Agricultural Water Management 171 (2016) 89–107

Contents lists available at ScienceDirect

Agricultural Water Management

journal homepage: www.elsevier.com/locate/agwat

A system dynamics based socio-hydrological model for agricultural

wastewater reuse at the watershed scale
Hanseok Jeong a , Jan Adamowski b,∗
a
Institute of Green Bio Science and Technology, Seoul National University, Gangwon 25354, Republic of Korea
b
Department of Bioresource Engineering, McGill University, 21 111 Lakeshore Road, Sainte-Anne-de-Bellevue, QC, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this study was to develop and verify a socio-hydrological model using system dynamics
Received 9 September 2015 (SD), thereby combining a deterministic conceptual hydrological model and a social model incorporating
Received in revised form 16 March 2016 population, land use, economics, technology, and policy dimensions. Applied to a central South Korean
Accepted 24 March 2016
watershed where wastewater is reused for paddy irrigation, the present model was verified in terms
of structure and behavior. Structural validity was confirmed when expected simulation sensitivity and
Keywords:
consistency criteria were met during behavior sensitivity and extreme conditions tests. The model’s
System dynamics modeling
behavioral validity in predicting hydrological processes including evapotranspiration, stream flow, and
Korea
Socio-hydrology
groundwater level, was also confirmed as the calibrated model performance during the validation period
Socio-hydrological modeling showed good agreement with those of the Soil and Water Assessment Tool (SWAT) model, validated
Urbanization for the study watershed, as well as observed groundwater levels. The values of Nash-Sutcliffe efficiency
Wastewater reuse (ENS ), percent bias (PBIAS), and R2 which compared model results with those of the SWAT model were
0.77, 3.0%, and 0.79, respectively, for the evapotranspiration, and 0.69, 1.4%, and 0.75, respectively, for the
stream flow, while the generated and observed groundwater levels exhibited a linear relationship with
an R2 value of 0.70. The validated model indicated that urbanization within the study watershed could
lead to increased stream flow and greater wastewater reuse. Instream flow regulation led to a decrease
in stream flow tied to a lower base flow, and a decrease in social benefits associated with a decline in
wastewater reuse. An assessment was made of the SD-based socio-hydrological model’s usefulness when
acting as an element of an integrated framework in providing a better understanding of small-scale socio-
hydrological systems’ interactions and the underlying causes of general trends and problems. SD-based
socio-hydrological modeling was deemed a suitable decision-support framework for designing water
resource policies contributing to successful integrated water resources management practice.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction
To ensure that agricultural systems address the future chal-
lenges of global food security and environmental sustainability,
present systems must be altered (Foley et al., 2011; Halbe et al.,
2014; Kolinjivadi et al., 2014). Wastewater reuse for agriculture
can provide a solution which addresses both challenges simulta-
neously, especially when water is scarce (Jang et al., 2010; Chavez
et al., 2012; Jeong et al., 2014). Largely irrigated with wastewa-
ter, agricultural lands produce foods consumed worldwide (World
Health Organization (WHO), 2006; Hamilton et al., 2007; Jimenez
et al., 2010). Produced at a relatively constant rate, treated wastew-
ater generated in Korea has abundant fertilizing agents and is
∗ Corresponding author.
E-mail address: jan.adamowski@mcgill.ca (J. Adamowski).
widely used as irrigation water (Jang et al., 2012). This wastew-
ater reuse is likely to increase over the coming decades (Jeong
et al., 2016). However, public health issues and environmental con-
cerns must be addressed for a safe and rational implementation of
wastewater reuse in agricultural settings. Conducting long-term
social and environmental assessments is vital to the implemen-
tation of successful and sustainable water resources policies. A
modeling approach can aid in such assessments (Adamowski et al.,
2010; Nalley et al., 2012; Nalley et al., 2013; Valipour, 2014a, 2015;
Yaeger et al., 2014; Butler and Adamowski, 2015; Inam et al., 2015).
Successfully employed to predict the expected responses of
hydrological systems to proposed water resource policies (Tuppad
et al., 2010; Noory et al., 2011; Park and Roesner, 2012; Haidary
et al., 2013; Rathinasamy et al., 2013; Belayneh et al., 2014),
hydrological models have proven effective in physically inter-
preting natural phenomena. However, traditional hydrological
models do not consider anthropogenic impacts on water cycle

