water

Review
Application of Analytical Probabilistic Models in Urban Runoff
Control Systems’ Planning and Design: A Review
Ali Aldrees 1, * and Salisu Dan’azumi 1,2, *

1 Department of Civil Engineering, College of Engineering at Al-kharj, Prince Sattam bin Abdulaziz University,
Al-Kharj 16278, Saudi Arabia
2 Department of Civil Engineering, Bayero University Kano, Kano 700241, Nigeria
* Correspondence: [email protected] (A.A.); [email protected] (S.D.);
Tel.: +966-593364065 (A.A.); +234-8038547730 (S.D.)

Abstract: Urban stormwater is known to cause a myriad of problems, ranging from flooding to water
quality degradations. This paper provides an extensive review of analytical probabilistic model
(APMs) used in the design of urban runoff control systems. APMs are closed-form mathematical
expressions representing a long-term system’s output performance derived from the probability
distribution of the system’s input variables. Once derived, the APMs are easy to handle, allow for
sensitive analysis, and can be co-opted into optimization frameworks. The implementation of APM in
the planning and design of runoff control systems will not only help address the runoff quantity and
quality problems of urban stormwater, but will also go a long way in optimizing the benefits derived
from the systems. This paper reviews studies that document the negative impacts of urbanization on
runoff quantity and quality, and the best management practices (BMPs) used to mitigate the impacts.
Three design methodologies used in urban stormwater control systems were reviewed. A detailed
review of research on the development and use of APMs in urban stormwater management in various
runoff control systems is presented, and recommendations are proffered.

Keywords: best management practices; low-impact development; water-sensitive urban design;

blue-green infrastructure; sponge cities
Citation: Aldrees, A.; Dan’azumi, S.
Application of Analytical Probabilistic
Models in Urban Runoff Control
Systems’ Planning and Design: A 1. Introduction
Review. Water 2023, 15, 1640. Urbanization causes a disruption of the natural water cycle. The clearing of land
https://doi.org/10.3390/w15091640 surfaces reduces evapo-transpiration processes that intercept and return rainfall to the
Academic Editors: Dafang Fu,
atmosphere. Grading the land involves filling depression storage and the removal of topsoil
Haifeng Jia, Jiangyong Hu and while subsoil is compacted. The construction of impervious surfaces such as roads, roofs
Wei-Shan Chen and paved walkways reduces infiltration and increases surface runoff. Rainfall that used
to infiltrate the ground now runs off the surface at an increased rate, depending on the
Received: 20 March 2023 level of changes made to the land surface. These changes cause an increase in the peak
Revised: 15 April 2023
runoff and total volume of runoff. Conversely, the time of concentration of the catchment is
Accepted: 19 April 2023
decreased, which causes flows across the land surfaces to occur at faster rates. This effect
Published: 22 April 2023
is further aggravated by artificial drainage systems that are designed to convey runoff to
rivers as quickly as possible. With the development of impervious surfaces, infiltration into
the soil is reduced, thus reducing the quantity of water available to recharge aquifers and
Copyright: © 2023 by the authors.
feed-in the base-flow during dry weather periods [1–5].
Licensee MDPI, Basel, Switzerland. In addition to increasing the runoff quantity, urbanization also affects the runoff
This article is an open access article quality negatively, by increasing the concentration of pollutants carried by stormwater [6].
distributed under the terms and As runoff runs over roads, parking lots, rooftops of urbanized areas, it picks up a variety
conditions of the Creative Commons of pollutants and transports them to downstream water bodies. The receiving water
Attribution (CC BY) license (https:// body is affected by the cumulative impact of urban activities from the entire watershed,
creativecommons.org/licenses/by/ and the resultant changes in stormwater quantity and quality are felt in the downstream
4.0/). waters [7,8].

Water 2023, 15, 1640. https://doi.org/10.3390/w15091640 https://www.mdpi.com/journal/water

Water 2023, 15, 1640 2 of 23

Several attempts have been made by researchers to quantify the effects of urbanization
on stormwater runoff entering receiving water bodies. Todeschini (2016) [9] studied the
modifications of stormwater runoff and water quality caused by increased imperviousness
in the Bivio Vela industrial catchment in northern Italy. Runoff flows generated from
51 rainfall events of smaller and higher intensities were monitored. The results showed
that the conversion of 33% of pervious to impervious surfaces had resulted in a great
increase in peak flows, the volume of flows, the number and duration of combined sewer
overflows and the pollutants mass discharge. Likewise, Schütte and Schulze (2017) [10]
studied the effects of land-use changes due to the proposed urbanization of two sub-
catchments of uMngeni catchment (South Africa) on hydrological flows. The study used
ACRU software to model the current (2017) and future flows that may arise as a result of the
conversion of agricultural lands to impervious surfaces. The results show that increases in
impervious surfaces would result in a significant increase in stormflows due to a change in
rainfall–runoff conversion caused by the reduced evapotranspiration. Urbanization within
a watershed has a number of negative impacts on downstream waters. These impacts
include: changes to stream flow and stream geomorphology, degradation of water quality
and impact on aquatic habitat [1,4].
Stormwater best management practices (BMPs) are techniques, measures, or structural
controls that are used in a given set of conditions to manage the quantity and/or improve
the quality of stormwater runoff in the most cost-effective manner [11]. They are designed
facilities or modified natural environments that help control the quantity as well as improve
the quality of urban stormwater. Urban runoff control systems can be classified in to two
BMPs: (1) Methods that are used to reduce the volume of stormwater runoff that will
otherwise flow into the receiving water bodies. These methods allow the infiltration of
the stormwater into the ground, thereby aiding in groundwater recharge. (2) Methods
that remove pollutants from the stormwater through filtration, sedimentation, absorption,
biological uptake, etc. [12,13]. However, most stormwater BMPs serve both purposes.
The most commonly employed stormwater BMPs include various types of stormwater
ponds, filtration practices, vegetated channel practices, wetlands, pervious pavements and
rainwater tanks [9]. Green roofs iare also among the most commonly used stormwater
BMPs [14–16].
Due to their importance, sustainable stormwater management concepts have been
given different names all over the world. Qi et al. (2021) [17] presented a comprehensive
list of them, as synonymously adopted all over the world. The list ranges from BMPs to
low-impact development, integrated urban drainage systems, sustainable urban drainage
systems, stormwater control measures, water-sensitive urban design, resilience cities and
sponge cities. Nature-based solutions, green infrastructure, or blue–green infrastructure
are other terminologies also used to refer to methods used in mitigating the impact of
flood risk related to urban stormwater [18–20]. In its quest to restore its cities following
the negative consequences of stormwater runoff due to urbanization, China developed the
sponge cities plan in 2013, aimed at promoting source control. The concept uses natural
methods to retain rainwater, thereby recharging groundwater, reducing flooding and water
pollution problems, and gradually restoring the natural hydrology of the cities. The sponge
cities pilot scheme started in 30 cities, and following the successes recorded, the concept is
now being adopted at the national level [21,22].
This review paper compiles research on analytical probabilistic models’ (APMs’)
applications to urban stormwater management over the last 35 years, when the models
began to debut. A search of the literature was carried out in the SCOPUS and Google
Scholar databases using different combinations of the terms: “analytical probabilistic
models” AND “stormwater management” OR other BMPs such as “detention ponds,
bioretention cells, green roofs, pervious pavements, rain garden, etc.” A total of 183 entries
were returned by Google Scholar, while SCOPUS returned 45 entries in the first search. The
search was repeated and different entries were returned. The entries were filtered and a
total of 126 published articles found to be relevant were reviewed. This attempt to compile
Water 2023, 15, x FOR PEER REVIEW 3 of 24

Water 2023, 15, 1640 3 of 23

was repeated and different entries were returned. The entries were filtered and a total of
126 published articles found to be relevant were reviewed. This a empt to compile and
review studiesstudies
and review on theonusetheof
useAPMs
of APMsin urban stormwater
in urban stormwatermanagement
management is, is,
to to
thethe
best
best
knowledge
knowledge ofof
the authors,
the authors,the first
the ofof
first itsits
kind inin
kind this area
this ofof
area research. Figure
research. Figure1 shows
1 shows the
the
chronological
chronologicalorder
orderofof
the
thearticles.
articles.

Figure 1. Number of published articles on APMs of stormwater management.

Figure 1. Number of published articles on APMs of stormwater management.
2. Design of Urban Runoff Control Systems
2. Design of Urban
Runoff controlRunoff
systems Control
can beSystems
designed based on three approaches: a design storm,
Runoff control
continuous systems
simulation andcan be designed
analytical based on
probabilistic three approaches: a design storm,
models.
continuous simulation
The design stormand analytical
approach usesprobabilistic
the statisticalmodels.
analysis of rainfall to establish the IDF
The of
curves design storm
an area. approach
A design usesconsists
storm the statistical analysis
of rainfalls of rainfall
of various to establish
durations, the IDFto
developed
curves
serveof asanthearea.
inputA design
that thestorm
runoff consists
controlofsystem
rainfalls
may of experience
various durations,
during developed
its life timeto[5].
serve
The as therunoff,
peak input that the runoff
generated fromcontrol system
the design may is
storm, experience duringthe
routed through its runoff
life time [5].
control
The peak runoff, generated from the design storm, is routed through the runoff control
facility in order to estimate the facility’s capacity. Many authorities specify design storms
facility in order
in their to estimate
stormwater the facility’s
regulations. capacity. commonly
For instance, Many authorities specify design
used regulations storms
specify that
inpost-development
their stormwater regulations.
peak runoff For instance,
be less commonly used regulations
than pre-development specifyhaving
peaks for storms that
post-development
recurrence intervals peakof runoff
2, 5, and be10less thanwith
years, pre-development
an emergency peaks for spillway
overflow storms having re-
of 100-year
currence intervals of 2, 5, and 10 years, with an emergency overflow spillway of 100-year
recurrence interval capacity [23]. The duration of the design storm is chosen to be the
recurrence interval capacity [23]. The duration of the design storm is chosen to be the same
same as the critical duration or the time of concentration of the catchment. The design
asstorm used for
the critical the control
duration or the system
time ofisconcentration
selected based on risk
of the factors, The
catchment. suchdesign
as thestorm
risk of
overflow to downstream conveyance structures and land-use.
used for the control system is selected based on risk factors, such as the risk of overflow The assumption being
to downstream conveyance structures and land-use. The assumption being used in theof
used in the development of the design storm concept is that the recurrence interval
the resultingofrunoff
development is thestorm
the design same concept
as the recurrence
is that the interval
recurrence of the rainfall
interval producing
of the resulting the
event [24]. This assumption ignores the effect of storm separation time
runoff is the same as the recurrence interval of the rainfall producing the event [24]. This on runoff generation
and the temporal
assumption ignores variations
the effect of instorm
stormseparation
hyetograph pattern,
time on runoffwhich are known
generation andtotheaffect
tem- the
behavior of runoff. Due to this, the design storm approach suffers from severe criticisms
poral variations in storm hyetograph pa ern, which are known to affect the behavior of
voiced by many researchers, particularly in the design of stormwater storage systems [24].
runoff. Due to this, the design storm approach suffers from severe criticisms voiced by
Detailed shortcomings of the design approach can be found in [25].
Water 2023, 15, 1640 4 of 23

Instead of analyzing the rainfall history to extract concise information, continuous

simulation involves the conceptual representation of the catchment and a meteorological
input over a longer period. In this case, the effects of storm separation time and the temporal
variation of the storm event are captured. The long-term rainfall data are input directly
into the continuous simulation software [5]. The result of continuous simulation is the
response of the catchment to the rainfall input. However, despite the numerous advantages
of the continuous simulation approach, its major problem is its computational burden,
requiring a large number of simulations to evaluate system configuration [26]. The models
are also extensive, requiring a large computer memory and time for the system analysis [25].
Many types of computer software have been developed for the continuous simulation
of a catchment’s response to rainfall input, but SWMM is by far the most widely used
software. Todeschini et al. (2012) [27] used SWMM to investigate the effectiveness of design
configurations and operating conditions of stormwater detention tanks in combination
with flow regulators upstream. The results suggest that the optimum performances were
obtained by regulating flows into the detention tanks to values between 0.5 to 1 liters per
hectare, and tank volumes of 35 to 50 m3 per hectare of impervious area, for catchments
ranging from 4.8 to 48 hectors. These values gave the maximum annual pollutant mass
entrapped by the tank and minimum volume of stormwater sent to downstream treatment
plants. Mobley and Culver (2014) [28] used the SWMM model of a 12 ha residential
site, near Fort Collins, Colorado, to modify the design of a detention pond such that
the ecological flows downstream of the catchment were preserved, while maintaining
regulatory flow requirements. Pereira et al. (2019) [29] used PC-SWMM to simulate surface
runoff, and predict the impacts of urbanization and the use of detention ponds in an urban
sub-catchment in Brasilia. Thirteen different configurations of ponds were tested and the
best configuration of ponds was observed to be located near the outlet of the catchment.
This configuration was found to reduce flood peak discharge by 10 to 30%, and reduce
nutrient load by 40 to 60% and COD by 46%. Continuous simulation, though data-intensive
and time-consuming, is most commonly used to design or modify the design of existing
runoff control systems.

