMM model (case
study: Shahrekord City)
Majid Omidi Arjenaki1 · Hamed Reza Zarif Sanayei2 · Heisam Heidarzadeh2 · Niloofar Aghili Mahabadi1

Received: 3 April 2020 / Accepted: 24 June 2020

© Springer Nature Switzerland AG 2020

Abstract
In recent decades, the accumulation of residential regions has increased due to the expansion of urbanization. This has led
to an increase in the impermeable areas and, thus, increasing surface runoff in the cities. Therefore, it is necessary to control
surface runoff using Low Impact Development (LID) methods in cities. In this research, using the hydraulic-hydrological
SWMM model, the collection network of surface runoff of Shahrekord city was simulated in 2, 5, and 10 years return peri-
ods. The calibrationof the model was performed in two rainfall events. Afterward, validation was performed using sensitiv-
ity analysis with calibrated values, which NSE, RMSE, and% BIAS indices showed good simulation accuracy. After this
stage, three methods of the green roof, permeable pavement, and rain barrels were located in 14 selected sub-catchments of
Shahrekord. The results of these methods on the volume and peak runoff of selected sub-catchment showed that green roof,
permeable pavement, and rain barrels, respectively, reduce the volume and discharge peak runoff by 46, 21, and 25%, on
average. Moreover, increasing the rainfall period would increase the effectiveness of using these methods. Also, the results
of investigating the value of the discharge and volume of runoff within the drainage canals, the capacity of the canals, and
downstream of selected sub-catchment, indicated the reduction of these values due to the application of LID methods. As a
result, it was concluded that the effect of using a green roof was greater than the others.

Keywords Rainfall · Runoff · Shahrekord · SWMM model · LID methods

Introduction life and financial damages are caused by the flood. Among
these floods, urban flooding is more important because it
For a long time, flooding has been one of the main issues of damages greatly the other areas. Considering the expansion
human society. This phenomenon is one of the hydroclimate of urbanization and growing industrialization, urban floods
events and is a kind of natural disaster that threatens human are intensified because the percentage of impermeable sur-
societies. Annually, in different parts of the world, various faces has increased and this will increase the volume of run-
off. In general, one of the most widespread effects of urban
development is the change in the amount and shape of the
* Majid Omidi Arjenaki flow resulting from rainfall and in contrast, the amount of
[Link].2885@[Link]
rainfall infiltration has been reduced into the deep reser-
Hamed Reza Zarif Sanayei voir and as a result, the volume and peak of the flood are
zarif@[Link]
increased. Nowadays, in urban areas, special methods have
Heisam Heidarzadeh been used to increase water penetration in soil, increase
Heidarzadeh@[Link]
resistance against the water flow, and runoff storage to
Niloofar Aghili Mahabadi restore the natural system. All modern methods of runoff
[Link]@[Link]
management are inspired by nature to reconstruct the condi-
1
Civil Engineering, Water Tendency and Hydraulic tions of rainfall-runoff in nature and reduce soil erosion and
Structures, Sharekord University, Sharekord, Iran runoff pollution rates (Wulliman and Thomas 2005). The
2
Faculty of Engineering, Assistant Professor of Civil use of green management technologies, which include Best
Engineering, Sharekord University, Sharekord, Iran Management Practice (BMP) and Low Impact Development