http://dx.doi.org/10.1016/j.agwat.2016.03.019
0378-3774/© 2016 Elsevier B.V. All rights reserved.
90 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
dynamics (Wagener et al., 2010), thereby limiting their ability to
fully mimic the response of hydrological systems to evolving cir-
cumstances. This has led to an event-oriented modeling approach
based on linear thinking, largely ignoring interactions between
system components. To develop sustainable water resource man-
agement strategies and achieve long-term water security for
stakeholders and their environment, hydrological modeling must
undergo a paradigm shift to provide predictions which consider
whole system behavior (Wagener et al., 2010; Halbe et al., 2013;
Straith et al., 2014; Valipour, 2014b).
The new field of socio-hydrology considers humans and their
actions as an integral part of water cycle dynamics. Comprising
three principal domains — historical (Wescoat, 2013; Zlinszky and
Timar, 2013; Kandasamy et al., 2014; Liu et al., 2014), comparative
(Chang et al., 2014), and process (Di Baldassarre et al., 2013; Elshafei
et al., 2014; Viglione et al., 2014) — socio-hydrology attempts to
describe the dynamics and co-evolution of coupled human-water
systems. Whereas historical and comparative socio-hydrology are
somewhat qualitative, process socio-hydrology represents the
quantitative science required for modeling human-water systems
and predicting possible future trends in their states. Therefore, pro-
cess socio-hydrology can contribute to bringing about a paradigm
shift in hydrologic modeling, allowing for a better understanding
of how hydrological systems co-evolve with human activities.
Seeking to achieve greater insights into causal relationships
within human-water systems, process socio-hydrology employs
nonlinear, organic thinking (i.e., systems thinking; Sivapalan et al.,
2012). System dynamics (SD), a branch of systems thinking (Ford,
1999), was introduced by Forrester (1961) to support studies
related to representing complex systems and analyzing their
dynamic behavior (Hjorth and Bagheri, 2006). In a manner similar
to other traditional hydrological models, the use of SD in hydro-
logical modeling is focused on predictive simulations (Ahmad and
Simonovic, 2004; Elshorbagy and Ormsbee, 2006; Ghashghaei et al.,
2013). For example, Li and Simonovic (2002) established a concep-
tual hydrological model using SD to estimate snowmelt-derived
flooding. Likewise, Khan et al. (2009) demonstrated the applica-
tion of SD in simulating water balance in a rice paddy system,
as well as surface-groundwater dynamic interactions in an irri-
gation area. Descriptive integrated modeling coupled with an SD
approach has also served in multi-subsystem feedback modeling
of water resources systems for strategic policy testing and selec-
tion (Simonovic et al., 1997; Simonovic and Fahmy, 1999; Guo
et al., 2001; Stave, 2003; Madani and Marino, 2009; Ahmad and
Prashar, 2010; Qi and Chang, 2011; Venkatesan et al., 2011). How-
ever, these applications have not included a detailed analysis of
hydrological processes or focused on managing hydrological sys-
tems in such a manner as to implement integrated water resources
management (IWRM) through a scenario-based approach. Towards
these goals, Davies and Simonovic (2011) developed a SD-based
integrated model incorporating the global climate system, carbon
cycle, economy, population, land use, agriculture, novel versions
of the hydrological cycle, global water use, and water quality to
provide insights into the nature and structure of the connections
between water resources and socio-economic and environmen-
tal change. Similarly, Wei et al. (2012) developed a complex SD
model reflecting the interactions between water resources, envi-
ronmental flow, and socio-economic factors. They assessed the
socio-economic impacts caused by allocating different levels of
environmental flow in China’s Weihe River Basin. In the case of
wastewater reuse issues, Venkatesan et al. (2011) used SD to eval-
uate the impacts on water quality and energy consumption of
indirect and direct potable water reuse as water resources manage-
ment options. Nasiri et al. (2013) created an SD model with different
hypothetical levels of water reuse implementation in order to
simulate and optimize the overall water system cost, accounting
for future scenarios of population, economic growth and climate
change. These studies mainly focused on evaluating the effects of
water resources policies on a targeted system, whereas traditional
hydrological modeling of wastewater reuse have focused on simu-
lating the relevant hydrological process (Kim et al., 2008; Lee et al.,
2010; Jeong et al., 2016).
Despite the number of studies on the subject of hydrological
modeling and watershed management using SD, few SD modeling
studies have sought to identify the interactions between physical
hydrological processes and human activities within a process socio-
hydrological framework. Such models could underpin the practice
of IWRM through an enhanced understanding of coupled human
and water systems dynamics (Sivapalan et al., 2012). The present
study’s objectives were therefore to: (i) develop a watershed-scale
socio-hydrological model using SD to understand the ongoing pro-
cesses and predict possible future trends for a central South Korean
watershed which reuses wastewater for paddy irrigation, (ii) eval-
uate the resultant model for its structure and behavior through
structural and behavioral validity tests, and (iii) employ the model
to understand the impacts of urbanization and instream flow regu-
lation on the socio-hydrological systems of the study watershed in
order to implement sustainable wastewater reuse for agricultural
purposes.
2. Socio-hydrological model
Based on a series of causal loops and model equations, a socio-
hydrological model wherein agricultural water use served as the
nexus of social and hydrological systems was developed (Fig. 1).
This model can be classified as a semi-lumped hydrological model
because the change in land use can be considered by using the
infiltration ratio for the given land use type, while the watershed
system is regarded as a single point in space without dimen-
sions. As in other conceptual hydrological models, modeling the
hydrological system was based on watershed-specific water bal-
ance equations. These included three main stock variables: stream
flow volume, soil water volume, and groundwater volume, rep-
resenting stream flow, interflow, and base flow, respectively. The
social system accounted for changes in population, land use, eco-
nomics, technology, and agricultural wastewater reuse policy. The
resulting model simulates the demand for wastewater reuse, as
influenced by policy and economic feasibility. Such a demand leads
to irrigation capacity development for wastewater reuse, thereby
bringing about changes to the entire hydrological system through
the impacts of wastewater reuse on stream flow and the groundwa-
ter cycle. Changes in hydrological systems can also cause changes in
the policy-making environment resulting in delayed changes in the
overall socio-hydrological system. These feedback changes define a
scenario leading to a balanced state at a specific point in time, along
with the coevolution which occurs between social and hydrological
systems.
2.1. Causal loop diagram
A conceptual modeling step, the causal loop diagram (e.g., Inam
et al., 2015) was used in this study to describe the causal rela-
tionships and dynamics between social and hydrological systems
related to agricultural wastewater reuse. Overall system dynamics
are driven by either reinforcing (+)¨ ör balancing (−)
¨ f̈eedback loops,
where change in one variable feeds back to either reinforce or limit
(i.e., balance) the initial change.
The causal loop diagram developed for this study includes
elements from the hydrological cycle, agricultural water sup-
ply, demand for reclaimed wastewater irrigation, and associ-
ated economic factors, along with changes in land use and
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107 91

Fig. 1. Schematic of a system dynamics-based socio-hydrological model designed to elucidate interactions between social and hydrological systems in a watershed in which
wastewater is reused for paddy irrigation.

population (Fig. 2). Population growth leads to an increase in is the difference between the birth and death rates, of the total
domestic water use and expansion of urban lands. In turn, greater population.
domestic water use leads to greater effluent discharge and stream  t
flow, thereby contributing to groundwater through groundwater P (t) = P (t0 ) + np (t) dt (1)
recharge. By decreasing agricultural water demand and increasing t0
the area of impervious surfaces, urban expansion results in a decline
in both infiltration rate and soil water content. Groundwater np (t) = rg × P (t) (2)
recharge causes a decrease in stream flow, with the two elements
Land use changes are a complicated process, governed by
forming a balancing feedback loop. An increase in groundwater
interactions among several factors including population growth,
increases base flow, thereby increasing stream flow. Groundwa-
economics, and regulations (Lambin et al., 2001). When it is impos-
ter and stream flow therefore form a reinforcing feedback loop,
sible to foresee and include all economic and regulatory changes,
with greater base flow leading to lower groundwater reserves and
such changes can be simulated in SD modeling by using a limited
vice-versa. By increasing soil water content, precipitation decreases
number of factors (Ahmad and Prashar, 2010). For example, in this
agricultural water demand, promoting interflow and percolation,
study, land use changes were restrictively simulated by consider-
thereby increasing stream flow and groundwater, respectively.
ing only the population growth within the Osan River watershed.
As interflow and percolation both decrease soil water content,
Therefore, this study does not reflect the effects of economic volatil-
they form a balancing feedback loop with respect to soil water
ity and regulations within the watershed. Changes in land use were
content. Greater evapotranspiration decreases the soil water con-
simulated by assuming that only agricultural and natural lands
tent, thereby increasing agricultural water demand, which, in turn,
were changed into urban land; this assumption is based on the
increases the agricultural water withdrawal, thereby promoting
analysis of land use change in the study watershed, which has
groundwater withdrawal and wastewater reuse. An increase in
undergone a dramatic urbanization.
groundwater withdrawal results in a decrease in agricultural water
The urban land area (Aurb (t) , ha) was calculated by cumulating
withdrawal, the two elements forming a balancing feedback loop.
the changes in the agricultural land area (Aagr (t) , ha) and natural
Greater groundwater irrigation costs leads to lower groundwater
land area (Anat (t) , ha) over time. The urbanization rates for agri-
withdrawal, thereby increasing the demand for wastewater reuse,
cultural land (rurba , hamonth−1 ) and natural land (rurba , hamonth−1 )
and forming a further balancing feedback loop. Greater wastew-
were employed to calculate the total land use change into urban
ater reuse demand leads to greater wastewater reuse, thereby
land.
decreasing agricultural water groundwater withdrawal. Further-  t
more, greater wastewater reuse leads to farm and social benefits,
Aurb (t) = Aurb (t0 ) + (rurba + rurbn ) dt (3)
thereby forming a reinforcing feedback loop that increases wastew-
t0
ater reuse demand. Likewise, greater wastewater reuse demand
increases wastewater irrigation costs and decreases groundwa-
 t