3. Analytical Probabilistic Models (APMs)

APMs are closed-form analytical expressions of a system’s output performance derived
from the probability distribution of the system’s input variables. The approach was initially
proposed by Benjamin and Connel (1970) [30]. The approach was first applied to water
resources by Eagleson (1972) [31] in water resources engineering, and later extended to
the area of stormwater management by many researchers [25,32]. In APMs of urban storm
water management, continuous rainfall data are divided into individual events using a
pre-defined minimum storm separation time, and the APM parameters are developed
using the rainfall statistics [32–34]. The input variable is rainfall characteristics (i.e., rainfall
depth, duration, inter-event time), and the output is the catchment’s response to the rainfall
input (i.e., runoff event volume, peak discharge, etc.). Using derived probability theory, the
probability distribution of output variables can be determined analytically if the probability
distributions of the input variables are known and the functional relationship between
them is either monotonically increasing or decreasing [35].
The application of APMs in urban stormwater management involves the following
stages, as described in [24–26,36–46]:
i. Selection of a case study of urban catchment and determination of its physiographic
information (catchment area, slope, proportion of imperviousness, drainage length,
depression storage, runoff coefficient);
ii. Discretization of the long-term continuous rainfall data into individual events using
the minimum inter-event time definition;
iii. Use of probability distributions, such as Exponential, Weibull, etc., to fit rainfall depth,
duration and intervention time;
iv. Development of APM parameters for the rainfall station;
Water 2023, 15, 1640 5 of 23

v. Development of APMs for stormwater characteristics such as runoff volume and peak
runoff captured by the facility, total runoff, pollutants captured and treated by the
facility, volume of spilled runoff, etc.
The APMs, once derived, were validated with other known approaches such as design
storm and continuous simulation, and the results were generally found to be in good
agreement [5,24,26,36–45]. Guo (2001) [5] assessed the suitability of using APMs in the
design of urban flood control detention ponds, alongside other approaches, namely, design
storm and continuous simulation, using a hypothetical catchment in Chicago. The results
show that all the three approaches generated similar results for the prediction of peak flow
of various durations from small urban catchments. Quader and Guo (2006) [46] studied
the discrepancy in peak flood estimations between design storm and the APM approach.
The effects of sub-catchment aggregation and time of concentration, as represented in the
two approaches, were examined. A case study catchment of Cataraqui North in Ontario
(Canada) was selected, and MIDUSS software was used for the design storm modeling.
The results were found to be in good agreement, with only 25% discrepancy between the
two peaks generated using the two approaches, which were attributed to sub-catchment
aggregation and the choice of time of concentration.
Exponential PDF is the most widely used distribution to model rainfall characteristics,
particularly in Canada and the USA, where the APMs of stormwater management were
developed. Hassini and Guo (2016) [47] used long-term rainfall data from seven rain-gauge
stations in northern USA to test the validity of using one-parameter exponential distribution
in modeling rainfall characteristics (depth, duration and inter-event time). Poisson’s
and Chi-square tests were used. It was found that exponential distribution fits rainfall
characteristics well, and was therefore recommended for APM. The exponential PDF is used
in more than 80% of the research papers reviewed herein. However, other distributions were
tested to determine their fit to rainfall characteristics in other regions of the world. In this
regard, Bacchi et al. (2008) [33] compared the use of one-parameter exponential distribution
and two-parameter Weibull distribution to model rainfall characteristics for three stations
in Italy. The results indicate that the Weibull distribution fits the Italian climate better. The
distribution was combined with rainfall–runoff transformation to derive the PDF for the
runoff volume and overflow volume of a storage facility, from which the design failure
probability can be obtained. Balistrocchi et al. (2009) [48] also used Weibull distribution to
model rainfall characteristics for Milano rainfall stations in Italy, while Generalized Pareto
Type 3 distribution was used to model rain storm depth with long durations in Toronto,
Canada [49]. Pareto and Gamma-2 PDF were found to work well for rainfall depth and
duration in Spain [50]. Weibull distribution was also found to fit rainfall characteristics
in Poland [51]. Log-normal distribution was found to fit rainfall characteristics for some
stations in Korea [52], while three-parameter exponential distribution was used to model
rainfall characteristics in Busan (Korea) as an improvement to single-parameter exponential
distribution [53].
In rainfall–runoff process modelling from urban catchments, Guo and Adams (1998a) [36]
used the exponential distribution to model the frequency distribution of rainfall depth,
duration, and inter-event time, from which a closed-form expression for the average
annual runoff event volume and the runoff event volume return period was derived. A
hypothetical catchment with a different runoff coefficient and various soils was used to test
the model against similar results obtained from a numerical simulation model (SWMM).
A close agreement between the analytical model and the simulation model was obtained.
Similarly, a close agreement was obtained between the results of runoff event volumes
and average annual runoff volume with a specified recurrence interval. Guo and Adams
(1998b) [37] also used the expression for runoff event volume and its duration together, with
the catchment’s average time of concentration, to derive an expression of peak discharge
rate, using the assumption of a triangular inflow hydrograph and exponential distribution
for rainfall depth. A closed-form analytical expression for the exceedance probability of
peak discharge per rainfall event and its return period was determined. The results from the
Water 2023, 15, 1640 6 of 23

analytical model compared favorably with those obtained using a continuous simulation
model, SWMM.
Guo et al. (2012) [54] further improved the APM by incorporating both Hortonian and
saturated overland flow mechanisms into the model to cater for the increasing use of low-
impact development (LID) techniques in urban watersheds, whose infiltration is always
below the natural infiltration. The PDFs of runoff event volume under the two scenarios
of infiltration excess and saturation excess runoff were derived separately, and combined
to obtain APM expressions for the expected value of the runoff event and its recurrence
interval, as well as the average annual runoff volume. The results were compared with
continuous simulation from HEC-HMS and there was very good agreement. Hassini and
Guo (2017) [55] derived APM expressions for the exceedance probability of peak discharge
in a small urban catchment considering triangular and trapezoidal hydrographs. Rainfall
data from Sherburne, Minnesota in the USA were used and a hypothetical catchment with
different times of concentration, imperviousness and soil types was assumed. Design storm
(using HEC-HMS model) was used to predict the peak discharges at different return periods.
The peak discharges generated from the developed APM were found to be comparable
with those of the design storm. Hassini and Guo (2020) [56] further added the effect of
saturation excess and infiltration excess runoff to their previous work [24] to develop APMs
for runoff event volume and exceedance probability of peak discharge in a small urban
catchment. Rainfall data from seven stations in the USA were used and a test catchment in
Hamilton (Canada) was used. The results of the APMs were found to be comparable to
those of design storm, with a percent difference ranging from 0.1% to 18%. Hassini and Guo
(2021, 2022) [57,58] recently developed a new and more accurate APM that can be used for
the design of runoff control systems. An APM for the frequency distribution of runoff event
volume was developed considering infiltration and saturation excess runoff generation.
The new model can effectively estimate runoff volume with different recurrence intervals,
and was found to be very sensitive to changes in soil saturation.
APMs have the ability to model flood routing. Guo and Zhuge (2008) [59] developed
analytical probabilistic expressions of flood routing to determine the probability distribu-
tion of peak outflow from a channel reach with and without a detention pond in between.
The outflow hydrographs were obtained, and the results of the analytical models were com-
pared with those of a single-event design storm using stormwater modeling software—the
MIDUSS and SWMM surface runoff routing algorithms. It was shown that the analytical
models compared well with the design storm. However, the use of different surface routing
models gave variations in the results of about 20%. Guo et al. (2009) [60] developed
closed-form analytical probabilistic channel routing equations for determining the flood
frequency distribution downstream of a catchment, given the catchments’ characteristics
and APM parameters derived from rainfall data. The model was verified by comparing
its results with those of HEC-HMS continuous simulation using 25 reaches and rainfall
data from Halifax, Canada. The results of the flood peak attenuations were presented as a
function of storage-delay time and return period, which can be used for watershed and
stormwater management purposes. Guo and Markus (2011) [61] enhanced the versatility
of APMs applied to small watersheds by incorporating SCS-CN for the determination of
rainfall excess from the catchments, and Clark’s unit hydrograph for runoff routing. The
results of the APM were compared with those of design storm using the HEC-HMS model.
Twelve watersheds were used in Chicago under urbanizing conditions, and the results
show that the analytical model can be used in stormwater management.
Flood peak estimation is another area wherein APMs were also found to be useful.
Guo and Dai (2009) [62] expanded the ability of the analytical model to cater for a larger
catchment and the master planning of a drainage system. A probabilistic rainfall areal
reduction method was used. Both the APM and design storm approaches were used to
obtain rainfall frequencies and flood peaks. A case study catchment of the Ganaraska river
watershed, Canada was simulated using the OTTHYMO model, and rainfall data from
Toronto Pearson International Airport station were used. The results show the capability
 reduction method was used. Both the APM and design storm approaches were use
obtain rainfall frequencies and flood peaks. A case study catchment of the Ganaraska r
Water 2023, 15, 1640 watershed, Canada was simulated using the OTTHYMO model, and 7rainfall of 23 data f
Toronto Pearson International Airport station were used. The results show the capab
of APMs in accurately representing the effects of rainfall characteristics across diffe
geographical
of APMs in accurately regions, and
representing the their
effects effects on flood
of rainfall frequency. across different
characteristics
geographical regions, In the
andcasetheirofeffects
modeling the pollutants
on flood frequency. build-up and wash-off from urban catchm
In the caseBehera et al. (2006)
of modeling [45] assumed
the pollutants build-up that:
and(1)wash-off
rainfall duration,
from urban inter-event
catchments, time, pollu
build-up
Behera et al. (2006) [45]and wash-off
assumed follow
that: (1) an exponential
rainfall duration,distribution;
inter-event (2)time,
the wash-off
pollutant load is unif
over the follow
build-up and wash-off entire catchment and depends
an exponential on the
distribution; (2)runoff volumeload
the wash-off generated, to derive an
is uniform
over the entire ical expressions
catchment for the PDF
and depends on theof runoff
wash-off load, generated,
volume expected value of pollutant
to derive analytical event wash
expressions for the PDF of wash-off load, expected value of pollutant event wash-off mean
load, annual average wash-off load and the long-term average pollutant event
centration
load, annual average (EMC). load
wash-off The analytical models were
and the long-term calibrated
average and verified
pollutant against values
event mean
served in a test catchment, and a good agreement
concentration (EMC). The analytical models were calibrated and verified against valueswas obtained.
The APMsand
observed in a test catchment, canabe used
good to screenwas
agreement runoff control alternatives in order to determ
obtained.
The APMs additional
can be used datatorequirements.
screen runoffSimilarly, the APMs can
control alternatives in be used
order toindetermine
sensitivity analys
additional datadetermine
requirements. the most important
Similarly, parameters,
the APMs can bewhich
used makes long-term
in sensitivity meteorological
analysis to c
putation
determine the most simplerparameters,
important and more economic,
which makes supports
long-termdecision-making
meteorological and eases stormw
compu-
system
tation simpler and more design
economic,[25]. supports
The APMs are computationally
decision-making and eases efficient
stormwaterand can be easily im
system
design [25]. The APMs are computationally efficient and can be easily implemented in or cont
mented in a spreadsheet or computer program, as compared to design storm
a spreadsheet or ouscomputer
simulation [54]. Therefore,
program, as compared analytical models
to design stormcanorbecontinuous
used as ansimula-
alternative to t
consuming
tion [54]. Therefore, continuous
analytical models simulation.
can be used as an alternative to time-consuming
continuous simulation. One of the greatest advantages of APMs over design storm and continuous sim
One of thetion is theadvantages
greatest option of co-opting
of APMs them into anstorm
over design optimization framework.
and continuous The optimizatio
simulation
is the option of runoff controlthem
co-opting systems
intocan
an be classified based
optimization on the objective
framework. function (i.e.,
The optimization of runoff q
tity, quality
runoff control systems can be and/or cost),based
classified uncertainty (deterministic
on the objective functionor stochastic),
(i.e., runoff and control appr
quantity,
quality and/or(static
cost), or dynamic)(deterministic
uncertainty [63]. Genetic algorithm, particle
or stochastic), andswarm
controloptimization,
approach (static ant colony o
mization, artificial bee colony, simulated annealing,
or dynamic) [63]. Genetic algorithm, particle swarm optimization, ant colony optimization,harmony search and cuckoo se
are some
artificial bee colony, of the optimization
simulated techniques
annealing, harmony that can
search andbe applied
cuckoo to optimize
search are somefloodof control
the optimization tems [17].
techniques that can be applied to optimize flood control systems [17].