13
Vol.:(0123456789)
 Modeling Earth Systems and Environment

(LID), restructures the natural conditions of the catchment, retention. The purpose of their study was to save 80–85%
before urban development (Park et al. 2008a). On the other of the rain in Beijing city of China. Using the implementa-
hand, traditional urban runoff management methods are tion of the green roof, rain garden, and permeable pavement
unable to control and maintain the quality of surface run- during the 35-years return period, they concluded that by
off solely so that modern methods should be employed to implementing 30% green roofing, 10% atmospheric garden-
compensate for the reduction of permeable surfaces. Sev- ing, and 35% permeable veneering, of the total area, rainfall
eral studies have been carried out in the field of urban flood storage increases from 59.9 to 82.2% (Randall et al. 2019).
modeling using LID-BMP methods to reduce urban runoff. Poursahebi et al. investigated the effect of LID-BMP meth-
Hsu et al. used a combination of two SWMM models and a ods using the SWMM model on the runoff rate of 22 districts
two-dimensional surface flow diffusion model to simulate of Tehran. This study included five scenarios of current con-
flooding in the Taiwan urban catchment. They concluded ditions, green roof, permeable pavement, rain garden, and
that a hybrid model is appropriate for urban flooding due to green roof-permeable pavement in five periods of 2, 5, 10,
the overflow of water from collection networks of surface 50, and 100 years for 19 sub-catchment in this region. The
runoff (Hsu et al. 2000). Lowe applied the SWMM model results showed that the effectiveness of all three methods had
to design an underground sewage system. He pointed out an upward trend in decreasing runoff with the return period
that the SWMM model is thought to an only tool for sur- of up to 50 years, but the effect of these methods decreased
face runoff analysis so that it can also be used to analyze in the 100 years return period. The implementation of the
and design underground sewage (Lowe 2010). Jang et al. green roof and the rain garden had the most and least effec-
used the SWMM model to manage urban runoff before and tive in reducing runoff, respectively (Poursahebi et al. 2019).
after urban development and used the model in four regions The present study aimed at simulating the rainfall-runoff
of South Korea. A comparison of the results with previous process in the urban area of Shahrekord and then investigat-
studies for the same region showed that the SWMM model ing the effect of using LID methods on the volume, dis-
can deal with the combining different models such as the charge, and finally dimensions of the downstream drainage
shorter and longer duration of the discharge peak for post- network. An overview of the research process is shown in
development conditions (Jang et al. 2007). Park et al. simu- Fig. 1.
lated flow hydrograph and volume of pollutant loads using
the SWMM model to investigate the effect of watershed seg-
mentation and spatial separation of sewage drainage network
in South Korea. The results showed that the model had a Materials and methods
good estimation of peak discharge and runoff volume and
the effect of spatial resolution on surface runoff results was This study was conducted on the urban area of Shahre-
not significant (Park et al. 2008b). Soleymani et al. simulated kord in the southwest part of Iran. The SWMM software
runoff using the SWMM model assuming traditional and was employed to simulate the rainfall-runoff process. The
modern scenarios in the Tehran province of Iran. This study regions were divided into 114 sub-catchment and the status
revealed that a new system had better performance com- of the surface runoff collection network was investigated
pared to the traditional system and it is capable of reducing during the 2-, 5-, and 10-years rainfall return periods. In the
floods by 59% in this region (Soleymani et al. 2015). Wang following, 3 methods of the green roof, permeable pave-
et al. evaluated and modeled runoff reduction tools using ment, and rain barrels have been applied in 14 selected sub-
the SWMM model. They used three scenarios of perme- catchments according to theirs. Finally, the volume and dis-
able pavement, rain collecting, and a green roof, in which charge rate of runoff and the dimensions of runoff collection
the permeable pavement was the best tool to reduce runoff network downstream of these 14 selected sub-catchments
volume and reduced runoff by about 30% (Wang et al. 2017). were investigated before and after applying these methods.
Campisano et al. (2017) evaluated the LID-BMP methods
using the SWMM model in long-term and low-probability
rainfalls and it was shown that the accuracy of using this Introduction of the study area
system is higher in the long-term rainfalls. Tuomela et al.
(2019) investigated the quality of runoff and its pollutant The Shahrekord city is located on the eastern slope of the
loads using SWMM model and concluded that modeling of Zagros Mountains the southwest of Iran. The longitude of
runoff quality with constant mean concentrations of pollut- Shahrekord is between 50° 46′ 54′′ and 50° 55′ 43′′ and its
ants, when rainfall-runoff data are available from an outlet of latitude is 32° 18′ 03′′–32° 23′′12′. The highest elevation in
the catchment, would be largely unknown and they certainly the urban area of the city is 2220 m above sea level and its
cannot be commented on them. Randall et al. investigated area is estimated at 4756.23 ha (Ebrahimi 2019). Figure 2
the impact of using LID methods on urban runoff and rainfall shows an overview of the city’s location.