ter irrigation cost, which form a balancing feedback loop for each Aagr (t) = Aagr (t0 ) − rurba dt (4)
causal relationship. t0
 t
Anat (t) = Anat (t0 ) − rurbn dt (5)
t0
2.2. Model equations
The urbanization rates, as a function of land type, were for-
2.2.1. Changes in population and land use mulated by accounting for the net population and urbanization
The total population (P (t) , person) can be calculated by adding factors of agricultural land (furba , haperson−1 ) and natural land
the net population change (np (t) , personmonth−1 ) to the existing (furbn , haperson−1 ). These parameters were determined by analyz-
population (P (t0 ) , person). The net population change is calculated ing the change in the total population and land use in the study area
by multiplying the population growth rate (rg , month−1 ), which between 2007 and 2009; agricultural and natural lands were found
92 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Fig. 2. Causal loop diagram describing the dynamics between the hydrological system and agricultural wastewater reuse in the study watershed. The relations between
system components in causal loop diagram are briefly presented to capture the essential casual relationships. A “+” sign on an arrow indicates a reinforcing relationship
between two variables, whereas a “−” sign indicates a balancing relationship between two variables.
to be converted into urban land as 0.008 and 0.002 ha per person,
respectively.
rurba = furba × np (t)
rurbn = furbn × np (t)
2.2.2. Economics
The net cost of groundwater and reclaimed wastewater irriga-
tion were evaluated to determine their economic impact. These
were amongst the most important factors in determining the
amount of wastewater reuse. The net cost of groundwater irrigation
(Cg (t) , $103 month−1 ) was calculated using the unit capital cost of
groundwater irrigation (Cg (t) , $103 10−6 m−3 ), the groundwater
(c)
irrigation capacity development (Dg (t) , 106 m3 month−1 ), the unit
operation cost of groundwater irrigation (Cg (t) , $103 10−6 m−3 ),
(o)
and the quantity of groundwater irrigation (Ig (t) , 106 m3 month−1 ).
(c) (o)
Cg (t) = Cg (t) × Dg (t) + Cg (t) × Ig (t)
The net cost of reclaimed wastewater irrigation
(Crw (t) , $103 month−1 ) was calculated using the unit capital
cost of reclaimed wastewater irrigation (Crw (t) , $103 10−6 m−3 ),
(c)
the reclaimed wastewater irrigation capacity development
(Drw (t) , 106 m3 month−1 ), the unit operation cost of reclaimed
wastewater irrigation (Crw (t) , $103 10−6 m−3 ), the quantity of
(o)
reclaimed wastewater irrigation (Irw (t) , 106 m3 month−1 ), and
the farm benefits (Bfarm (t) , $103 month−1 ) and social benefits
(Bsocial (t) , $103 month−1 ) achieved from reclaimed wastewater
irrigation.
(c) (o)
Crw (t) = Crw (t) × Drw (t) + Crw (t) × Irw (t) − Bfarm (t) − Bsocial (t) (9)
Generally, farm benefits from wastewater irrigation occur in the
form of reduced fertilizer application rates and increased yields.
However, the two benefits do not occur concurrently as yield
increases are expected in conditions that do not reduce the amount
of fertilizer used for crop growth (Jeong et al., 2014). Given that
reduced fertilization is becoming conventional practice in wastew-
ater irrigation areas, it was assumed to be the only farm benefit.
Farm benefits are found to occur in particular months: basal,
tillering, and panicle stage fertilizations are made to paddy fields
in May, June, and July, respectively.
 
(6) Irw (t)
Bfarm (t) = pfer × fuse × rr × tfer × Apaddy (t) × (10)
Sa (t)
(7)
 
where pfer is the fertilizer price $103 Mg −1 , fuse is the fertilization
 
rate Mgha−1 , rr is the reduction rate of fertilizer use, tfer is the fer-
tilization timing, Apaddy (t) is the area of paddy fields (hamonth−1 ),
 
and Sa (t) is the agricultural water supply 106 m3 month−1 .
Social benefits from wastewater irrigation occur in the form of
water quality improvements (Bwq (t) , $103 month−1 ) and the acqui-
sition of water resources (Bwr (t) , $103 month−1 ):
Bsocial (t) = Bwq (t) + Bwr (t) (11)
Paddy fields naturally purify irrigation water (Takeda and
Fukushima, 2006), thereby improving water quality. This benefit
was also found to occur in the case of wastewater irrigation (Kim
(8)
et al., 2008; Jang et al., 2012), in a study which used total nitrogen
(T-N) as a water quality indicator.
Bwq (t) = Irw (t) × rfr × pur × cTN × lTN (12)
where rfr is the ratio of agricultural return flow, pur is the purifi-
cation rate of the paddy fields, cTN is the unit treatment cost of
T-N ($103 Mg −1 ), and lTN is the T-N load in the irrigation water
(Mg10−6 m−3 ).
As the amount of wastewater reuse for irrigation increases, the
need to develop new water resources for this purpose is reduced.
This represents a benefit because it enables the use of newly devel-
oped water resources for other purposes. This benefit is calculated
as:
Bwr (t) = ˛ × Irw (t) × cwrd (13)
where ˛ is the reuse ratio, which is defined as the ratio
of effluent discharge to the total stream flow (Jeong et al.,
2016). In the present model this was calculated by dividing the
effluent discharge (Eww (t) , 106 m3 month−1 ) by the stream flow
(Fstream (t) , 106 m3 month−1 ) on a monthly time step, where cwrd is
the unit cost of water resources development ($103 10−6 m−3 ).
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107 93

The total net irrigation cost (Cn (t) , $103 ) is a net present cost reclaimed wastewater irrigation capacity in a given month
that considers the discount factor (ˇ (t)), which is a function of the (Vrw (t) , 106 m3 ).
average nominal interest rate (in ) and time, expressed as the sum  
Irw (t) = Min prw (t) , Vrw (t) (23)
of the net irrigation costs of groundwater irrigation and reclaimed
wastewater irrigation. An agricultural water deficit (Lg (t) , 106 m3 month−1 ) occurs
 when the groundwater irrigation demand exceeds the actual
t
 
Cn (t) = ˇ (t) Cg (t) + Crw (t) dt (14) amount of groundwater irrigation. This leads to the additional
t0 development of groundwater irrigation capacity.
1 Lg (t) = g (t) − Ig (t) (24)
ˇ (t) = t
(15)
(1 + in )
The domestic water demand (dd (t) , 106 m3 month−1 ) is
estimated using the per capita domestic water demand
2.2.3. Water demands (pc)
(dd , 106 m3 person−1 month−1 ) and the population receiving
The potential demand for reclaimed wastewater
the water-supply.
(prw (t) , 106 m3 month−1 ) is a key factor for determining the
(pc)
amount of wastewater irrigation. This can be calculated using dd (t) = dd × P (t) (25)
a cap ratio (), representing a policy decision that accounts for
the target reclaimed wastewater usage for agriculture, and the Effluent discharge is constrained by the wastewater treatment
economics of reclaimed wastewater irrigation to groundwater capacity (Vww (t), 106 m3 ) in a given month and the total wastewater
irrigation (Eq. (16)). The potential demand was calculated for generated, which considers the wastewater generation ratio (ω)
domestic wastewater reuse (Nasiri et al., 2013), and the modified and domestic water demand.
 