4. APMtoApplication
4. APM Application to Urban
Urban Runoff Runoff
Control Control Systems
Systems
A schematic
A schematic diagram diagram
showing variousshowing various
stormwater stormwater
BMPs BMPs
is shown is shown
in Figure in Figure 2. A
2. APMs
havetobeen
have been applied applied
a wide varietytoofa stormwater
wide variety of stormwater
BMPs. BMPs.ofThe
The application APMsapplication
to these of APM
these systems
systems in discussed in discussed
under this section. under this section.

FigureBMPs.
Figure 2. Stormwater 2. Stormwater BMPs.

4.1. Detention Ponds/Stormwater Tanks

Detention ponds involve the temporary storage of runoff in ponds, basins or even
underground containers, and are meant to control the quantity as well as quality of urban
runoff downstream of a catchment [64,65]. The purpose of stormwater detention is to
reduce the flood damages caused by increased runoff due to imperviousness by limiting
Water 2023, 15, 1640 8 of 23

post-development peak discharges to be less than or equal to pre-development runoff [66],

or to a rate based on other criteria specified by the stormwater authorities in charge [67].
Furthermore, stormwater detention improves the quality of stormwater runoff in addition
to reducing the peak discharge [68]. The residence time resulting from stormwater detention
allows for the suspended particulate matter and adsorbed contaminants to settle [69,70].
As a BMP, detention ponds can help limit the pollutants loaded into receiving water bodies.
Many researchers have dedicated much attention to the application of APMs in de-
tention basins. Papa et al. (1997) [71] derived APM expressions for the pollution control
performance of detention ponds for different combinations of active to permanent pool
volumes. The results of the study have been compared to those simulated using SWMM
software. It was found that the degree of suspended solid removal in both cases was
comparable, with a difference of only 5 to 10% in extended dry ponds and 10 to 30% in
wet ponds. Guo and Adams (1999a) [42] derived analytical expressions for the probability
distribution of peak outflow rate from flood control detention ponds. The derived ana-
lytical expressions were used to determine the storage–discharge relationship required to
achieve the specified level of flood control at the facility. Using the runoff volume and
peak outflow rate presented in [36,37], the runoff rate exceedance probability per rainfall
event was derived based on different combinations of storage and outflow. Comparisons
were made between the results obtained from the analytical probabilistic model and similar
results obtained from SWMM software, and the results were found to be in good agreement.
Guo and Adams (1999b) [43] also used the expressions previously developed in [36,37] to
derive APM expressions for the long-term performance of a stormwater quality control
pond. The expression of flow capture efficiency was derived from the total spill volume,
while the volume-weighted average detention time of the basin was derived by taking
into account the variable inflow and outflow rates and the inter-runoff event time. The
APM expressions describing the detention time and the statistical solution of flow capture
efficiency were compared with similar values obtained from SWMM, and the results were
found to be in close agreement, thus confirming the validity of the assumptions made in
deriving the models.
Li and Adams (2000) [44] used an analytical probabilistic approach to derive runoff
quantity and quality control performances for urban runoff control systems. Rainfall was
first transformed to runoff, and the runoff transformed to overflow using the derived
analytical expressions. The runoff volume was also transformed to runoff pollution mass
load using the EMC concept, and was later transformed to total pollution mass discharge
load. The APM expressions for fraction of runoff overflow and total pollution mass
discharge load were used to derive closed-form APM expressions for the long-term runoff
control and long-term pollution control performances of the stormwater storage and
treatment systems. Comparisons between the runoff control performances predicted with
the analytical model (coded in computer programs called SUDS and EXSUDS) and those
obtained using a continuous simulation model STORM were conducted, and the results
were in reasonably good agreement.
Analytical expressions for runoff control performances using different forms of rainfall–
runoff transformations were developed [24,26,38–41]. Chen and Adams (2005a) [38] mod-
ified the rainfall–runoff transformation to consider infiltration rather than a common
runoff coefficient, and developed closed-form analytical expressions for runoff control
performances, including exceedance probability of a spill volume, expected value of a
spill volume, average annual volume and number of spills, and runoff capture efficiency.
The performance of the modified analytical model developed was tested against values
obtained from continuous simulation using SWMM and the analytical models developed
earlier by [25] for rainfall–runoff transformation (called ASTORM models), and good simi-
larities between the three results were obtained. Chen and Adams (2005b) [39] also used
the extended version of rainfall–runoff transformation, which divides the catchment into
pervious and impervious areas with different depression storages and runoff coefficients,
to develop APM expressions for the average annual number and volume of spills and the
Water 2023, 15, 1640 9 of 23

runoff control efficiencies. The results of the extended analytical model were compared
with those from ASTORM and SWMM, and the results were in good agreement, with the
extended model outperforming the ASTORM rainfall–runoff conversion model.
Chen and Adams (2006a) [40] used two types of rainfall–runoff transformations,
ASTORM and the extended ASTORM, to derive analytical expressions for stormwater
quality control based on build-up and wash-off functions. The appropriate models for
pollutant build-up and wash-off (designated as Type 1 and Type 2) were chosen, and
were combined to formulate the pollutant load model. Finally, the system quality control
measures were derived, which are closed-form expressions that can be used to evaluate
the long-term system behavior. Comparisons were made between the quality control
models developed with observed values, and the values predicted using SWMM gave
good estimates of system performance. Chen and Adams (2006b) [41] also used the
derived analytical expressions based on three different rainfall–runoff transformations
(i.e., ASTORM, Type 1 and Type 2) to derive APM expressions for stormwater quality
control measures. In this case, pollutant removal via the extended detention dry pond was
assumed to take place primarily through sedimentation, and TSS control was considered
as a surrogate measure of other pollutants’ removal. Closed-form APM expressions for
average annual volume of runoff, average annual number of spills, average annual runoff
control and pollutant removal efficiencies were derived. A comparison of the results from
Type 1, Type 2 and SWMM was conducted, and the results were in good agreement.
Chen and Adams (2007a) [26] used the ASTORM rainfall–runoff transformation,
extended ASTORM and the modified rainfall–runoff transformation to develop analytical
expressions for average annual runoff volume from an urban catchment. In the second
case, the Horton’s infiltration equation was slightly modified, in that the rainfall duration
was assumed to be a temporarily averaged constant. Model verification showed that both
of the two analytical models compared favorably with results obtained from SWMM. Chen
and Adams (2007b) [24] also used rainfall–runoff transformations, and pollutant build-up
and wash-off functions, to derive analytical expressions of the cumulative density function
(CDF) of pollutants load, as well as the expected value of pollutant EMC and average
annual pollutant EMC. In the rainfall–runoff transformation, two types of models were
proposed: the lumped parameter rainfall–runoff, and its extended form [39,44]. Two forms
of pollutant load model (Type 1 and Type II) were obtained, and the expected pollutants’
EMC and average annual pollutants’ EMC values were derived. The pollutant load models
were compared with observed values, and a good agreement was obtained. However, the
Type II load model was found to outperform Type I in the estimation of average annual
pollutants’ EMC.
Apart from the tremendous contributions made to the development of APM in relation
to detention pond’s analysis and design, coming from Canada and USA, some important
contributions coming from Italy are noticeable. Becciu and Raimondi (2014) [72] derived
APM expressions for the overflow spill of stormwater detention ponds. Two management
rules regarding the emptying of the pond were considered. Likewise, the probabilities
for spilled volumes varied from zero to one, corresponding to no spill and a spill volume
equal to the storage capacity of the pond, respectively. Data of rainfall series from Milano-
Monviso, Italy, were used. The resulting analytical expressions can act as very valuable tools
that can be used to estimate the overflow probability and the probability of a specific spilled
volume. Raimondi and Becciu (2015) [73] used rainfall statistics, detention pond outlet
operation rules, storage volume and maximum outflow to derive APM expressions for the
pre-filling probability of detention ponds. As in their previous paper, the same management
rules regarding the pond’s emptying were considered. The results can be used to estimate
the pond’s volume and out flow rate as a function of pre-filling probability. A comparison
of the analytical results with continuous simulation, using case study rainfall data from
Monviso, Milano (Italy), showed a very good agreement, thus confirming the applicability
of the method in the design and performance assessment of stormwater detention basins.
Becciu and Raimondi (2015) [74] derived a similar expression for the PDF of a detention
Water 2023, 15, 1640 10 of 23