13
Modeling Earth Systems and Environment

Fig. 1  Research steps

Introduction of SWMM model obtained by combining the equations of motion (Eq. 1) and
Continuity (Eq. 2) for a gradually varied unsteady flow.
This model was designed by the US Environmental Pro- These equations are:
tection Agency in 1969–1971. This model is a dynamic
𝜕A 𝜕Q
simulator of rainfall-runoff (single event and continuous) + = 0, (1)
𝜕t 𝜕x
with the ability to account for evaporation, snowmelt, wet-
lands, deep infiltration, and subsurface flows. In this model,
flood wave is estimated using the kinematic method and a 𝜕Q 𝜕(Q2 ∕A) 𝜕H
+ + gA + gAsf + gAhL = 0. (2)
combination of ground and canalized flow elements. The 𝜕t 𝜕x 𝜕x
features of this model are the physical basis, the distributive That x is the distance from the beginning of the conduit;
view, the possibility of separately examining permeable and t is the time; A is the cross-section of flow; H is the head
impermeable areas, and the ability to simulate the nonlinear of the conduit; sf is the frictional slope in the conduit; hl is
response responses of the catchment to additional rainfall. the local energy loss per unit length of conduit, and g is the
The hydraulic part of this model is capable of routing runoff acceleration of gravity.
generated by three methods of steady flow, kinematic wave, This model simulates runoff flow in the form of a sub-
and dynamic wave. Flow routing in canals and junctions basin drainage network and all water passage channels and
is performed using the full form of St. Venant equations

13
 Modeling Earth Systems and Environment

Fig. 2  Location of the study

area

provides water surface profiles along and at the junction of obtained using land use maps and tables in ASCE2 as an
the channels. Therefore, this model can be used in urban average weight according to the area of each type of use.
runoff management and drainage network design (Tsihrintzis The input information includes hydrological information
and Hamid 1997). One of the unique features of this model is and hydraulic information. The hydrological information
the ability to simulate and apply LID methods. The model is including single event rainfall information for the calibra-
also capable of qualitatively examining water and simulating tion and validation of the model using the data was extracted
the effects of LID methods on runoff quality. from Shahrekord Meteorological Department and rainfall
pattern information, such as intensity–duration–frequency
Data required for simulation diagrams of rainfall in different return periods were obtained
from the studies conducted by Borujeni et al. (2011). This
In this study, the SCS1 method and Curve Number (CN) diagram is shown in Fig. 3.
were used to model the deep penetration of water into the Likewise, the hydraulic information consisted of sub-
soil, due to the essential and lack of need for large amounts catchments specifications (area, slope, percentage of
of data. It is noteworthy that the amounts of CN were impermeable areas, topography, etc.), which obtained using
Shahrekord detailed plan maps as well as land use maps
using the Civil 3D software. The information regarding the
canals and their dimensions was also obtained using the
1
Soil conservation service. maps of the comprehensive plan for surface water disposal
2
American society of civil engineers.

13
Modeling Earth Systems and Environment

Fig. 3  The curves of intensity–

duration–frequency of rainfall
in Shahrekord Station

in Shahrekord. In this study, the effects of snowmelt and Figures 4 and 5 have shown the rainfall intensity in
evaporation on the amount of surface runoff were neglected. these two dates and the observational and simulated runoff
depth diagrams over time. The simulated graph matches
well with the observational graph so that the results of
Calibration and verification of the model the simulation are reliable. According to the diagrams,
the flow within the canals has continued after the end of
To compare the observed hydrographs with calibrated the rainfall, and there was still runoff in the abovemen-
simulation hydrographs, it is necessary to use some statisti- tioned nodes for several hours after the end of the rainfall.
cal and probability functions. Some of the most common Calibration continued for several hours after the end of the
of these functions used in urban hydrology studies are the rain. Also, the maximum depth of runoff in both canals
Nash–Sutcliffe coefficient (Randall et al.), the Root Mean was different due to different dimensions of the canals and
Square Error (RMSE), and the BIAS%, which are presented corresponding to the amount of rainfall in the two nodes.
in Eqs. 3–5, respectively.
∑n � �2
i=1
Pi − Oi Sensitivity analysis
NSE = 1 − ∑ �
n �2 , (3)
O − ̄
O
Figure 6, presents the diagram of the sensitivity of param-
i=1 i i