form was employed in this study. Eww (t) = Min ω × dd (t) , Vww (t) (26)

 
crw
prw (t) = Max  × a (t) 1 − ,0 (16) 2.2.4. Facility capacities
cg The development of facility capacities reflects the demand
for irrigation and wastewater treatment, and technological sup-
where a (t) is the agricultural water withdrawal (106 m3 month−1 ),
port. The process of each facility development was mainly
crw is the unit average cost of reclaimed wastewater irrigation, and
adopted and modified from the study of Nasiri et al. (2013).The
cg is the unit average cost of groundwater irrigation. crw and cg are
capacities of groundwater irrigation, reclaimed wastewater
calculated as:
irrigation, and wastewater treatment all increase when the



t
crw (t) dt
t
cg (t) dt capacity developments (D (t) , 106 m3 month−1 ) exceed depreciation
t0 t
crw =
t
, cg =
0 t (17) (d (t) , 106 m3 month−1 ).
t0
Irw (t) dt t0
g (t) dt
dVg (t)
= Dg (t) − dg (t) (27)
The agricultural water supply was computed using the dt
paddy field area and the unit agricultural water supply dVrw (t)
= Drw (t) − drw (t) (28)
(sua (t) , 106 m3 month−1 ha−1 ), which was calculated by using dt
the irrigation efficiency and the net irrigation water requirement dVww (t)
from the evapotranspiration input data. = Dww (t) − dww (t) (29)
dt
Sa (t) = sua (t) × Apaddy (t) (18) Each facility is developed to meet increasing demand and to
compensate for capacity depreciations. The capacity reserve ratio
The agricultural water withdrawal was estimated by exclud- () and development delay (T, month) were employed in the mod-
ing the amount of reservoir irrigation (Ir (t) , 106 m3 month−1 ) from eling process of capacity development.
the agricultural water supply. The reservoir irrigation was calcu-     
lated using the agricultural water supply, paddy fields’ area, and Dg (t) = Max g (t) 1 + g − Vg (t) /Tg , dg (t) (30)
reservoir-irrigated area (Arir (t) , hamonth−1 ), i.e., the paddy field  
area irrigated from agricultural reservoirs among the total paddy Drw (t) = Max [prw (t) (1 + rw ) − Vrw (t)] /Trw , drw (t) (31)
fields in the Osan River watershed.  
Dww (t) = Max [ω × dd (t) (1 + ww ) − Vww (t)] /Tww , dww (t) (32)
a (t) = Sa (t) − Ir (t) (19) Capacity depreciation assumes a fixed asset degradation rate
Arir (t) for each facility, which is calculated with the depreciation rate (␦,
Ir (t) = Sa (t) × (20) month−1 ).
Apaddy (t)
dg (t) = ıg × Vg (t) (33)
The groundwater irrigation demand (g (t) , 106 m3 month−1 ) was
calculated by excluding the amount of reclaimed wastewater irri- drw (t) = ırw × Vrw (t) (34)
gation from the agricultural water withdrawal.
dww (t) = ıww × Vww (t) (35)
g (t) = a (t) − Irw (t) (21)
2.2.5. Infiltration and percolation
The groundwater irrigation must meet groundwater irrigation Infiltration (I (t) , 106 m3 month−1 ) is determined by the smaller
demand but is constrained by the groundwater irrigation capacity of the infiltration rates, calculated by considering the ratio (rinf )
in a given month (Vg (t) , 106 m3 ). of infiltration to precipitation (Pre (t) , 106 m3 month−1 ), the dif-
  ference in storage capacity (SC, 106 m3 ) and soil water volume
Ig (t) = Min g (t) , Vg (t) (22)
(V soilwater (t) , 106 m3 ). When the infiltration rate calculated by the
The reclaimed wastewater irrigation must meet potential infiltration ratio and precipitation exceed the potential infiltration
demand for reclaimed wastewater but is constrained by the (SC − Vsoilwater (t − 1)) in a given month, the potential infiltration is
94 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
determined to be the actual infiltration. To consider the state of land
cover, the infiltration ratio is determined using an area-weighted
average, which considers each infiltration ratio (rinfu for urban land,
rinfa for agricultural land, and rinfn for natural land) as a function of
the type of land use.
 
I (t) = Min rinf × Pre (t) , SC − Vsoilwater (t − 1)
    
rinfu × Aurb (t) + rinfa × Aagr (t) + rinfn × Anat (t)
rinf (t) =
Aurb (t) + Aagr (t) + Anat (t)
Using the percolation coefficient (cper , month−1 ), percolation
(Per (t) , 106 m3 month−1 ) was modeled as contributing to the
groundwater volume (Vgw (t) , 106 m3 ) as a certain portion of the
soil water volume over field capacity (FC, 106 m3 ).
Per (t) = cper × (Vsoilwater (t) − FC)
Using the groundwater recharge coefficient (cgre , month−1 ),
groundwater recharge (Gre (t) , 106 m3 month−1 ) was modeled as a
portion of the stream flow volume (Vstream (t) , 106 m3 ) that con-
tributes to the groundwater.
Gre (t) = cgre × Vstream (t)
2.2.6. Evapotranspiration
In hydrological modeling, if one does not consider spatial
distribution, watershed evapotranspiration can be estimated by
multiplying a certain constant by a reference crop evapotranspi-
ration estimated using the FAO Penman–Monteith equation (Allen
et al., 1998). The watershed evapotranspiration coefficient was
calculated using the SWAT model validated for the study water-
shed (Jeong et al., 2016). The actual watershed evapotranspiration
(ET (t) , 106 m3 month−1 ) was calculated by multiplying the water-
shed evapotranspiration coefficient (cETmon ) by the reference crop
evapotranspiration (ET0 (t) , 106 m3 month−1 ) when the soil water
volume, excluding the soil water content at the wilting point
(WP, 106 m3 ) in a given month, exceeds the reference crop evap-
otranspiration. Other cases were calculated by subtracting the soil
water content at the wilting point from the soil water volume.
ET (t) = cETmon × ET0 (t) (Vsoilwater (t) − WP) ≥ ET0 (t)
ET (t) = Vsoilwater (t) − WP (Vsoilwater (t) − WP) < ET0 (t)
Evaporation (ET (t) , 106 m3 month−1 ) from free water surfaces in
the watershed was calculated by multiplying an evaporation coef-
ficient (ceva , month−1 ) by the stream flow volume.
E (t) = ceva × Vstream (t)
2.2.7. Water balance
The surface water balance considers the stream flow volume,
inflow (IN (t) , 106 m3 month−1 ), effluent discharge, evapora-
tion, stream flow, and groundwater recharge. The inflow
consists of the surface flow (Fsurface (t) , 106 m3 month−1 ),
interflow (Finter (t) , 106 m3 month−1 ), and base flow
(Fbase (t) , 10 m month−1 ). The stream flow was estimated as
6 3
the amount of the stream flow volume flowing out multiplied by a
coefficient of stream flow (cstream , month−1 ).
dVstream (t)
= IN (t) + Eww (t) − E (t) − Fstream (t) − Gre (t)
dt
IN (t) = Fsurface (t) + Finter (t) + Fbase (t)
Fstream (t) = cstream × Vstream (t)
The soil water balance considered the soil water volume, infil-
tration, evapotranspiration, interflow, and percolation.
dVsoilwater (t)
= I (t) − ET (t) − Finter (t) − Per (t)
dt
The groundwater balance considered the groundwater volume,
percolation, groundwater recharge, base flow, and groundwater
loss (Fgwloss (t) , 106 m3 month−1 ) flowing out of the watershed.
dVgw (t)
= Pre (t) + Gre (t) − Fbase (t) − Fgwloss (t) (47)
dt
(36)
 2.2.8. Runoff and return flow
The surface flow was estimated as precipitation in excess of the
(37)
saturated soil’s storage capacity.
Fsurface (t) = Pre (t) − I (t) (48)
The interflow was estimated using the interflow coefficient
(cint , month−1 ), assuming that interflow only occurs when the soil
water volume exceeds the soil’s field capacity.
(38) Finter (t) = cint × (Vsoilwater (t) − FC) (49)
Using the base flow coefficient (cbase , month−1 ), the base flow
was modeled as a certain proportion of the groundwater volume,
where hydraulic head represents the difference between the soil
water level and that of the river.
(39) Fbase (t) = cbase × Vgw (t) (50)
Groundwater loss, indicating that groundwater is flowing out
of the study watershed, was estimated using the groundwater loss
coefficient (cgwloss , month−1 ).
Fgwloss (t) = cgwloss × Vgw (t) (51)
The return flow (Freturn (t) , 10 6
m3 month−1 ),
which represents
the portion of the applied irrigation water that is not consumed in
the paddy fields but instead flows into streams (Kim et al., 2009),
was modeled as a certain proportion of the agricultural water
supply using an agricultural return flow ratio obtained from the
Ministry of Land, Transport and Maritime Affairs (2011).
Freturn (t) = rrf × Sa (t) (52)
Further details about parameters used in the model for the Osan
River watershed are presented in Tables 1A and 1B.
(40)
3. Materials and methods
(41)
3.1. Study watershed
Socio-hydrological modeling was used in central South Korea’s
Osan River watershed (Fig. 3), which is the spatial boundary con-
(42) dition for this study. With an area of 98.3 km2 , a drainage density
of 0.52 km/km2 , a main river length of 16.49 km, and a drainage
slope of 10.21%, the Osan River watershed extends between lat-
itude N 37◦ 05 56 and N 37◦ 14 05 and longitude E 127◦ 01 29
and E 127◦ 09 38 . The Osan River watershed has a temperate
monsoon climate, with a mean annual temperature of approxi-
mately 12 ◦ C, and a mean annual precipitation of approximately
1312 mm. Groundwater level was measured in real time at the
Osan Gwol-Dong groundwater level monitoring station (Fig. 3),
managed by the Ministry of Land, Infrastructure, and Transport.
The Osan River watershed contains the Osan wastewater treat-
ment plant (WWTP) (Fig. 3), which has a treatment capacity of
(43) 140,000 m3 day−1 . There are some agricultural pumping stations
that take in treated wastewater, which has been diluted with
(44)
stream water, to irrigate paddy fields downstream. This type of
(45) agricultural activity is referred to as indirect wastewater reuse.
In the Osan River watershed, which is currently undergoing
rapid urbanization, rice lodging caused by the effects of effluent dis-
charge on the irrigation water has occasionally been reported, and
stream water quality has been found to fall short of local standards
(46)
(Jeong et al., 2016). The demand for sound indirect wastewater
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107 95