pond’s spilled volume in order to evaluate its efficiency. Becciu et al. (2015) [75] also
derived APMs of retention time in stormwater detention ponds. The analytical formulae
developed can be used for the design of pond storage corresponding to a specified retention
time that ensures some pollutants are removed from the pond. The APM expressions were
validated against results from a continuous simulation using the case study in Monviso,
Milano (Italy), and were found to fit very well. Raimondi et al. (2022) [76] derived APM
expressions for the probability of runoff volume and residual storage in sustainable urban
drainage systems. The models were applied to two catchments in Genova and Milano
(Italy) using rainfall data from Monviso station. In both cases, the results were compared
with those from continuous simulation, and were found to be accurate.
Due to the shorter rainfall durations compared to the corresponding dry spell be-
tween the rainfall events, some researchers have considered rainfall arrival as a marked
Poisson’s process, and modeled rainfall characteristics stochastically [77–80]. Wang and
Guo (2019) [81] used analytical stochastic models (ASM) to describe the runoff capture
efficiency of detention ponds as a power function, rather than linear. The ASM results
were compared with the results of an SWMM continuous simulation using a case study
catchment area located in Jackson, Mississippi. The values of the root mean square error
(RMSE), Nash–Sutcliff efficiency (NSE) and correlation coefficient (R) for runoff capture
efficiency were 0.021, 0.994 and 0.9983, while these values for average pond fullness level
were 0.012, 0.998 and 0.9997, respectively. This indicates the applicability of ASMs.
Stormwater retention basins can also be analyzed by using stochastic water balance
to develop analytical models. Parolari et al. (2018) [82] developed a stochastic water
balance model of stormwater retention ponds under passive and active outlet conditions.
Analytical expressions of the steady-state and joint PDF of water level and valve closure
time, which can be used to define the water level and flow duration curves of the basin,
were derived. The model’s performance was tested by taking observations of water levels
from a retention pond located in Ann Arbor, MI, USA. He results show that the model
accurately predicts the water level PDF, which can be used to form a basis for evaluating
the changes in rainfall–runoff due to climate change and land-use.
Stormwater detention tanks are used to mitigate the impact of sewer overflow. Bal-
istrocchi et al. (2009) [48] applied APMs to develop a CDF of the overflow volume and
pollutant load distribution of a sewer tank. Weibull distribution was used to model rainfall
characteristics. Analytical expressions of performance indices such as the decrease in the
annual runoff volume and ratio of pollutant load captured by the tank were derived. The
model was verified with SWMM continuous simulation, using the urban catchment of Bres-
cia, Italy, and the results were found to be satisfactory. Andres-Domenech et al. (2010) [50]
derived analytical PDFs of the number of overflows, volume of overflows and overflow
reduction efficiency of a stormwater tank. Rainfall data from Valencia and Santander, Spain
using different probability distributions were tested. Pareto and Gamma-2 PDFs were
found to fit well. The analytical results regarding long-term volumetric flow and overflow
reduction efficiencies were compared with those of IW continuous simulation, and were
found to be similar. Becciu and Raimondi (2012) [83] developed APM expressions for
the pre-filling probability of stormwater tanks. The effects of minimum inter-event time
definition on outflow rate and storage volume were investigated using rainfall data from
Monviso, Milano, Italy. The results of the APM were compared to the results of continuous
simulation, and it was shown that the APM underestimated the pre-filling probability due
to some assumptions made in the development of the model. Thus, the model needs to
be refined further. Stormwater tanks, designed using APM, have also been found to be
capable of improving the quality of sewer discharges from catchments along the Tyrrhenian
coast of Italy [84].
Detaining runoff in stormwater detention ponds for a longer period improves the
quality of the treated runoff, but this poses the risk of overflow from subsequent rainfall,
which may generate runoff. There is an optimal detention time in the facilities such
that the trade-off between runoff and pollution control is addressed [18]. There is also
Water 2023, 15, 1640 11 of 23

a need to minimize the cost of building the facility, while at the same time achieving
the objectives. Papa and Adams (1997) [85] used APM expressions to develop a dynamic
programming model for the optimization of the cost of building detention ponds in multiple
parallel catchments, subject to meeting runoff quality control constraints. Shamsudin et al.
(2014) [86] used long-term rainfall data to obtain the rainfall characteristics and develop
APM parameters for a catchment in Malaysia. The APM was coded via particle swam
optimization (PSO) to develop a methodology that addresses the trade-off between the
runoff and pollution control performances of detention ponds. The detention pond’s
volume and outlet were appropriately sized such that a least-cost design was obtained.
Dan’azumi et al. (2013a) [35] developed APM parameters relevant to the rainfall
characteristics of Malaysian cities, and used the parameters to develop an optimization
algorithm via PSO that can be used to optimize detention time in wet detention ponds
such that they give the best pollution control performance [87]. Behera and Teegavarapu
(2015) [88] used the APM expressions of pollution control in extended wet detention ponds
to compare the results of three optimization techniques: dynamic programming (DP), non-
linear programming (NLP) and genetic algorithm (GA). They sought to obtain the optimal
values of pollution control, pond depth, storage volume and release rate of ponds treating
urban stormwater from multiple sub-catchments releasing their outflow into a common
downstream point, such that the quality control target at the downstream river could be
met at minimum cost. The results show that the NLP and GA provided an improved
solution compared to the DP.

4.2. Rainwater Harvesting System/Rainwater Tanks

Rainwater tanks, consisting of rain-barrels and cisterns, are rainwater harvesting
systems (RHS) that store rainwater for household use and reduce the volume of runoff
generated from urban surfaces. The use of rainwater tanks reduces water consumption
from municipal supply, and thus reduces the water bill. The water stored in the tank can
be used for gardening and toilet flushing, thus reducing municipal water consumption.
Some rainwater tanks have two compartments: the rainwater tank itself and an infiltration
facility, which aids in groundwater recharge [89,90].
Raimondi and Becciu (2014a) [89] developed APM expressions to estimate the prob-
ability of meeting the water demand using rainwater tanks as a function of household
population and number of storm events occurring, using long-term rainfall data from
35 years at the Milano-Monviso station. The results of the study can be used to determine
whether it is efficient to use rainwater harvesting alone, or in combination with munici-
pal water supply. Raimondi and Becciu (2014b) [90] developed APMs for the design of
multi-use rainwater tanks. These rainwater tanks were designed to have two basins: a
rainwater basin and an infiltration basin. A trade-off between the risk of water shortage
in the basin and the risk of overflow was studied. The results of a case study in a catch-
ment in Milan, Italy, show that the probability of complete rainwater use in a household
depends on the period of regulation, with weekly regulation yielding a higher probability
compared to daily regulation. Additionally, the probability of overflow was high for a
small storage volume and low infiltration rate. Becciu et al. (2016) [91] improved on their
previous models by considering the effect of re-filling during the regulation period, and
developed an analytical expression to estimate the CDF of active storage in the rainwater
tank. The results were compared with those of a continuous simulation model using data
from Milano, Italy, and there was a good agreement.
Guo and Baetz (2007) [78] derived an analytical expression that could be used to
design rainwater storage units in green buildings, focusing on the rate of water use in the
building, the climate characteristics of the area and the reliability of the system. The APM
was applied to a hypothetical catchment in Chicago and Montana, USA, and it was shown
that the APM provided an efficient approach to designing the system. De Paola and De
Martino (2013) [92] studied the efficiency of four stormwater tank configurations using
SWMM, and applied the semi-probabilistic approach to determine the qualitative and
Water 2023, 15, 1640 12 of 23

quantitative stormwater capture efficiencies of the most efficient tank configuration. It was
concluded that the analytical approach provided similar results to continuous simulation.
Kim et al. (2012) [93] used mass balance equations for each component of a rainwater tank
to develop APM expressions for the rainfall–runoff reduction in an RHS. The PDF and
CDF of runoff from the catchment and the RHS were derived, and the expected value of
runoff volume was determined. The model was applied to a dormitory building in Seoul
(Korea) to design an RHS and to estimate the runoff reduction achieved as a result of it.
Di Chiano et al. (2023) [94] used APM expressions to derive the CDF of active storage
in RHS. Active storage was considered as a function of rainfall moments, water demand
and mean number of chained events under deficit conditions. The results of the model
were compared with those of continuous simulation, using rainfall data from Monviso,
Milano (Italy), focusing on a case study of RHS in Milan. An average normalized RMSE of
0.033, under three demand conditions, was obtained between the APM and the continuous
simulation, suggesting a very good prediction.
Stochastic mass balance equations of RHS have been used to develop analytical models
for RHS systems. Guo and Guo (2018a) [95] derived an ASM that could be used to determine
the size of an RHS using a differential mass balance equation. Analytical expressions of a
rainwater tank’s efficiency in terms of water supply reliability, required storage volume
and its runoff reduction benefits were derived. The stochastic models, developed using
rainfall data from five different climates (Atlanta, Concord, Detroit, Flagstaff and Billings)
in the USA, were validated against the results obtained from SWMM continuous simulation
and also those of Guo and Baetz (2007) [78]. The values of mean Nash–Sutcliffe efficiency
(NSE), root mean square error (RMSE) and correlation coefficient of 0.98, 0.035 and 0.99,
respectively, were obtained, indicating a good result. Pelak and Porporato (2016) [96]
modeled rainfall as a marked Poisson’s process, and developed an analytical expression
that optimizes the volume of a rainwater harvesting system at minimum cost. The volume
was expressed as a function of rainfall parameters, roof area, water use rate, and the
cost of the cistern and that of the external water source. The cost consists of fixed and
distributed costs. The result of the study can be used to size an RHS in any climate. This
will help reduce urban stormwater runoff and water consumption from public mains. Sim
and Kim (2020) [97] used stochastic mass balance to develop an analytical model for the
quantification of the water supply and stormwater interception efficiency of an RHS. In
the study, the sensitivity of the RHS to climate change was evaluated, and the model was
assessed using rainfall data from Busan (Korea). The results of the analytical model were
compared with those derived using multiple regression. The R2 and RMSE values for
water supply and stormwater interception efficiency ranged from 0.91 to 0.96 and 0.026 to
0.033, respectively. Cheng et al. (2021) [98] also used water balance to develop a stochastic
model of an RHS. Due to the random occurrence of rainfall, the reliability of the model was
expressed in terms of the fraction of time for which the RHS satisfies water demand. The
model was applied to three RHSs in Toronto, Canada, and was found to have high accuracy.