� eters, roughness coefficient of permeable zones (N-imper-

∑n
(Pi − Oi )2 vious), coefficient of permeability of zones (N-pervious),
RMSE = i=1
, (4) the height of reservoir of permeable zones (DS tore imper-
n
vious), the height of permeable zones of storage (DS tore
∑n previous), and equivalent Width (W) and the Curve Num-
2
i=1 (Pi − Oi ) ber (CN).
BIAS% = ∑n × 100, (5) Figure 6, indicates that the equivalent width (W) and the
i=1
(Oi )2
Curve Number (CN) had a direct relationship with the run-
where Oi is the observed runoff; Pi is the simulated runoff; off. As the two parameters increase, the runoff increases as
Ō i is the average of observed runoff and n is the total number well. In contrast, other parameters are inversely related to
of data. the amount of runoff. Therefore, their increase decreases
%BIAS is the total error in the volume of the flow and it the volume of runoff. Also, among the mentioned param-
indicates the mean difference of the volume simulated by the eters, the equivalent width (W) has the highest sensitivity so
model with the observed flow. that by increasing it by 20% in all sub-catchments, the total
To calibrate the model, two rainfall events were randomly runoff volume of the catchment increases by 1.636%. After
selected for two runoff events on October 4 in 2019 and performing the calibration and sensitivity analysis on the six
October 23 in 2019. Their diagrams are shown in Figs. 4 abovementioned parameters, finally, verification was carried
and 5. out using the optimum parameters.

13
 Modeling Earth Systems and Environment

Fig. 4  The diagram of observed

and simulated runoff depth at
node 174 on 4 October 2019
history

Fig. 5  The diagram of observed

and simulated runoff depth at
node 362 on 23 October 2019
history

Verification than 0.5 the results are acceptable. The best value of this
coefficient is also 1, and the closer the values are to this
After performing the calibration on the sensitive parameters number, the higher the simulation accuracy is.
of the model, in two events October 26 in 2019, and October
29 in 2019, the runoff depth parameter was verified using Introducing LID methods
the optimal values of the SWMM model Figs. 7 and 8 dem-
onstrate the runoff depth parameters and the optimal values Low Impact Development (LID) methods are an innovative
of the SWMM model. approach in the management of urban flood, which model
Finally, to estimate the performance indicators of the and imitate nature. In this method, the rain is decentralized
SWMM model in the simulation of surface runoff in the at large and small scales at the source of rainfall. In this
calibration and validation steps, 3–5 equations were used method, decentralized rains are managed at large and small-
and the results were presented in Table 1. scaled locations at the source of rainfall. The purpose of
According to the results of Table 1, NSE > 0.84 in the various LID methods is to simulate the hydrological con-
calibration stage and NSE > 0.93 in the validation stage. On ditions of the catchment before urban development. The
the other hand, we know that if the NSE coefficient is greater majority of the methods include infiltration, purification,

13
Modeling Earth Systems and Environment

Fig. 6  Sensitivity analysis chart

of effective parameter on runoff
rate

Fig. 7  The diagram of observed

and simulated runoff depth at
node 174 on 26 October 2019
history

storage, evaporation, and maintenance of runoff near the The amount of coverage by this method was various in each
source of rainfall. case and was calculated according to the capacity of each sub-
In this study, three types of LID methods including a catchment extracted from the land use maps. In Table 2 the
green roof, permeable pavement, and rain barrel were used. amount of coverage by each of these three methods in different
These methods were applied in 14 selected sub-catchments sub-catchment is expressed.
of the city, which most of them are located in the northern Also, the input parameters for simulating the LID methods
and northeastern parts of the catchment. In general, said it are presented in Table 3. These values include information on
is notable that they are located in the upstream parts, which soil properties, the drainage layer, and the retention layer of
are the source of city runoff production. Figure 9 shows the these methods (Bai et al. 2019; Poursahebi et al. 2019).
location of this sub-catchment.

13
 Modeling Earth Systems and Environment

Fig. 8  The diagram of observed

and simulated runoff depth at
node 362 on 29 October 2019
history

Table 1  Values of SWMM model performance indicators for depth Discussion and results
parameter
Rainfall events Stage NSE RMSE (m) BIAS (%) According to Table. 4, applying green roof, permeable pave-
ment, and rain barrels resulted in the reduction of the volume
04-Oct-19 Calibration 0.843 2.632 4.108
and peak discharge of runoff in rainfall with a 2-year return
23-Oct-19 Calibration 0.898 1.39 3.053
period in the 14 sub-catchments.
26-Oct-19 Validation 0.937 1.155 1.456
According to Table 4, it is observable that the implemen-
29-Oct-19 Validation 0.973 1.673 2.704
tation of the green roof and permeable pavement methods