Fig. 3. Location of the study watershed in South Korea and its monitoring networks including the site of the groundwater level monitoring station and meteorological station.

Fig. 4. Sensitivity of (b) potential demand for reclaimed wastewater, (c) total net present cost of irrigation, and (d) base flow to the (a) urbanization factor of agricultural
land. 50, 75, 95, and 100% confidence bounds are shown. (*) Denotes agricultural land.
96 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107

Table 1A
List of socio-hydrological modeling parameters for the Osan River watershed.

Parameters Descriptions Value Unit

cstream Coefficient of stream flow 0.98a month−1

cint Coefficient of interflow 0.20a month−1
cbase Coefficient of base flow 0.55a month−1
cETmon Coefficient of watershed evapotranspiration varies -b
ceva Coefficient of evaporation 0.03a month−1
cper Coefficient of percolation 0.10a month−1
cgre Coefficient of groundwater recharge 0.10a month−1
cgwloss Coefficient of groundwater that flows out of watershed 0.05a month−1
SC Storage capacity 20a 106 m3
WP Wilting point 2a 106 m3
FC Field capacity 5a 106 m3
rinf Ratio of infiltration to precipitation CVc –
rinfu Infiltration ratio of urban land 0.36a –
rinfa Infiltration ratio of agricultural land 0.40a –
rinfn Infiltration ratio of natural land 0.63a –
rfr Ratio of agricultural return flow 0.35 –
$103 10−6 m−3
(c)
Cg (t0 ) Unit capital cost of groundwater irrigation capacity 111
$103 10−6 m−3
(c)
Crw (t0 ) Unit capital cost of reclaimed wastewater irrigation capacity 100
$103 10−6 m−3
(o)
Cg (t0 ) Unit operation cost of groundwater irrigation 70
$103 10−6 m−3
(o)
Crw (t0 ) Unit operation cost of reclaimed wastewater irrigation 120
cwrd Unit cost of water resources development 445 $103 10−6 m−3
cTN Unit treatment cost of total nitrogen 5.917 $103 Mg −1
pfer Fertilizer price 0.4975 $103 Mg −1
fuse Amount of fertilizer use 0.11 Mgha−1
rr Reduction rate of fertilizer use 0.2 –
a
Indicates a calibrated value.
b
Indicates dimensionless unit.
c
Indicates a value calculated from the infiltration ratios of urban, agricultural, and natural land.

Table 1B
List of socio-hydrological modeling parameters for the Osan River watershed (continued).

Parameters Descriptions Value Unit

pur Purification rate of paddy fields 0.45 -a

tfer Fertilization timing varies –
lTN T-N load in irrigation water 6.26 Mg10−6 m−3
 Cap ratio of wastewater reuse 0.5 –
ip Inflation rate 0.00334 –
in Nominal interest rate 0.00447 –
P(t0 ) Population 297,560 person
(pc)
dd Per capita domestic water demand 295 106 m3 person−1 month−1
Apaddy (t0 ) Total area of paddy fields 1799 ha
Arir (t0 ) Reservoir irrigated area 1232 ha
Aurb (t0 ) Urban land area 3365 ha
Aagr (t0 ) Agricultural land area 2413 ha
Anat (t0 ) Natural land area 4053 ha
rg Population growth rate varies month−1
g Reserve ratio of groundwater irrigation capacity 0.1 –
rw Reserve ratio of reclaimed wastewater irrigation capacity 0.1 –
ww Reserve ratio of wastewater treatment irrigation capacity 0.1 –
Tg Development delay of groundwater irrigation capacity 48 month
Trw Development delay of reclaimed wastewater irrigation capacity 48 month
Tww Development delay of wastewater treatment capacity 24 month
ıg Depreciation rate of groundwater irrigation capacity 0.001667 month−1
ırw Depreciation rate of reclaimed wastewater irrigation capacity 0.001667 month−1
ıww Depreciation rate of wastewater treatment irrigation capacity 0.001667 month−1
furba Urbanization factor of agricultural land 0.008 haperson−1
furbn Urbanization factor of natural land 0.002 haperson−1
a
Indicates dimensionless unit.