4.3. Green Roofs

A green roof is a rooftop garden. These are used to provide shade, reduce the temper-
atures of the roof surface and surrounding air, and to moderate the heat island effect [99].
Green roofs comprise four layers: a vegetation layer, a substrate layer, a drainage layer
and a waterproof layer. Some green roofs have a water storage layer combined with the
drainage layer for holding more rainwater. Vegetation is planted on top of the substrate
layer, where rainwater is retained. Excess rainwater from the roof is drained through the
drainage layer [100].
Researchers also explored the application of APMs to green roof design and analysis.
Zhang and Guo (2013a) [79] derived analytical expressions for runoff generation from
green roofs. The results obtained from the analytical models were compared with those
of continuous simulation using the LID module of SWMM, and also from the field results
derived from a real case study in Portland, USA. The results of the APM were found to
Water 2023, 15, 1640 13 of 23

be in good agreement with both. Additionally, Guo et al. (2014) [101] derived analytical
expressions for long-term average runoff reduction rates (defined as the ratio of total runoff
captured to that of total runoff generated) and the irrigation water requirement of green
roofs. The performance values of the APM in terms of runoff reduction rates and irrigation
time fraction at different growing medium depths under semi-arid climate (Atlanta, USA)
and humid climate (Billings, USA) conditions were compared with those from continuous
simulation using SWMM, and it was concluded that the APM can be used as an alternative
to SWMM in the planning, design and management of green roofs.
Guo (2016) [102] further refined the work of [101] by considering rainfall occurrence as
a stochastic process to derive a stochastic differential equation of green roofs. The stochastic
water balance equation was formulated to determine the mean and PDF of the moisture
contents of green roofs. The accuracy of the model, in terms of runoff reduction rates and
irrigation time fraction, was evaluated by testing the results against those of SWMM contin-
uous simulation using four sets of rainfall data from Billings, Phoenix, Atlanta and Boston
(USA), and using sandy and loamy soils as growing media. The comparison of results
between SWMM and the stochastic model implied the good correlation coefficients of 0.993
and 0.995, respectively, for runoff reduction rates and irrigation time fraction. Most studies
assume that the RHS is dry at the beginning of the rainfall event. However, some moisture
retention is possible in the roof at the beginning of the next rainfall period. Raimondi and
Becciu (2020) [100] considered the possibility of pre-filling from previous rainfall events to
develop an APM for the design of green roofs. The APM was tested against the results of
continuous simulation, using rainfall data from the Milano Monviso (Italy) station. The
results show that the model compared well with continuous simulation. Thus, the APM
can be used for the optimization of the design of green roofs. Raimondi et al. (2022a) [103]
extended their work from 2020 to develop an APM that could be used to determine the
thickness of the substrate layer of green roofs as a function of runoff reduction. The results
of the study compare well with those obtained from continuous simulation.
Raimondi et al. (2022b) [104] used APM to develop a model for the survival of
vegetation on green roofs without the need for irrigation. The thickness of substrate
medium and risk of vegetation withering were combined to design the green roof. The
APM was tested using two green roofs in Milano and Calabria (Italy). The results from
the analytical model compared excellently with those of continuous simulation. Guo et al.
(2022) [105] used a stochastic model of rainfall to model the hydrologic and hydraulic
process occurring on green roofs. Both the saturation excess runoff and infiltration excess
runoff were considered, and analytical equations that can be used for the quantification
of the performance of green roofs were derived. The results of the analytical model were
compared with those of continuous simulation, and were found to be accurate. Raimondi
et al. (2023) [106] used APMs to determine the probability that runoff from a green roof will
exceed a certain threshold, given the substrate thickness, climatic variable and moisture
content of the roof. The reduced retention capacity of the system due to previous rainfall
was also considered. The analytical model was tested using a case study in Milano (Italy),
and the results were similar between the model and the continuous simulation.

4.4. Filtration Practices

Filtration practices are surface or underground practices that reduce the volume of
runoff by infiltration through the soil. They provide a performance that is independent
of local conditions, and their designs are applicable to roadside and congested urban
conditions. According to [107], bioretention cells and sand filters are amongst the filtration
practices commonly used for small to medium catchment basins, because they usually
occupy only 2 to 3% of the drainage area, and hence are suitable in dense urban settings.
Sand and gravel filters are also commonly used as filtration practices for the management
of urban stormwater [108]. Other filtration practices include pervious pavements, etc.
Water 2023, 15, 1640 14 of 23

4.4.1. Bioretention Cells/Biofilters/Rain Gardens/Impervious Area Disconnection

Bioretention systems are shallow landscaped depressions, commonly located in park-
ing lots or within residential land-use areas, that are designed to incorporate many of the
pollutant removal mechanisms that are operated in forested ecosystems. They are also
known as biofilters or rain gardens [13]. Stormwater treatment in a bioretention cell is
achieved through sedimentation, filtration, soil adsorption, micro-biological decay pro-
cesses and the uptake of pollutants by plants [109]. The components of a bioretention area
include a grass buffer strip, planting soil, plant material, a ponding area with surface mulch,
an underground sand bed, an organic layer and infiltration chambers [110].
Daly et al. (2011) [13] tracked the water balance of a biofilter by considering its inflow
variability, filter media and vegetation type. An analytical model for the long-term PDF
of soil moisture content of the filter, and the statistics of outflow, evapotranspiration and
overflow, were derived. The total nitrogen removal performance was also estimated from
the model. The results of the analytical model were tested with real data collected from a
biofilter in Malborne, Australia, and it was shown that the model could be used to assess
the performance of biofilters across different climates.
Zhang and Guo (2014) [111] modeled runoff from both pervious and impervious
urban surfaces to develop closed-form APM expressions for the stormwater runoff capture
efficiency of bioretention cells. The results obtained were compared to those of an SWMM
continuous simulation model, and close agreement was observed, thus validating the
APM expression. However, some assumptions were made regarding the amount of water
present in the cells before any rainfall event, which need to be considered in extreme
cases. Accordingly, Guo et al. (2020) [112] refined the work of [111] to consider wider
ranges of application. A dynamic water balance was considered to stochastically model the
hydrology of bioretention cells. Analytical expressions for the long-term runoff capture
efficiency, the fraction of time for which the cell processes runoff, the average water stored
inside a cell, and the storage capacity required to achieve capture efficiency were derived.
These four performance indices generated by the ASM were compared with the results of
continuous simulation, and close agreements were obtained, thus verifying the applicability
of the ASM.
The resilience and reliability of using bioretention cells as runoff control systems was
studied by [113]. APM expressions were used to evaluate the resilience indices related
to the system’s robustness, rapidity and serviceability in the context of extreme runoff
events. The results of the APM were compared with those generated using the continuous
simulation SWMM. Resilience indices of 0.66 to 1.0 and 0.73 to 1.0, respectively, were
observed for the APM and SWMM. The reliability index found ranged from 56% to 100%
and 60% to 100% for the APM and SWMM, respectively.
Impervious area disconnection is a system that works in a similar way to bioretention
cells. Runoff from urban surfaces (roof tops, pavements) is made to pass through pervious
surfaces (grassed area), where processes such as infiltration and pollutant removal occur,
thus reducing the volume of surface runoff. The time of concentration in the catchment
is also reduced, thereby reducing the peak discharge from the catchment. Wang et al.
(2019) [81] determined the effect of impervious area disconnection on runoff reduction from
two urban catchments in the USA. Two different catchments’ soils (sandy and loamy) were
used. The runoff reduction due to impervious area disconnection was examined using
different imperviousness ratios. The results of the APM and SWMM were compared, and
in all cases, impervious area disconnection was found to significantly reduce the volume
of runoff to the sewer system, and the APM results compared very well with those of the
SWMM. Zhang and Guo (2013b) [80] studied the hydrologic operations of a rain garden
to derive analytical expressions for its long-term runoff capture efficiency. The results
from the APM model were compared with those of SWMM simulations, and a very good
agreement between the APM and continuous simulation results was observed. The APM
was applied to rain gardens in Atlanta and Flagstaff in USA to demonstrate the sensitivity
of runoff capture efficiency to specific model parameters.
Water 2023, 15, 1640 15 of 23

4.4.2. Infiltration Trenches/Basins

Infiltration trenches are rectangular excavations with void-forming materials, such as
gravel aggregates, which are designed to receive, filter, store and infiltrate urban stormwater.
They aid in reducing urban runoff and improving groundwater recharge. They also assist
in sediment and heavy metal removal from stormwater [114]. Guo and Gao (2016) [115] de-
rived analytical expression for the total annual overflow volume and total runoff reduction
rate of infiltration basins. The results of the APM were compared with those of SWMM
continuous simulation using rainfall data from Atlanta and Billings (USA), and the results
were found to be consistent, with a relative difference of less than 10%. Guo and Guo
(2018b) [116] derived APM expressions for the overflow frequency and stormwater capture
efficiency of non-vegetated infiltration facilities, such as infiltration trenches, infiltration
chambers, dry wells, etc. In deriving the expressions, infiltration was assumed to occur at
the bottom only. The results from the analytical expressions were compared with those of
SWMM simulations in relation to a case study of a catchment in Concord, New Hampshire
(USA), using sandy and loamy soils, and the two sets were found to be in good agreement.
The average absolute difference and average relative difference between the APM and
SWMM were found to be 0.04% and 5%, respectively. Wang and Guo (2020a) [117] analyzed
the water balance of infiltration-based BMPs by considering infiltration through their sides
and bottom, in an attempt to overcome the shortcomings of [108]. The mean degree of
saturation and mean runoff capture efficiency were derived, and the results of the analytical
model were compared with those of SWMM. Two soils, sandy and loamy, were used, and
the rainfall data from two climate conditions (Billings and Jackson) were used to develop
the APM model. The results were found to be reasonably comparable, with the largest
absolute relative difference being less than 10%.
Following the design guidelines released by the Atlanta and New Durham authorities,
Wang and Guo (2020b) [118] applied the analytical models they had developed earlier
in [117] to a practical design analysis of infiltration trenches. Runoff values, generated using
rainfall data from hypothetical catchments, in the two locations were assumed, and the
performance of the trench was assessed as a function of its soil type, footprint dimensions,
drain time and infiltration conditions. The results of the runoff reduction ratio indicate
that the conditions of infiltration through the sides, the bottom, and both combined have
profound effects on the runoff reduction ratio, by up to 15%. The runoff reduction ratio
was found to be most highly affected by changes in soil type and trench dimensions.

4.4.3. Permeable Pavements

Pervious pavements consist of pavements made with porous blocks or porous asphalt
that permits water to infiltrate. Pervious pavements may also be made from impervious
blocks that are fitted in such a way that water can pass between them. They can be used in
road surfaces with light traffic or in car parks. The infiltration rate through the pavement
may be as high as 1000 mm/h in new developments, although this value may reduce to 10%
of the original value over the lifetime of the pavement [119]. Zhang and Guo (2015) [120]
derived analytical expressions for the runoff capture efficiency of a permeable pavement
as an LID system to mitigate the impact of urban stormwater. SWMM simulations were
run on the modeled pavement in order to validate the APM expressions, and the results
showed little discrepancy. It is recommended that the APM results be compared with those
of a case study on real-life pavements.
Stochastic differential equations of permeable pavements were used by Guo et al.
(2018) [121] to model the dynamic water balance of their system. Rainfall and the corre-
sponding net inflow were represented as a marked Poisson process to develop the PDF of
inflow volume, and to derive analytical expressions for the long-term stormwater capture
efficiency and moisture content of permeable pavements. The results of the APM were
compared with those simulated using SWMM using data from the four climates of Atlanta,
New Durham, Charlotte and Flagstaff (USA), and were found to be very similar.
Water 2023, 15, 1640 16 of 23

Three runoff control systems—bioretention cell, permeable pavement and green

roofs—were compared to determine the most cost-effective. The runoff reduction efficien-
cies and licecycle costs of implementing each of them were considered. APM expressions
were combined with a genetic algorithm for the optimization. The objective function was
to maximize runoff reduction capacity and minimize the lifecycle cost. The results show
that the bioretention cell had the greatest runoff reduction capability, but given the high
land cost in urbanized areas, permeable pavements are the most reasonable option [122].

4.5. Vegetated Open Channel Practices

These are systems designed to treat stormwater runoff in a swale or channel formed
by check dams or other processes. Usually, they do not allow for quantity control, and
are therefore combined with other stormwater BMPs to meet regulations. These systems
directly receive runoff from an impervious surface; they have a temporary ponding time of
less than 48 h and feature a 6-inch drop onto a protected shelf to minimize the clogging
potential of the inlet [4]. Up to the time of submitting this review, no publications have
been found on the application of APMs to vegetated open channel practices. This issue can
be explored by future researchers. Two of the common types of vegetated open channel
practices include grassed swales (dry/wet) and grassed channels.
According to [123], grassed swales are broad, shallow earthen channels used to treat
stormwater runoff using flood-tolerant and erosion-resistant grasses. Filtering via these
practices occurs through the vegetation, a subsoil matrix, and infiltration into the underlying
soils. Grassed swales feature gentle longitudinal slopes, with check dams perpendicular to
the flow so as to slow down the flows and allow the particulates to settle. There are two
types of grass swales—dry swales, which have a filter bed of prepared soil laid over an
under-drain system, and wet swales, which are designed to sustain moisture conditions
that support wetland vegetation [124].
Grassed channels are used in pretreatment practices that provide nominal treatment,
because they lack the filter media present in grassed swales. They act by allowing the
infiltration of some runoff from small storms into areas with pervious soils, and are therefore
most highly applicable to other structural stormwater BMPs [123]. They help in reducing
the effect of imperviousness, and provide aesthetic benefits. Grassed channels are designed
for use on <4% flat slopes with infiltration rates greater than 0.27 inches per hour. The
stormwater runoff takes an average of 5 min to flow from the top to the bottom of the
channel. For efficient usage, the channel should be used to treat small drainage areas of
less than 5 acres. For the effective removal of particles, the grass of the channel should be
maintained at a height of 3 to 4 inches [4].