Fig. 9  The location of sub-catchment with LID

13
Modeling Earth Systems and Environment

Table 2  The percentage of covered areas of selected sub-catchment and sub-catchment 11 with a 12.75% decrease in runoff had
Method Green roof % Permeable Rain barrel
the highest and lowest reduction, respectively. Besides, in
pavement 2 years of the return period, the implementation of the rain
Sub-catchment % % barrels showed that sub-catchment 46 with 35.15% and sub-
catchment 42 with 15.63% had the highest and lowest reduc-
8 46 23.25 32.7
tion in runoff volume.
9 45 24.82 34.1
In green roof method, sub-catchments 11 and 43 with a
10 43 28 22.1
50% decrease in a peak discharge of runoff and sub-catch-
11 51 13.1 18.6
ment 97 with a 40.54% decrease in the discharge peak of
42 46.3 22.8 15.2
runoff, respectively, had the highest and lowest discharge
43 49.9 16.7 21.7
peak, respectively. In the permeable pavement method, sub-
46 50 15.8 34.9
catchment 10 with a 28.57% decrease in a discharge peak of
59 45.1 24.8 25.9
runoff and sub-catchment 11 with an 11.90% decrease in a
65 46.2 23 31.9
discharge peak of runoff had the highest and lowest decrease,
73 43.1 28.2 24.3
respectively. Also, using the rain barrels method during the
94 47.3 21.2 21.4
2-years return period, it was found that sub-catchment 9 with
97 43.3 27.8 25.7
37.50% and sub-catchment 42 with 12.50% had the highest
108 50 16 20.1
and lowest reduction of discharge peak of runoff.
109 48.5 19.2 24.2
The effect of the using LID methods on runoff of selected
sub-catchments during the 5 and 10 years return periods
was also investigated, in which the results are presented in
had the most and least effect on decreasing the volume of Tables 5 and 6.
runoff during the 2-years return period, respectively. In the The results of applying LID methods on the volume and
green roof method, sub-catchment 11 with a 51% decrease in peak runoff of selected sub-catchment showed that green
runoff volume and sub-catchment 97 with a 42.98% decrease roof, permeable pavement, and rain barrels, respectively,
in runoff had the highest and lowest reduction in runoff reduced the volume and discharge peak of runoff by 46,
volume, respectively. In the permeable pavement method, 21 and 25%, on average. Furthermore, according to the
sub-catchment 73 with a 28.33% decrease in runoff volume results of the LID methods in 14 selected sub-catchments in

Table 3  Input parameters for LID Parameters Permeable Green roof Rain barrel Unit
simulating the LID methods pavement

Surface Berm height 30 – mm

Manning coefficient 0.05 0.1 –
Pavement Thickness 150 mm
Void ratio 0.21 –
Permeability 100 mm h−1
Soil Thickness 12 mm
Porosity 0.5 –
Field capacity 0.2 –
Wilting point 0.1 –
Hydraulic conductivity 0.5 mm h−1
Slope 10 %
Storage Thickness 900 mm
Porosity 75 –
Hydraulic conductivity – mm.h−1
Barrel height 1500 mm
Drainage Flow exponent 0.5 0.5 –
Offset height 6 6 mm
Drain delay – 6 H
Slope – 1 1 %
Plant cover – 0 90 %

13
13
Table 4  Volume and peak discharge of runoff of selected sub-catchments, before and after applying LID methods for 2-year return period
Without LID Green roof Permeable pavement Rain barrel
3 3 3
S C Total Peak ­(m /s) Total Decrease % Peak ­(m /s) Decrease % Total Decrease % Peak ­(m /s) Decrease % Total Decrease % Peak ­(m3/s) Decrease %
Volume Volume Volume Volume
(× 103m3) (× 103m3) (× 103m3) (× 103m3)
(× 103m3)