reuse systems is expected to increase in the future in order to meet
changes in the seasonal demand for irrigation water and safe agri-
cultural food products. In order to implement a successful indirect
wastewater reuse policy that remains effective under changing cir-
cumstances imposed by urbanization, it is necessary to understand
the ongoing process and to predict possible future trends using a
process socio-hydrology framework.
3.2. Model inputs
Daily weather records from 1964 to 2012, including precipi-
tation, maximum and minimum temperatures, relative humidity,
mean wind velocity, and solar radiation, were obtained from the
Suwon National Meteorological Station, which is located approxi-
mately 7 km north of the Osan River watershed. These records were
used in evaluating the behavioral validity of the established model.
Representative Concentration Pathway 8.5 (RCP 8.5) scenarios
were obtained from the Korean Meteorological Administration and
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
downscaled using the method devised by Alcamo et al. (1997) to
generate future weather data. These records were used to verify the
structural validity of the established model and to anticipate future
trends in the study watershed. The historical reference crop evap-
otranspiration was estimated using the FAO Penman–Monteith
equation (Allen et al., 1998) and historical weather data. Future
reference crop evapotranspiration was estimated using the FAO
Penman–Monteith equation and a transfer function model. The
transfer function model was established with stepwise multiple
linear regression. Land use data for 2007 and 2009, obtained from
the Ministry of the Environment, were used to characterize land use
in the study watershed. The types of land use in the study watershed
were classified as urban, agricultural, and natural. The distribution
of land use in 2007 was 19.4% urban, 36.8% agricultural, and 43.8%
natural, shifting to 34.2% urban, 24.5% agricultural, and 41.3% natu-
ral by 2009. The future population was estimated using population
projections from 2010 to 2060 conducted by the Korean National
Statistical Office (National Statistical Office (NSO, 2012).
3.3. Model evaluation
In the case of a model that uses SD, the model’s validity must be
verified in terms of its structural and behavioral aspects. Indeed, a
causal-descriptive model must not only predict behavior, but also
explain how the behavior is generated through the validity of the
model’s internal structure (Barlas, 1996). Model structural valid-
ity tests whether the established model reasonably represents the
phenomenon that is simulated through the set of relations between
the elements used in the model (Khan et al., 2009). In this study,
behavior sensitivity and extreme condition tests, both powerful
and practical indirect structural tests (Barlas, 1996; Khan et al.,
2009), were employed to evaluate model structural validity.
Model behavioral validity refers to how well the simulated
behavior reproduces the observed behavior, i.e., the real sys-
tem’s anticipated trends (Stave, 2003; Qudrat-Ullah, 2012). In the
present study, the behavior of evapotranspiration, streamflow, and
groundwater level was tested to ensure they represent the hydro-
logical processes they model. The simulated evapotranspiration
and streamflow were evaluated in a single watershed dimension
and at the outlet of the study watershed, respectively, by compar-
ison with simulated results from a version of the Soil and Water
Assessment Tool (SWAT) model validated (by the authors) for the
study watershed (Jeong et al., 2016). Model-generated groundwa-
ter levels were evaluated by indirect comparison with the observed
data from the Osan Gwol-Dong groundwater level monitoring sta-
tion. As proposed by Moriasi et al. (2007), the performance statistics
of Nash-Sutcliffe efficiency (ENS ) (Nash and Sutcliffe, 1970), root
mean square error of observations standard deviation ratio (RSR)
(Singh et al., 2004), and percent bias (PBIAS) were used to evaluate
the model behavioral validity on a monthly time step. The accuracy
of model performance was also determined as either sufficient or
insufficient according to whether the coefficient of determination
(R2 ) value for a linear regression between modeled and SWAT-
computed or observed values was greater than or lesser than 0.6,
which was the criteria used by Santhi et al. (2001).
3.4. Simulation scenarios
A variety of scenarios were constructed to understand the
dynamics of urbanized socio-hydrological systems (Table 2). A
baseline scenario, including population growth and changes in land
use under initial urbanization conditions served as a standard. In
using the extreme conditions test to verify the established model’s
structural validity, maximum wastewater reuse (MWR) and no
wastewater reuse (NWR) scenarios were prepared by changing the
cap ratio of the baseline scenario to 0.8 or 0.0, respectively. The
97
Table 2
Summary of the scenarios used in model validity tests and applications.
Scenario Descriptions
Baseline Baseline scenario with initial conditions
aiming to supply 50% of the needed
irrigation water (by using a cap ratio of
0.5) through indirect wastewater reuse
while considering changes in the
population and land use
Maximum wastewater reuse for Maximum use of reclaimed wastewater for
agriculture (MWR) agriculture by changing the cap ratio
from 0.5 to 0.8 in the baseline scenario
No wastewater reuse for agricultureNo use of reclaimed wastewater for
(NWR) agriculture by changing the cap ratio
from 0.5 to 0.0 in the baseline scenario
Population and land use change Population and land use remain constant
control (PLC) as in the baseline scenario
Land use change control (LC) Land use remains constant as in the
baseline scenario
Instream flow regulation (IFR) Restricting stream water intake by
including instream flow in the MWR
scenario
population and land use change control (PLC) allows a scenario to
be modeled that does not take into account the baseline scenario’s
standard changes in population and land use, while the land use
change control (LC) allows for a scenario that does not incorporate
the baseline scenario’s land use changes. The instream flow regu-
lation (IFM) scenario is employed to represent the response of the
social system to excessive indirect wastewater reuse by restricting
the stream water intake. The PLC, LC, and IFM scenarios were pre-
pared for model application. The impacts of population growth on
wastewater reuse for agriculture can be evaluated by comparing
the PLC and LC scenarios, while the impacts of changes in land use
can be assessed by comparing the baseline and LC scenarios. The
impacts of instream flow regulation were evaluated by comparing
the MWR and IFM scenarios.
4. Results and discussion
4.1. Model validity
4.1.1. Structural validity
Structural validity assessed through the behavior sensitivity and
extreme conditions tests was projected for 2010 and 2030, a period
over which the socio-hydrological systems were modeled to change
rapidly as a result of the effects of urbanization and changes in
wastewater reuse practice. Graphical presentations were employed
to make it possible to understand the overall behavioral changes of
the target variable and to give an intuitive grasp of the response
behavior of the system under the designed model structure and
extreme conditions. Behavior sensitivity tests were performed for
the urbanization factor of agricultural land, the cap ratio, and the
unit operation cost of the reclaimed wastewater irrigation, because
these parameters represent social, policy, and economic systems,
respectively, and can have significant impacts on hydrological pro-
cesses.
For a given growth in population, the greater the urbanization
factor of agricultural land, the more agricultural land was converted
to urban land. The behavior sensitivities to the urbanization factor
of agricultural land (base value = 0.008 ha person−1 ) were tested
by varying the value from 0.004 to 0.012 ha person−1 (Fig. 4a). For
a given degree of urbanization, potential demands increased over
time, rising by roughly 11% between 2010 and 2030 (Fig. 4b). The
total net present cost of irrigation was also reduced as the amount of
agricultural land decreased, changing by approximately 10% com-
pared to the initial conditions (Fig. 4c). Urbanization directly causes
changes in the hydrological cycle through its impacts on infiltration
98 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Fig. 5. Sensitivity of (b) potential demand for reclaimed wastewater, (c) total net present cost of irrigation, and (d) base flow to the (a) cap ratio. 50, 75, 95, and 100%
confidence bounds are shown.
and by causing changes in the behavior of groundwater through
groundwater irrigation. Base flow only showed a small sensitivity
to urbanization between 2010 and 2030 (2%), likely because infil-
tration and groundwater irrigation’s impacts on base flow offset
each other (Fig. 4d).
The cap ratio related to the amount of reclaimed wastewa-
ter used for agriculture increases when greater reuse is targeted.
Behavior sensitivities to the cap ratio (base value = 0.5) were tested
by varying the value from 0.2 to 0.8 (Fig. 5a). The potential demand
for reclaimed wastewater rose as the cap ratio increased, grow-
ing by roughly 60% between 2010 and 2030. Potential demand was
shown to accurately reflect changes in the cap ratio, indicating that
the quantity of irrigation is mainly determined by the policy level
regarding reclaimed wastewater (Fig. 5b). The total net present cost
of irrigation decreased as the reused amount of reclaimed wastew-
ater, which reflects the potential demand, increased. The cap ratio
was found to influence the total irrigation cost, which increased
by 95% from 2010 to 2030 (Fig. 5c). Reclaimed wastewater irri-