4.6. Other Stormwater BMPs

Other types of stormwater BMPs that are used to control urban stormwater runoff
include: constructed wetlands, dry wells, artificial marshes, oil/greet separators, catch
basins, etc. [119,125]. Their effectiveness can be represented via a decrease in the SCS curve
number of the basins. Perez-Pedini et al. (2005) [126] determined the optimal number
and location of infiltration facilities in a watershed for the purpose of peak flow reduction
at the watershed outlet. The watershed was discretized into 4553 hydrologic response
units, whereby each unit represents a 120 × 120 m plot of the watershed. Different types
of infiltration-based BMPs were conceptualized as binary integers that decrease the curve
number of hydrologic response unit by five. The results of the optimization show that
the optimal number and location of infiltration-based BMPs depends on various factors,
such as flow travel time, catchment network connectivity, land-use, contributing area, and
distance to the channel. APMs of stormwater management can be applied to constructed
wetlands, dry wells, artificial marshes, catch basins, etc., in future studies.
Water 2023, 15, 1640 17 of 23

5. Recommendations for Future Direction

Urban stormwater raises flood and water pollution problems for many communities
across the world, and the cost of the damage cannot be easily quantified. This paper has
reviewed the literature on APM applications in urban stormwater management. Once
derived, the APM models can easily be co-opted into any optimization frameworks, thereby
giving the freedom to maximize benefits and minimize cost. The following recommenda-
tions are given:
(a) The APM parameters were obtained from analyses of the long-term data on rainfall
depth, duration and inter-event time. To make them more applicable, it is necessary
to develop a comprehensive database of APM parameters describing rainfall charac-
teristics in cities across the world, for the purpose of runoff control systems design;
(b) Most rain-gauge stations, particularly in developing countries, record daily rainfall
only. Urban catchments have shorter times of concentration, and studies in these parts
of the world have to rely on rainfall disaggregation techniques, whereby daily rainfall
is broken down to an hourly or even sub-hourly time scale, which may raise some
reliability problems. There is a need for a database of finer-resolution rainfall data.
The provision of a large network of hourly and sub-hourly rain-gauge data will not
only be useful to urban hydrologists, but also to other professionals. It will also help
in reducing the uncertainties caused by rainfall disaggregation. Another source of
uncertainty is the spatial distribution of rain-gauge stations used to develop rainfall
characteristics and APM parameters. Research is required into the effects of the spatial
distribution of rain-gauge stations on the reliability of rainfall characteristics;
(c) The APM parameters were derived based on minimum inter-event times of 2 h, 6 h,
12 h and 24 h. In the case of small urban catchments, with faster concentration, it
is recommended that a database of APM parameters based on a smaller discretized
inter-event time, such as 5 min, 15 min, 30 min or 1 h, be developed. This requires
the archiving of rainfall data at a sub-hourly resolution, which could then be used to
develop its own database;
(d) There is uncertainty about the inter-event time value to be used in rainfall event aggre-
gation from a continuous time series. This calls for further research on its reliability;
(e) The APMs are mostly based on the exponential distribution of rainfall characteristics.
Rainfall characteristics were also found to follow other distributions, such as Gamma,
Weibull, and log-normal. A distribution fit test for other PDFs needs to be undertaken
in different climates;
(f) A decision support system that incorporates meteorological, catchment and runoff
control systems’ characteristics altogether needs to be developed, which can then
eventually be used in the design and real-time control of these systems;
(g) The design of some systems, such as rainwater tanks, involves the consideration of
rainfall variability vis-à-vis water demand and the cost of municipal water consump-
tion. Likewise, designing detention ponds for runoff quantity and pollution control
involves conflicting objectives. There is a need for studies that embed APMs into
optimization techniques so as to derive optimum benefits from the runoff control
systems at the least cost;
(h) Climate change is known to affect the design of stormwater conveyance and storage
systems. There is a research gap regarding the effect of climate change on runoff
control systems designed using APMs. The impact of climate change on the reliability
of the systems needs to be investigated, so as to ensure their design functions are met;
(i) There is a research gap regarding the APMs related to the runoff reduction efficiency
and pollution control performance of vegetated open channel technologies, such as
swales, grass channel, etc.;
(j) Different runoff control systems have been reviewed in this paper. Some systems may
be more suitable to specific climates. There is a research gap in the determination of
the best system for each specific geographical area.
Water 2023, 15, 1640 18 of 23

6. Conclusions
Urban stormwater runoff is detrimental to downstream drainage systems and to
receiving water bodies. The risks range from flooding to water pollution. This paper has
reviewed the literature on runoff control systems, such as detention basins, rain gardens,
rainwater harvesting system, bioretention cells, pervious pavements, infiltration trenches,
etc. The design of runoff control systems can be carried out using the traditional design
storm approach, continuous simulation and APMs. The major flaw of the design storm
approach is its inability to capture the effects of inter-event time. That is, the design storm
assumes that the recurrence interval of runoff is the same as that of the rainfall that causes it.
The continuous simulation approach, on the other hand, is laborious and time-consuming,
thus making it unsuitable for use at the planning stage of a runoff control project. APMs,
however, are more compact, easy to use, and offer a direct way to conduct sensitivity
analyses in routine planning projects. Moreover, APMs are flexible and can be co-opted
into an optimization framework. Despite their simplicity, the APMs provide results that
are as accurate as those of continuous simulation. This paper offers an extensive review of
the applications of APMs to urban stormwater management.

Author Contributions: S.D. and A.A. jointly contributed to the production of this review paper. All
authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by Prince Sattam bin Abdulaziz University project number
PSAU/2022/01/23514.
Acknowledgments: The authors extend their appreciation to Prince Sattam bin Abdulaziz Univer-
sity for funding this research work through the project number PSAU/2022/01/23514. We also
acknowledge the inputs of anonymous reviewers.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or
in the decision to publish the results.

Abbreviations

ACRU Agricultural Catchments Research Unit

APM Analytical Probabilistic Models
ASM Analytical Stochastic Models
ASTORM Analytical STORM
BMPs Best Management Practices
CDF Cumulative Distribution Function
COD Chemical Oxygen Demand
DP Dynamic Programming
EMC Event Mean Concentration
EX-SUDS Extended Sustainable Urban Drainage System
GA Genetic Algorithm
HEC-HMS Hydrologic Engineering Center’s Hydrologic Modeling System
IDF Intensity–Duration–Frequency
MIDUSS Microcomputer Interactive Design of Urban Stormwater Drainage Systems
NLP Non-Linear Programming
NSE Nash–Sutcliff Efficiency
OTTHYMO Ottawa Hydrological Model
PC-SWMM Personal Computer–Storm Water Management Model
PDF Probability Density Function
PSO Particle Swam Optimization
RMSE Root Mean Square Error
SCN-CN Soil Conservation Service–Curve Number
STORM Stormwater Management Software
SUDS Sustainable Urban Drainage system
SWMM Storm Water Management Model
Water 2023, 15, 1640 19 of 23

References
1. Steg, R. Summary of the Stormwater Sources in the Flatheadd Lake Basin; Montana Operations Office, USEPA: Montana, WY, USA, 2010.
2. McCuen, R.H. Hydrologic Analysis and Design; Prentice Hall: Upper Saddle River, NJ, USA, 2004.
3. Birkinshaw, S.J.; Kilsby, C.; O’donnell, G.; Quinn, P.; Adams, R.; Wilkinson, M.E. Stormwater Detention Ponds in Urban
Catchments—Analysis and Validation of Performance of Ponds in the Ouseburn Catchment, Newcastle upon Tyne, UK. Water
2021, 13, 2521. [CrossRef]
4. GSMM. General Application Structural Stormwater Controls. Georgia Stormwater Management Manual, Vol. 2. Atlanta Re-gional
Commission, GA, 2016. Available online: http://www.georgiastormwater.com/ (accessed on 3 October 2022).
5. Guo, Y. Hydrologic Design of Urban Flood Control Detention Ponds. J. Hydrol. Eng. 2001, 6, 472–479. [CrossRef]
6. Methods, H.; Durrans, S.R. Stormwater Conveyance Modeling and Design; Bently Institute Press: Exton, PA, USA, 2001.
7. Jia, H.; Hu, J.; Huang, T.; Chen, A.S.; Ma, Y. Urban Runoff Control and Sponge City Construction. Water 2022, 14, 1910. [CrossRef]
8. Jeng, H.A.C.; Englande, A.J.; Bakeer, R.M.; Bradford, H.B. Impact of urban stormwater runoff on estuarine environmental quality.
Estuar. Coast. Shelf Sci. 2005, 63, 513–526. [CrossRef]
9. Todeschini, S. Hydrologic and Environmental Impacts of Imperviousness in an Industrial Catchment of Northern Italy. J. Hydrol.
Eng. 2016, 21, 05016013. [CrossRef]
10. Schütte, S.; Schulze, R. Projected impacts of urbanisation on hydrological resource flows: A case study within the uMngeni
Catchment, South Africa. J. Environ. Manag. 2017, 196, 527–543. [CrossRef] [PubMed]
11. USEPA. Preliminary Data Summary of Urban Stormwater Best Management Practices; Report No EPA-821-R-99-012; United States
Environmental Protection Agency: Washington, DC, USA, 1999.
12. Sharifi, S.; Massoudieh, A.; Kayhanian, M. A Stochastic Stormwater Quality Volume-Sizing Method with First Flush Emphasis.
Water Environ. Res. 2011, 83, 2025–2035. [CrossRef]
13. Daly, E.; Deletic, A.; Hatt, B.E.; Fletcher, T.D. Modelling of stormwater biofilters under random hydrologic variability: A case
study of a car park at Monash University, Victoria (Australia). Hydrol. Process. 2011, 26, 3416–3424. [CrossRef]
14. Li, Y.; Babcock, R.W. Green roof hydrologic performance and modeling: A review. Water Sci. Technol. 2013, 69, 727–738. [CrossRef]
15. Krasnogorskaya, N.; Longobardi, A.; Mobilia, M.; Khasanova, L.F.; Shchelchkova, A.I. Hydrological Modeling of Green Roofs
Runoff by Nash Cascade Model. Open Civ. Eng. J. 2019, 13, 163–171. [CrossRef]
16. Berndtsson, J.C. Green roof performance towards management of runoff water quantity and quality: A review. Ecol. Eng. 2010,
36, 351–360. [CrossRef]
17. Qi, W.; Ma, C.; Xu, H.; Chen, Z.; Zhao, K.; Han, H. A review on applications of urban flood models in flood mitigation strategies.
Nat. Hazards 2021, 108, 31–62. [CrossRef]
18. Ellis, J. Sustainable surface water management and green infrastructure in UK urban catchment planning. J. Environ. Plan. Manag.
2013, 56, 24–41. [CrossRef]
19. Maes, J. and Jacobs, S. Nature-based solutions for Europe’s sustainable development. Conserv. Lett. 2017, 10, 121–124. [CrossRef]
20. O’Donnell, E.; Thorne, C.; Ahilan, S.; Arthur, S.; Birkinshaw, S.; Butler, D.; Dawson, D.; Everett, G.; Fenner, R.; Glenis, V.; et al. The
blue-green path to urban flood resilience. Blue-Green Syst. 2019, 2, 28–45. [CrossRef]
21. Yin, D.; Xu, C.; Jia, H.; Yang, Y.; Sun, C.; Wang, Q.; Liu, S. Sponge City Practices in China: From Pilot Exploration to Systemic
Demonstration. Water 2022, 14, 1531. [CrossRef]
22. Jia, H.; Wang, Z.; Zhen, X.; Clar, M.; Yu, S.L. China’s sponge city construction: A discussion on technical approaches. Front.
Environ. Sci. Eng. 2017, 11, 18. [CrossRef]
23. Debo, T.N.; Reese, A.J. Municipal Stormwater Management; Lewis Publishers: Boca Raton, FL, USA, 2003.
24. Chen, J.; Adams, B.J. A derived probability distribution approach to stormwater quality modeling. Adv. Water Resour. 2007,
30, 80–100. [CrossRef]
25. Adams, B.J.; Papa, F. Urban Stormwater Management Planning with Analytical Probabilistic Models; John Wiley & Sons: New York,
NY, USA, 2000.
26. Chen, J.; Adams, B.J. Development of analytical models for estimation of urban stormwater runoff. J. Hydrol. 2007, 336, 458–469.
[CrossRef]
27. Todeschini, S.; Papiri, S.; Ciaponi, C. Performance of stormwater detention tanks for urban drainage systems in northern Italy.
J. Environ. Manag. 2012, 101, 33–45. [CrossRef]
28. Mobley, J.T.; Culver, T.B. Design of outlet control structures for ecological detention ponds. J. Water Res. Plan. Manag. ASCE 2014,
140, 250–257. [CrossRef]
29. Souza, F.P.; Costa, M.E.L.; Koide, S. Hydrological Modelling and Evaluation of Detention Ponds to Improve Urban Drainage
System and Water Quality. Water 2019, 11, 1547. [CrossRef]
30. Mielke, P.W.; Benjamin, J.R.; Cornell, C.A. Probability, Statistics and Decision for Civil Engineers. J. Am. Stat. Assoc. 1971, 66, 923.
[CrossRef]
31. Eagleson, P.S. Dynamics of flood frequency. Water Resour. Res. 1972, 8, 878–898. [CrossRef]
32. Adams, B.J.; Fraser, H.G.; Howard, C.D.D.; Hanafy, M.S. Meteorological Data Analysis for Drainage System Design. J. Environ.
Eng. 1986, 112, 827–848. [CrossRef]
33. Bacchi, B.; Balistrocchi, M.; Grossi, G. Proposal of a semi-probabilistic approach for storage facility design. Urban Water J. 2008,
5, 195–208. [CrossRef]
Water 2023, 15, 1640 20 of 23