8 1.4 0.23 0.76 45.71 0.13 43.48 1.07 23.57 0.18 21.74 0.94 32.86 0.16 30.43
9 0.92 0.16 0.51 44.57 0.09 43.75 0.7 23.91 0.12 25 0.61 33.70 0.1 37.50
10 2.48 0.42 1.41 43.15 0.24 42.86 1.79 27.82 0.3 28.57 1.94 21.77 0.33 21.43
11 2.51 0.42 1.23 51 0.21 50 2.19 12.75 0.37 11.9 2.05 18.33 0.35 16.67
42 0.96 0.16 0.52 45.83 0.09 43.75 0.74 22.92 0.12 25 0.81 15.63 0.14 12.50
43 0.78 0.12 0.39 50 0.06 50 0.65 16.67 0.1 16.67 0.61 21.79 0.1 16.67
46 1.65 0.27 0.82 50.3 0.14 48.15 1.39 15.76 0.23 14.81 1.07 35.15 0.18 33.33
59 2.26 0.38 1.24 45.13 0.21 44.74 1.70 24.78 0.29 23.68 1.67 26.11 0.29 23.68
65 1.26 0.21 0.68 46.03 0.12 42.86 0.97 23.02 0.17 19.05 0.86 31.75 0.15 28.57
73 2.93 0.50 1.67 43 0.29 42 2.1 28.33 0.36 28 2.22 24.23 0.38 24
94 1.32 0.22 0.7 46.97 0.12 45.45 1.05 20.45 0.17 22.73 1.04 21.21 0.17 22.73
97 2.28 0.37 1.3 42.98 0.22 40.54 1.65 27.63 0.27 27.03 1.7 25.44 0.28 24.32
108 4.71 0.75 2.36 49.89 0.39 48 3.97 15.71 0.64 14.67 3.78 19.75 0.61 18.67
109 4.89 0.81 2.53 48.26 0.42 48.15 3.96 19.02 0.66 18.52 3.71 24.13 0.62 23.46
Modeling Earth Systems and Environment
 Table 5  Volume and peak discharge of runoff of selected sub-catchments, before and after applying LID methods for 5-year return period
Modeling Earth Systems and Environment

Without LID Green roof Permeable pavement Rain barrel

3 3
S C Total Peak ­(m /s) Total Decrease % Peak ­(m /s) Decrease % Total Decrease Peak Decrease Total Decrease Peak Decrease %
Volume Volume Volume % ­(m3/s) % Volume % ­(m3/s)
(× 103m3) (× 103m3) (× 103m3) (× 103m3)
(× 103m3)

8 1.9 0.32 1.03 45.79 0.17 46.88 1.46 23.16 0.25 21.88 1.28 32.63 0.22 31.25
9 1.26 0.21 0.69 45.24 0.12 42.86 0.95 24.60 0.16 23.81 0.83 34.13 0.14 33.33
10 3.38 0.57 1.92 43.20 0.32 43.86 2.44 27.81 0.41 28.07 2.64 21.89 0.44 22.81
11 3.42 0.57 1.67 51.17 0.28 50.88 2.98 12.87 0.5 12.28 2.79 18.42 0.47 17.54
42 1.33 0.22 0.72 45.86 0.12 45.45 1.03 22.56 0.17 22.73 1.13 15.04 0.19 13.64
43 1.14 0.17 0.58 49.12 0.09 47.06 0.95 16.67 0.14 17.65 0.9 21.05 0.14 17.65
46 2.24 0.37 1.11 50.45 0.19 48.65 1.89 15.63 0.32 13.51 1.46 34.82 0.25 32.43
59 3.08 0.52 1.69 45.13 0.29 44.23 2.32 24.68 0.39 25.00 2.28 25.97 0.39 25.00
65 1.72 0.29 0.93 45.93 0.16 44.83 1.33 22.67 0.22 24.14 1.17 31.98 0.2 31.03
73 3.99 0.67 2.27 43.11 0.39 41.79 2.87 28.07 0.49 26.87 3.02 24.31 0.51 23.88
94 1.84 0.3 0.97 47.28 0.16 46.67 1.45 21.20 0.24 20.00 1.45 21.20 0.24 20.00
97 3.12 0.5 1.77 43.27 0.3 40.00 2.26 27.56 0.37 26.00 2.33 25.32 0.38 24.00
108 6.52 1.04 3.26 50.00 0.54 48.08 5.49 15.80 0.88 15.38 5.23 19.79 0.84 19.23
109 6.76 1.11 3.49 48.37 0.59 46.85 5.47 19.08 0.91 18.02 5.28 21.89 0.88 20.72

13
13
Table 6  Volume and peak discharge of runoff of selected sub-catchments, before and after applying LID methods for 10-years return period
Without LID Green roof Permeable pavement Rain barrel
3 3
S C Total Peak ­(m /s) Total Decrease % Peak ­(m /s) Decrease % Total Decrease Peak Decrease Total Decrease Peak Decrease %
Volume Volume Volume % ­(m3/s) % Volume % ­(m3/s)
(× 103m3) (× 103m3) (× 103m3) (× 103m3)
(× 103m3)