gation was found to affect groundwater irrigation, which had a
decisive effect on groundwater and base flow. Initially, the dif-
ference between base flows caused by the reclaimed wastewater
irrigation was negligible due to the delay in development of the irri-
gation facility for reclaimed wastewater irrigation. However, this
difference gradually increased over time and the base flow changed
by approximately 8% between 2010 and 2030 (Fig. 5d).
The economics of reclaimed wastewater irrigation influences
the amount of reclaimed wastewater that is reused. Behavior sensi-
tivities to the unit operation cost of reclaimed wastewater irrigation
(base value = $120,000 per million m3 ) were tested by varying the
value from $60 to $180,000 per million m3 (Fig. 6a). The unit oper-
ation cost showed a continued impact on the potential demand
for reclaimed wastewater over the simulation period. An increase
in the unit operating cost caused a significant (36%) reduction in
potential demand between 2010 and 2030 (Fig. 6b). The large vari-
ation in the total net present cost of irrigation, corresponding to
variations in unit operating cost, showed the impact of the unit
operation cost on the total irrigation cost to significantly increase
over time. The total net present cost corresponding to the unit oper-
ation cost changed by approximately 86% between 2010 and 2030
(Fig. 6c). The unit operation cost affects the base flow indirectly by
changing the potential demand. As unit operation cost increased
between 2010 and 2030, base flow decreased marginally (approx.
1%; Fig. 6d).
Extreme condition tests were performed for the MWR and
NWR scenarios and compared to the initial conditions of the base-
line scenario. For the potential demand for reclaimed wastewater,
the MWR showed the maximum potential demand, reflecting the
cap ratio for the simulation period. In contrast, the NWR did
not generate any potential demand (Fig. 7a). Reclaimed wastew-
ater irrigation under the MRW regime, which was limited by the
irrigation capacity of the reclaimed wastewater, showed a ten-
dency to increase steadily. In contrast, the NRW regime showed
no reclaimed wastewater irrigation (Fig. 7b). The groundwater irri-
gation capacity of the MRW was at almost constant capacity once
the reclaimed wastewater irrigation capacity of the MRW reached
a level capable of satisfying the potential demand for reclaimed
wastewater (c. 2020), whereas the installed capacity of the NRW
appeared to increase steadily, even after a significant number of
groundwater irrigation facilities were developed (Fig. 7c). Ground-
water irrigation under the MRW regime showed a tendency to
decline after the irrigation facility for reclaimed wastewater irriga-
tion was developed to meet its demand, whereas under the NRW
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Fig. 6. Sensitivity of (b) potential demand for reclaimed wastewater, (c) total net present cost of irrigation, and (d) base flow to the (a) unit operation cost of reclaimed
wastewater irrigation. 50, 75, 95, and 100% confidence bounds are shown. (*) Denotes reclaimed wastewater irrigation.
regime a continuous increase in groundwater irrigation occurred
under its limited facility capacity (Fig. 7d).
In the behavior sensitivity test, if the model behavior is sur-
prisingly sensitive to certain parameters, it may indicate a flaw
in the model structure (Saysel and Barlas, 2001). Structural flaws
could be also revealed by extreme condition tests, which compare
the model generated behavior to the anticipated behavior of the
real system under the same extreme conditions, which can often
be anticipated by human beings (Barlas, 1996). The indirect struc-
tural tests presented above showed the model structure to exhibit
the expected sensitivity to the selected parameters in the behavior
sensitivity tests, yielding reasonable and consistent results under
both extreme conditions. This supports the validity of the model
structure.
4.1.2. Behavioral validity
Monthly watershed evapotranspiration simulated by the site-
validated SWAT model between 1971 and 1990 was used for
model calibration. A trial-and-error method was employed to
adjust the parameters related to watershed evapotranspiration
(Tables 1A and 1B). The quality of model performance for the
calibration period was determined to be quite satisfactory given
ENS , RSR, PBIAS, and R2 values of 0.73, 0.51, −2.6%, and 0.80,
respectively (Table 3). Simulated watershed evapotranspiration
accurately mimicked the SWAT simulation results for the cali-
bration period (Fig. 8). After model calibration (1971–1990), the
accuracy of the socio-hydrological model was tested by compari-
son with monthly watershed evapotranspiration simulated by the
site-validated SWAT model (1991–2000). The validation results of
the socio-hydrological model showed good performance given that
values of ENS , RSR, PBIAS, and R2 were 0.77, 0.47, 3.0%, and 0.79,
99
Table 3
Model performance evaluation statistics for monthly evapotranspiration and stream
flow, compared with the simulation results of the site-validated SWAT model, over
the entire simulation period.
Simulation phases Evapotranspiration Stream flow
ENS RSR PBIAS R ENS 2
RSR PBIAS R2
Calibration (1971–1990) 0.73 0.51 −2.6% 0.80 0.66 0.58 3.9% 0.73
Validation (1991–2010) 0.77 0.47 3.0% 0.79 0.69 0.56 1.4% 0.75
respectively (Table 3). Graphical evaluation of accuracy between
the established socio-hydrological model and the verified SWAT
model for the validation period also showed good agreement
(Fig. 8).
Using the calibrated parameters (Tables 1A and 1B), the same
procedure was employed to verify stream flow predictions. Given
values of ENS , RSR, PBIAS, and R2 of 0.66, 0.58, 3.9%, and 0.73
(Table 3), respectively, model performance for the calibration
period (1971–1990) was determined to be very good. Performance
during the validation period (1991–2000) was marginally better,
with ENS , RSR, PBIAS, and R2 values of 0.69, 0.56, 1.4%, and 0.75,
respectively (Table 3). Graphical comparisons showed that the
socio-hydrological model reproduced the stream flow behavior of
the SWAT model with sufficient accuracy, although the difference
in simulation time step between the two models led to small differ-
ences for high monthly stream flows (Fig. 9). According to statistical
and graphical model evaluation, model performance during both
calibration and validation periods showed very good agreement in
simulating monthly stream flow.
The generated monthly groundwater level, derived from the
constructed groundwater level-base flow relationship, was com-
pared with the observed monthly groundwater level data in
100 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Fig. 7. Comparisons of simulation results under extreme boundary conditions. * denotes reclaimed wastewater. a and b respectively indicate the extreme boundary conditions
that yield the maximum and minimum irrigation amounts of reclaimed wastewater.
the study watershed from 2005 to 2012. The established socio-
hydrological model was determined to be capable of simulating
the actual behavior of the groundwater with relative accuracy;
observed and model-based groundwater levels exhibited a linear
relationship with an R2 value of 0.70 (Fig. 10).
Validating the behavior of the social components is difficult due
to lack of data. To assess the accuracy of the simulated capacity of
the Osan WWTP, which reflects population changes and technical
support and considers capacity development and depreciation, this
capacity was compared to that in the master plan (Osan, 2012). The
capacity of the WWTP was set to the current monthly treatment
capacity (3,600,000 m3 ) and the installed capacity was simulated
by taking into account changes in the population and wastewa-
ter generation ratio. The simulation result appears to provide a
good estimate, as the installed capacity in the 2030s converges to
4,200,000 m3 , which agrees with the capacity forecast in the master
plan (Fig. 11). Changes in the WWTP capacity according to popu-
lation growth corresponds to one of the six dynamic patterns in
SD modeling (Ford, 2009), indicating that the target variables of an
established model can reproduce the general trends of expected
behavior (Stave, 2003).
4.2. Model applications
4.2.1. Urbanization impacts
The impacts of urbanization are mainly caused by population
growth and changes in land use. These factors influence hydrologi-
cal dynamics (Rose and Peters, 2001; DeFries and Eshleman, 2004)
and affect indirect wastewater reuse by increasing the quantity of
wastewater effluent (Jeong et al., 2016). These were evaluated for
the period of 2010–2059 by running the established model with
the generated future weather data.
Population growth requires a large-scale WWTP, and the result-
ing effluents affect the hydrological dynamics and agricultural
activities of the stream water which effluents flow into. An increase
in population increases stream flow by causing an increase in efflu-
ent discharge. Population changes were well reflected in future
trends of stream flow (Table 4). Population growth, which increases
stream flow, contributed to groundwater recharge and resulted
in an increase in groundwater levels. Future trends in ground-
water levels showed a similar tendency to those of stream flow
because the differences in the groundwater level between the PLC
and LC scenarios accurately reflect differences in population. The
increase in effluent discharge led to an increase in the reuse ratio
and increased the rate of reuse of reclaimed wastewater (Table 4).
As land changes from an agricultural to urban setting, the
infiltration rate is reduced. The difference in the mean annual infil-
tration appeared to continue to increase until the 2030s as land
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107 101