34. Branham, T.L.; Behera, P.K. Development of a Rainfall Statistical Analysis Tool for Analytical Probabilistic Models for Urban
Stormwater Management Analysis. In Proceedings of the World Environmental and Water Resources Congress, Providence, RI,
USA, 16–20 May 2010. [CrossRef]
35. Dan’azumi, S.; Shamsudin, S.; Aris, A. Development of analytical probabilistic model parameters for urban stormwater manage-
ment. Sains Malays. 2013, 42, 325–332.
36. Guo, Y.; Adams, B.J. Hydrologic analysis of urban catchment with event based probabilistic models, Runoff volume. Water Resour.
Res. 1998, 34, 3421–3432. [CrossRef]
37. Guo, Y.; Adams, B.J. Hydrologic analysis of urban catchment with event based probabilistic models, 2. Peak discharge rate. Water
Resour. Res. 1998, 34, 3433–3443. [CrossRef]
38. Chen, J.; Adams, B.J. Urban Storm Water Control Evaluation with Analytical Probabilistic Models. J. Water Resour. Plan. Manag.
2005, 131, 362–374. [CrossRef]
39. Chen, J.; Adams, B.J. Analysis of storage facilities for urban stormwater quantity control. Adv. Water Resour. 2005, 28, 377–392.
[CrossRef]
40. Chen, J.; Adams, B.J. Analytical urban storm water quality models based on pollutant buildup and wash-off processes. J. Environ.
Eng. ASCE 2006, 132, 1314–1330. [CrossRef]
41. Chen, J.; Adams, B.J. Urban stormwater quality control analysis with detention ponds. Water Environ. Res. 2006, 78, 744–753.
[CrossRef] [PubMed]
42. Guo, Y.; Adams, B.J. An analytical probabilistic approach to sizing flood control detention facilities. Water Resour. Res. 1999, 35,
2457–2468. [CrossRef]
43. Guo, Y.; Adams, B.J. Analysis of detention ponds for stormwater quality control. Water Resour. Res. 1999, 35, 2447–2456. [CrossRef]
44. Li, J.Y.; Adams, B.J. Probabilistic Models for Analysis of Urban Runoff Control Systems. J. Environ. Eng. 2000, 126, 217–224.
[CrossRef]
45. Behera, P.K.; Adams, B.J.; Li, J.Y. Runoff Quality Analysis of Urban Catchments with Analytical Probabilistic Models. J. Water
Resour. Plan. Manag. 2006, 132, 4–14. [CrossRef]
46. Quader, A.; Guo, Y. Peak Discharge Estimation Using Analytical Probabilistic and Design Storm Approaches. J. Hydrol. Eng. 2006,
11, 46–54. [CrossRef]
47. Hassini, S.; Guo, Y. Exponentiality Test Procedures for Large Samples of Rainfall Event Characteristics. J. Hydrol. Eng. 2016, 21,
04016003. [CrossRef]
48. Balistrocchi, M.; Grossi, G.; Bacchi, B. An analytical probabilistic model of the quality efficiency of a sewer tank. Water Resour. Res.
2009, 45, W12420. [CrossRef]
49. Palynchuk, B.; Guo, Y. Threshold analysis of rainstorm depth and duration statistics at Toronto, Canada. J. Hydrol. 2008, 348,
535–545. [CrossRef]
50. Andrés-Doménech, I.; Montanari, A.; Marco, J.B. Stochastic rainfall analysis for storm tank performance evaluation. Hydrol. Earth
Syst. Sci. 2010, 14, 1221–1232. [CrossRef]
51. Szelag,
˛ B.; Suligowski, R.; Studziński, J.; De Paola, F. Application of logistic regression to simulate the influence of rainfall genesis
on storm overflow operations: A probabilistic approach. Hydrol. Earth Syst. Sci. 2020, 24, 595–614. [CrossRef]
52. Lee, M.; An, H.; Jeon, S.; Kim, S.; Jung, K.; Park, D. Development of an analytical probabilistic model to estimate runoff event
volumes in South Korea. J. Hydrol. 2022, 612, 128129. [CrossRef]
53. Kim, S.; Han, S. Urban Stormwater Capture Curve Using Three-Parameter Mixed Exponential Probability Density Function and
NRCS Runoff Curve Number Method. Water Environ. Res. 2010, 82, 43–50. [CrossRef]
54. Guo, Y.; Liu, S.; Baetz, B.W. Probabilistic rainfall-runoff transformation considering both infiltration and saturation excess runoff
generation processes. Water Resour. Res. 2012, 48, W06513. [CrossRef]
55. Hassini, S.; Guo, Y. Derived flood frequency distributions considering individual event hydrograph shapes. J. Hydrol. 2017,
547, 296–308. [CrossRef]
56. Hassini, S.; Guo, Y. Analytical Hydrological Model for the Planning and Design of Low-Impact Development Practices. In Recent
Advances in Environmental Science from the Euro-Mediterranean and Surrounding Regions, 2nd ed.; EMCEI 2019. Environmental
Science and Engineering; Springer: Cham, Switzerland, 2021. [CrossRef]
57. Hassini, S.; Guo, Y. Analytical Derivation of Urban Flood Frequency Models Accounting Saturation-Excess Runoff Generation.
J. Hydrol. 2020, 584, 124713. [CrossRef]
58. Hassini, S.; Guo, Y. Analytical Derivation of Urban Runoff-Volume Frequency Models. J. Sustain. Water Built Environ. 2022, 8,
04021022. [CrossRef]
59. Guo, Y.; Zhuge, Z. Analytical probabilistic flood routing for urban stormwater management purposes. Can. J. Civ. Eng. 2008,
35, 487–499. [CrossRef]
60. Guo, Y.; Hansen, D.; Li, C. Probabilistic approach to estimating the effects of channel reaches on flood frequencies. Water Resour.
Res. 2009, 45, W08404. [CrossRef]
61. Guo, Y.; Markus, M. Analytical Probabilistic Approach for Estimating Design Flood Peaks of Small Watersheds. J. Hydrol. Eng.
2011, 16, 847–857. [CrossRef]
62. Guo, Y.; Dai, J. Expanded analytical probabilistic stormwater models for use in watershed and master drainage planning. Can. J.
Civ. Eng. 2009, 36, 933–943. [CrossRef]
Water 2023, 15, 1640 21 of 23