8 2.28 0.38 1.27 44.30 0.21 44.74 1.76 22.81 0.29 23.68 1.54 32.46 0.26 31.58
9 1.51 0.25 0.85 43.71 0.14 44.00 1.13 25.17 0.19 24.00 0.99 34.44 0.17 32.00
10 4.06 0.68 2.37 41.63 0.39 42.65 2.92 28.08 0.49 27.94 3.16 22.17 0.53 22.06
11 4.11 0.68 2.08 49.39 0.34 50.00 3.57 13.14 0.6 11.76 3.35 18.49 0.56 17.65
42 1.62 0.27 0.9 44.44 0.15 44.44 1.25 22.84 0.21 22.22 1.37 15.43 0.23 14.81
43 1.43 0.21 0.76 46.85 0.11 47.62 1.19 16.78 0.18 14.29 1.13 20.98 0.17 19.05
46 2.7 0.45 1.39 48.52 0.23 48.89 2.27 15.93 0.38 15.56 1.76 34.81 0.29 35.56
59 3.7 0.62 2.09 43.51 0.34 45.16 2.78 24.86 0.47 24.19 2.74 25.95 0.46 25.81
65 2.07 0.35 1.15 44.44 0.19 45.71 1.59 23.19 0.27 22.86 1.41 31.88 0.24 31.43
73 4.79 0.81 2.81 41.34 0.46 43.21 3.45 27.97 0.58 28.40 3.63 24.22 0.61 24.69
94 2.23 0.37 1.22 45.29 0.2 45.95 1.76 21.08 0.29 21.62 1.76 21.08 0.29 21.62
97 3.75 0.61 2.2 41.33 0.36 40.98 2.72 27.47 0.45 26.23 2.8 25.33 0.46 24.59
108 7.9 1.26 4.1 48.10 0.65 48.41 6.65 15.82 1.07 15.08 6.33 19.87 1.02 19.05
109 8.17 1.35 4.38 46.39 0.71 47.41 6.61 19.09 1.1 18.52 6.2 24.11 1.04 22.96
Modeling Earth Systems and Environment
Modeling Earth Systems and Environment

different return periods showed that with increasing return

period, the effect of using these methods on the volume and
discharge peak of runoff is. These results have been demon-
strated in Figs. 10 and 11.
As illustrated in Figs. 10 and 11, the maximum reduc-
tion in volume and discharge peak of runoff during different
return periods was the result of the implementation of the
green roof method. After that, the rain barrel method was
ranked second and the permeable pavement method comes
in last. Due to the good performance of the permeable
pavement method in reducing surface runoff, it is observed
that the effect of this method on the reduction of runoff is
Fig. 10  Results of implementing LID methods in 14 selected sub- less than other methods since, according to Table 2, this
catchments on volume of runoff during different return periods method had a lower amount of coverage area in the selected
sub-catchments.

Results of applying LID methods on collection

network of runoff

One of the important applications of LID methods is to pre-

vent flooding and to raise the collection canals of surface
runoff. These methods can reduce the discharge, depth, and
ultimately the volume of runoff inside the canals and prevent
flooding in the passage of the city. For example, Figs. 12 and
13 show the discharge peak and volume of runoff in canal
226, located downstream of the sub-catchment 8, 9, 10, and
11. Figure 14 also shows the position of this canal and its
upstream sub-catchments.
Fig. 11  Results of implementing LID methods in 14 selected sub-
As shown in Figs. 12 and 13, at the moment of the rain-
catchments on peak discharge of runoff during different return peri-
ods fall peak, the reduction rate of volume and discharge peak

Fig. 12  Diagram of peak discharge of runoff in canal 226

13
 Modeling Earth Systems and Environment

Fig. 13  Diagram of runoff volume in canal 226

Fig. 14  The position of canal 226

of runoff which occurred due to the implementation of LID Figure 15 shows the filled capacity of canal 266 before and
methods, was more increased. Among these methods, the after performing each of the LID methods. (Capacity is the
green roof method had the highest decrease rate in volume ratio of the height of the runoff within the canal to the total
and discharge peak of runoff in canal 226. Also, after apply- canal height).
ing LID methods the comparison was carried out, which According to Fig. 15, the empty canal capacity in the green
showed that the use of green roof, permeable pavement, and roof method is more than other methods and the impact of
rain barrel reduced the volume of runoff by 24.36%, 11.08%, using permeable pavement and rain barrel methods on canal
and 12.65%, respectively, in the canal 226. capacity is almost the same. Therefore, according to this study,
New LID methods can reduce the volume and depth of run- the green roof method can be the best option for controlling
off within the collection canals of surface runoff and no longer and reducing runoff in Shahrekord city.
need to increase the dimensions to prevent the canal flooding.

13
Modeling Earth Systems and Environment

Fig. 15  The filled capacity of canal 266 before and after performing each of the LID methods

Conclusion also help to prioritize these methods for use in different

regions.
This research was carried out using the SWMM software in
the urban area of Shahrekord. Using this software to model
natural events, such as the process of converting precipita-
tion to runoff, can provide us a comprehensive view of the References
city and its floodplains. In this research, the calibration and
validation of the model in different historical events showed Bai Y, Zhao N, Zhang R, Zeng X (2019) Storm water management of
low impact development in urban areas based on SWMM. Water
that the simulation had good accuracy. Then, three types of 11:33
LID methods were used in 14 selected sub-catchments of the Borujeni HS, Emad K, Fattahi R (2011) Evaluation of analysis methods
city that the results are as follows: of short-term rainfall (case study: Shahrekord Synoptic Station).
J Water Sci Eng 1:7–21
Campisano A, Catania FV, Modica C (2017) Evaluating the SWMM
– The model was able to well simulate the process of con- LID Editor rain barrel option for the estimation of retention poten-
verting precipitation to the runoff in this area and cor- tial of rainwater harvesting systems. Urban Water J 14:876–881
rectly determine the condition of the surface runoff of the Ebrahimi A (2019) Assessing the impact of urban expansion and land
city and its canals. cover changes on land surface temperature in Shahrekord city. J
RS GIS Nat Resour 9:102–118
– LID methods could reduce the amount of surface runoff Hsu M-H, Chen SH, Chang T-J (2000) Inundation simulation for urban
based on the amount of the coverage area. drainage basin with storm sewer system. J Hydrol 234:21–37
– The effect of using these methods increases with increas- Jang S et al (2007) Using SWMM as a tool for hydrologic impact
ing rainfall return period. assessment. Desalination 212:344–356
Lowe SA (2010) Sanitary sewer design using EPA storm water man-
– At peak times of rainfall, the impact of using LID meth- agement model (SWMM) computer applications in engineering.
ods was greater. Education 18:203–212
– The results of applying LID methods in 14 selected sub- Park J et al (2008a) Analysis of runoff reduction with LID adoption
catchments showed that the volume and discharge of run- using the SWMM. J Korean Soc Water Environ 24:806–816
Park S, Lee K, Park I, Ha S (2008b) Effect of the aggregation level of
off in these sub-catchments are reduced by 46, 26, and surface runoff fields and sewer network for a SWMM simulation.
25% due to green roof, permeable pavement, and rain Desalination 226:328–337
barrel methods, respectively. Poursahebi A, Niri MZ, Ghoudarzi SM (2019) Simulating of LID-BMP
– In general, the green roof method was introduced as the methods on urban runoff (case study: District 22, Tehran). J Water
Wastewater 30
best method for reducing and managing runoff in Shah- Randall M, Sun F, Zhang Y, Jensen MB (2019) Evaluating Sponge
rekord based on this research. City volume capture ratio at the catchment scale using SWMM. J
Environ Manage 246:745–757
Finally, researchers are advised to use LID methods to Soleymani M, Behzadian K, Ardeshir A (2015) Evaluation of strategies
for modifying urban storm water drainage system using risk-based
optimally manage surface runoff in urban areas. Understand- criteria. J Water Wastewater 26:16–29
ing the influential indicators in selecting these methods can

13
 Modeling Earth Systems and Environment

Tsihrintzis VA, Hamid R (1997) Modeling and management of urban Wulliman J, Thomas P (2005) Learning from nature: reducing urban
stormwater runoff quality: a review. Water Resour Manage stormwater impacts lake line magazine. Urban Drainage and
11:136–164 Flood Control District, Denver
Tuomela C, Sillanpää N, Koivusalo H (2019) Assessment of stormwa-
ter pollutant loads and source area contributions with storm water Publisher’s Note Springer Nature remains neutral with regard to
management model (SWMM). J Environ Manag 233:719–727 jurisdictional claims in published maps and institutional affiliations.
Wang Y, Sun M, Song B (2017) Public perceptions of and willing-
ness to pay for sponge city initiatives in China resources. Conserv
Recycl 122:11–20

13