63. Shishegar, S.; Duchesne, S.; Pelletier, G. Optimization methods applied to stormwater management problems: A review. Urban
Water J. 2018, 15, 276–286. [CrossRef]
64. Gayer, J. Managing Urban Runoff in the Light of Integrated Water Resources Management. Ph.D. Thesis, Faculty of Water
Management and Melioration, Corvinus University of Budapest, Budapest, Hungary, 2004.
65. Drake, J.; Young, D.; McIntosh, N. Performance of an Underground Stormwater Detention Chamber and Comparison with
Stormwater Management Ponds. Water 2016, 8, 211. [CrossRef]
66. Chin, D. Water Resources Engineering; Prentice Hall: Hoboken, NJ, USA, 2006.
67. Shishegar, S.; Duchesne, S.; Pelletier, G. An integrated optimization and rule-based approach for predictive real time control of
urban stormwater management systems. J. Hydrol. 2019, 577, 124000. [CrossRef]
68. Hogan, D.M.; Walbridge, M.R. Best Management Practices for Nutrient and Sediment Retention in Urban Stormwater Runoff.
J. Environ. Qual. 2007, 36, 386–395. [CrossRef]
69. Papa, F.; Adams, B.J.; Guo, Y. Detention time selection for stormwater quality control ponds. Can. J. Civ. Eng. 1999, 26, 72–82.
[CrossRef]
70. Wang, G.-T.; Chen, S.; Barber, M.E.; Yonge, D.R. Modeling Flow and Pollutant Removal of Wet Detention Pond Treating
Stormwater Runoff. J. Environ. Eng. 2004, 130, 1315–1321. [CrossRef]
71. Papa, F.; Adams, B.J.; Bryant, G.J. Models for Wwater Qquality Ccontrol by Sstonnwater Pponds. J. Water Manag. Model. 1997, 5,
R195-01. [CrossRef]
72. Becciu, G.; Raimondi, A. Probabilistic analysis of spills from stormwater detention facilities. Urban Water II 2014, 139, 159–170.
[CrossRef]
73. Raimondi, A.; Becciu, G. On pre-filling probability of flood control detention facilities. Urban Water J. 2013, 12, 344–351. [CrossRef]
74. Becciu, G.; Raimondi, A. Probabilistic Analysis of the Retention Time in Stormwater Detention Facilities. Procedia Eng. 2015, 119,
1299–1307. [CrossRef]
75. Becciu, G.; Raimondi, A.; Brebbia, C. Probabilistic modeling of the efficiency of a stormwater detention facility. Int. J. Sustain. Dev.
Plan. 2015, 10, 1–11. [CrossRef]
76. Raimondi, A.; Di Chiano, M.G.; Marchioni, M.; Sanfilippo, U.; Becciu, G. Probabilistic modeling of sustainable urban drainage
systems. Urban Ecosyst. 2022, 1–10. [CrossRef]
77. Ross, S.M. Introduction to Probability Models; Academic Press: New York, NY, USA, 2007.
78. Guo, Y.; Baetz, B.W. Sizing of Rainwater Storage Units for Green Building Applications. J. Hydrol. Eng. 2007, 12, 197–205.
[CrossRef]
79. Zhang, S.; Guo, Y. Analytical Probabilistic Model for Evaluating the Hydrologic Performance of Green Roofs. J. Hydrol. Eng. 2013,
18, 19–28. [CrossRef]
80. Zhang, S.; Guo, Y. Explicit Equation for Estimating Storm-Water Capture Efficiency of Rain Gardens. J. Hydrol. Eng. 2013, 18,
1739–1748. [CrossRef]
81. Wang, J.; Guo, Y. Stochastic analysis of storm water quality control detention ponds. J. Hydrol. 2019, 571, 573–584. [CrossRef]
82. Parolari, A.J.; Pelrine, S.; Bartlett, M.S. Stochastic water balance dynamics of passive and controlled stormwater basins. Adv.
Water Resour. 2018, 122, 328–339. [CrossRef]
83. Becciu, G.; Raimondi, A. Factors affecting the pre-filling probability of water storage tanks. WIT Trans. Ecol. Environ. 2012, 164,
473–484. [CrossRef]
84. De Paola, F.; Ranucci, A. Analysis of Spatial Variability for Stormwater Capture Tank Assessment. Irrig. Drain. 2012, 61, 682–690.
[CrossRef]
85. Papa, F.; Adams, B.J. Application of derived probability and dynamic programming techniques to planning regional storm-water
management systems. Water Sci. Technol. 1997, 36, 227–234. [CrossRef]
86. Shamsudin, S.; Dan’azumi, S.; Aris, A.; Yusop, Z. Optimum combination of pond volume and outlet capacity of a storm-water
detention pond using particle swarm optimization. Urban Water J. 2014, 11, 127–136. [CrossRef]
87. Dan’Azumi, S.; Shamsudin, S.; Aris, A. Optimization of pollution control performance of wet detention ponds in tropical urban
catchments using particle swarm optimization. J. Hydroinform. 2012, 15, 529–539. [CrossRef]
88. Behera, P.K.; Teegavarapu, R.S.V. Optimization of a Stormwater Quality Management Pond System. Water Resour. Manag. 2014,
29, 1083–1095. [CrossRef]
89. Raimondi, A.; Becciu, G. Probabilistic Modeling of Rainwater Tanks. Procedia Eng. 2014, 89, 1493–1499. [CrossRef]
90. Raimondi, A.; Becciu, G. Probabilistic Design of Multi-use Rainwater Tanks. Procedia Eng. 2014, 70, 1391–1400. [CrossRef]
91. Becciu, G.; Raimondi, A.; Dresti, C. Semi-probabilistic design of rainwater tanks: A case study in Northern Italy. Urban Water J.
2016, 15, 192–199. [CrossRef]
92. De Paola, F.; De Martino, F. Stormwater Tank Performance: Design and Management Criteria for Capture Tanks Using a
Continuous Simulation and a Semi-Probabilistic Analytical Approach. Water 2013, 5, 1699–1711. [CrossRef]
93. Kim, H.; Han, M.; Lee, J.Y. The application of an analytical probabilistic model for estimating the rainfall–runoff reductions
achieved using a rainwater harvesting system. Sci. Total. Environ. 2012, 424, 213–218. [CrossRef] [PubMed]
94. Di Chiano, M.G.; Marchioni, M.; Raimondi, A.; Sanfilippo, U.; Becciu, G. Probabilistic Approach to Tank Design in Rainwater
Harvesting Systems. Hydrology 2023, 10, 59. [CrossRef]
Water 2023, 15, 1640 22 of 23

95. Guo, R.; Guo, Y. Stochastic modelling of the hydrologic operation of rainwater harvesting systems. J. Hydrol. 2018, 562, 30–39.
[CrossRef]
96. Pelak, N.; Porporato, A. Sizing a rainwater harvesting cistern by minimizing costs. J. Hydrol. 2016, 541, 1340–1347. [CrossRef]
97. Sim, I.; Kim, S. Probabilistic solution to rainwater harvesting system and its impact on climate change. Desalination Water Treat.
2020, 200, 407–421. [CrossRef]
98. Cheng, G.; Huang, G.; Guo, Y.; Baetz, B.W.; Dong, C. Stochastic Rainwater Harvesting System Modeling Under Random Rainfall
Features and Variable Water Demands. Water Resour. Res. 2021, 57, e2021WR029731. [CrossRef]
99. USEPA. Using Green Roofs to Reduce Heat Islands. 2023. Available online: https://www.epa.gov/heatislands/using-green-
roofs-reduce-heat-islands (accessed on 14 March 2023).
100. Raimondi, A.; Becciu, G. Performance of Green Roofs for Rainwater Control. Water Resour. Manag. 2020, 35, 99–111. [CrossRef]
101. Guo, Y.; Zhang, S.; Liu, S. Runoff Reduction Capabilities and Irrigation Requirements of Green Roofs. Water Resour. Manag. 2014,
28, 1363–1378. [CrossRef]
102. Guo, Y. Stochastic Analysis of Hydrologic Operation of Green Roofs. J. Hydrol. Eng. 2016, 21, 04016016. [CrossRef]
103. Raimondi, A.; Marchioni, M.; Sanfilippo, U.; Stroppa, F.F.; Becciu, G. Probabilistic estimation of runoff from green roofs. Int. J.
Comput. Methods Exp. Meas. 2022, 10, 13–25. [CrossRef]
104. Raimondi, A.; Marchioni, M.; Sanfilippo, U.; Becciu, G. Vegetation Survival in Green Roofs without Irrigation. Water 2021, 13, 136.
[CrossRef]
105. Guo, R.; Guo, Y.; Liu, S. Analytical Equations for Direct Quantification of Green Roofs’ Hydrologic Performance Statistics.
J. Hydrol. Eng. 2022, 27, 04022003. [CrossRef]
106. Raimondi, A.; Sanfilippo, U.; Marchioni, M.; Di Chiano, M.; Becciu, G. Influence of climatic parameters on the probabilistic design
of green roofs. Sci. Total Environ. 2023, 865, 161291. [CrossRef]
107. FHWA. Stormwater Best Management Practices in Ultra-Urban Setting: Selection and Monitoring. Washington DC, Federal
Highway Administration, U.S. Department of Transportation, 2000. Available online: https://rosap.ntl.bts.gov/view/dot/25727.
(accessed on 12 November 2022).
108. Błażejewski, R.; Murat-Błażejewska, S. Analytical solutions of a routing problem for storm water in a detention basin. Hydro-Log.
Sci. J. 2003, 48, 665–671. [CrossRef]
109. Weiss, P.T.; Gulliver, J.; Erickson, A.J. Cost and Pollutant Removal of Storm-Water Treatment Practices. J. Water Resour. Plan.
Manag. 2007, 133, 218–229. [CrossRef]
110. VASM. Virginia Stormwater Handbook; Virginia Department of Conservation and Recreation, Division of Soil and Water Conser-
vation: Richmond, VA, USA, 1999; Volume II. Available online: http://observatoriaigua.uib.es/repositori/suds_virginiab.pdf
(accessed on 5 October 2022).
111. Zhang, S.; Guo, Y. Stormwater Capture Efficiency of Bioretention Systems. Water Resour. Manag. 2013, 28, 149–168. [CrossRef]
112. Guo, R.; Guo, Y.; Zhang, S.; Zhu, D.Z. A Tool for Water Balance Analysis of Bioretention Cells. J. Sustain. Water Built Environ.
2020, 6, 04020013. [CrossRef]
113. Islam, A.; Hassini, S.; El-Dakhakhni, W. Resilience Quantification of Low-Impact Development Systems Using SWMM and a
Probabilistic Approach. J. Sustain. Water Built Environ. 2022, 8, 04022013. [CrossRef]
114. Fach, S.; Dierkes, C. On-site infiltration of road runoff using pervious pavements with subjacent infiltration trenches as source
control strategy. Water Sci. Technol. 2011, 64, 1388–1397. [CrossRef] [PubMed]
115. Guo, Y.; Gao, T. Analytical Equations for Estimating the Total Runoff Reduction Efficiency of Infiltration Trenches. J. Sustain.
Water Built Environ. 2016, 2, 06016001. [CrossRef]
116. Guo, R.; Guo, Y. Analytical Equations for Use in the Planning of Infiltration Facilities. J. Sustain. Water Built Environ. 2018, 4,
06018001. [CrossRef]
117. Fadhelab, S.; Rico, M.A.; Hana, D. Sensitivity of peak flow to the change of rainfall temporal pattern due to warmer climate.
J. Hydrol. 2018, 560, 546–559. [CrossRef]
118. Wang, J.; Guo, Y. Proper Sizing of Infiltration Trenches Using Closed-Form Analytical Equations. Water Resour. Manag. 2020, 34,
1–13. [CrossRef]
119. Butler, D.; Davies, J.W. Urban Drainage, 2nd ed.; Spon Press, Tailor & Francis Group: New York, NY, USA, 2004.
120. Zhang, S.; Guo, Y. An analytical equation for evaluating the stormwater capture efficiency of permeable pavement systems.
J. Irrig. Drain. Eng. ASCE 2015, 131, 06014004-1–06014004-9.
121. Guo, R.; Guo, Y.; Wang, J. Stormwater capture and antecedent moisture characteristics of permeable pavements. Hydrol. Process.
2018, 32, 2708–2720. [CrossRef]
122. Zeng, J.; Huang, G.; Mai, Y.; Chen, W. Optimizing the cost-effectiveness of low impact development (LID) practices using an
analytical probabilistic approach. Urban Water J. 2020, 17, 136–143. [CrossRef]
123. SMRC. Assorted Fact Sheets: Stormwater Management Practices; Stormwater Manager’s Resource Center, Center for Watershed
Protection: Ellicot City, MD, USA, 2020; Available online: http://www.stormwatercenter.net (accessed on 5 July 2020).
124. USEPA. Free Water Surface Wetlands for Wastewater Treatments: A Technology Assessment; EPA 832 S 99 032; Office of Water, USEPA:
Washington, DC, USA, 1999. Available online: http://www.epa.gov/owow/wetlands/pdf/FW_Surface_Wetlands.pdf. (accessed
on 9 August 2020).
Water 2023, 15, 1640 23 of 23

125. AWWA. National Water Quality Management Strategy: Australian Guidelines for Urban Stormwater Management; Australia and
New-Zealand; Australian Water Works Association: St. Leonards, NSW, Australia, 2000.
126. Perez-Pedini, C.; Limbrunner, J.F.; Vogel, R.M. Optimal Location of Infiltration-Based Best Management Practices for Storm Water
Management. J. Water Resour. Plan. Manag. 2005, 131, 441–448. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.