Fig. 8. Comparison of simulated monthly evapotranspiration between SWAT and the socio-hydrological model for the (a) calibration (1971–1990) and (b) validation
(1991–2010) periods.

Table 4 use continued to change. Change in infiltration was relatively small

Impacts of population change on the future trends in stream flow, groundwater
after the 2030s when land cover conditions became relatively con-
level, and amount of effluent reuse for the 50-year simulation period (2010–2059).
The results from the LC scenario show the impacts of population change. stant (Table 5), indicating that the type of land cover can have a
greater impact on infiltration than either weather or soil conditions.
Classification Scenario Decade
As was expected, the agricultural water demand also decreased as
2010s 2020s 2030s 2040s 2050s the expanse of agricultural land decreased. In the beginning, the
Population (person) PLC a
297,560 297,560 297,560 297,560 297,560 effects of changes in land use can have a significant impact on agri-
LCb 314,734 338,651 345,167 333,776 309,783 cultural water demands, but after the land cover conditions became
Monthly stream flow PLC 8.98 9.95 8.52 9.92 9.75 constant, increased evapotranspiration due to climate change was
(106 m3 ) LC 9.16 10.39 9.04 10.33 9.90 reflected in the increase in agricultural water demand (Table 5).
Monthly groundwater PLC 47.19 47.11 46.50 47.39 47.11
LC 47.27 47.31 46.76 47.57 47.18
These results are not surprising given that an increase in the
level (m)
Annual reuse amount PLC 0.21 0.62 0.85 0.84 0.81 expanse of impervious areas occurs as a consequence of urban-
of effluent (106 m3 ) LC 0.23 0.67 0.93 0.91 0.83 ization. The same results are achieved using traditional modeling
a
Indicates a scenario of population and land use change control. methods. However, interestingly, the present model shows increas-
b
Indicates a scenario of land use change control. ing groundwater levels despite a decreasing infiltration rate. This
results from a decrease in groundwater irrigation for agricul-
ture due to the reduction in the expanse of agricultural lands
102 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107

Fig. 9. Comparison of simulated monthly stream flow between SWAT and the socio-hydrological model for the (a) calibration (1971–1990) and (c) validation (1991–2010)
periods, along with similar comparisons of monthly flow duration curves for the (b) calibration and (d) validation periods.

Fig. 10. Comparison between observed and simulated groundwater levels.

(Table 5). The impacts of groundwater irrigation on groundwa-
ter levels can be further elucidated by comparing baseline and
LC scenario monthly groundwater level changes over the simula-
tion period (Fig. 12). The difference in groundwater levels between
the two scenarios increased after April when paddy irrigation
began. Thereafter the impact of groundwater irrigation was sim-
ulated to continue until the end of the year. This result indicates
that the anthropogenic activity of groundwater irrigation can have
a greater impact than infiltration, which is a common compo-
nent of the hydrological cycle that has an impact on groundwater
behavior. Based on these results, the use of socio-hydrological mod-
eling allows us to confirm the fact that anthropogenic activities
should be included when attempting to obtain a more fundamental
understanding of hydrological dynamics.
4.2.2. Impacts of instream flow regulation
Using more reclaimed wastewater would help meet policies
that encourage wastewater reuse; however, excessive use of
wastewater can cause environmental problems by reducing stream
flow. River authorities can restrict reuse through instream flow
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107 103

Fig. 11. Simulated facility capacity of the Osan wastewater treatment plant. (*) Indicates wastewater treatment plant.

Fig. 12. Mean monthly groundwater levels for the 50-year simulation period (2010–2059) showing increases in groundwater Table depth according to land use change. (*)
Indicates a scenario excluding land use changes from the baseline scenario.

regulation, thereby limiting excessive reuse. The impacts of
instream flow regulation were analyzed by simulating different
possible future trends in the socio-hydrological system for the
period extending from 2010 to 2059. The estimated mean min-
imum monthly flow (5.37 × 106 m3 month−1 ) of the Osan river,
the main stream in the Osan River watershed — obtained using
the site-validated SWAT model operating over a period of 30 years
(1981–2010) — was employed as the instream flow.
Restricting stream water intake through instream flow regu-
lation is generally known to secure stream flow. This study also
showed that instream flow regulation ensures stream flow dur-
ing the spring drought when the demand for agricultural water
in Korea is the largest (Fig. 13). However, there is also the pos-
sibility that regulation can cause a decrease in stream flow after
the spring drought due to a decreased base flow, which is affected
by groundwater irrigation, which compensates for reduced stream
water irrigation (Fig. 13 and Table 6).
Socio-hydrological modeling can also simulate social benefits
and hydrological behavior including agricultural water use. The
changes in social benefits showed that reusing a certain amount
of effluent brings the benefits of additional water and improved
water quality. The regulation associated with reclaimed wastewa-
ter irrigation causes a decrease in the amount of effluent that can
be reused, resulting in a loss of social benefits from agricultural
wastewater reuse (Table 6). The opportunity cost for the instream
flow regulation on wastewater reuse for the watershed can be eval-
uated from the coupled modeling approach.
104 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107

Fig. 13. Monthly stream flow for the 50-year simulation period (2010–2059) and (inset) comparison between two scenarios for monthly stream flow in 2036. (a and b )
Indicate the maximum use of reclaimed wastewater for paddy irrigation either considering (IFR) or not considering (MWR) the instream flow, respectively.

Table 5 4.3. Implications and limitations of SD-based socio-hydrological

Impacts of land use change on future infiltration, agricultural water demand,
modeling
groundwater level, and groundwater irrigation for the 50-year simulation period
(2010–2059). The results from the Baseline scenario show the impacts of land use
change. Traditional hydrological modeling can promote unsustainable
watershed management policies and practices because it uses a
Classification Scenario Decade
linear approach to evaluate the hydrological dynamics caused by
2010s 2020s 2030s 2040s 2050s anthropogenic activities (Mirchi et al., 2012). Conversely, a SD-
Annual infiltration LC a
44.66 44.25 41.56 46.52 46.58 based socio-hydrological modeling approach provides a framework
(106 m3 ) Baselineb 44.55 44.07 41.29 46.29 46.33 for integrating the individual physical processes of the hydrologic
Annual agricultural LC 23.82 23.16 24.25 23.56 25.10 cycle, along with socio-economic and policy systems, thus helping
water demand (106 m3 ) Baseline 22.18 19.47 19.63 19.04 20.29
LC 47.28 47.29 46.80 47.56 47.18
to conceptualize and simulate the complex dynamics of integrated
Monthly groundwater
level (m) Baseline 47.27 47.45 47.00 47.73 47.34 systems. This facilitates a holistic understanding of hydrologi-
Annual groundwater LC 1.66 2.82 2.82 2.70 2.79 cal systems that includes both anthropogenic and social factors.
irrigation (106 m3 ) Baseline 1.60 2.41 2.26 2.17 2.27 The socio-hydrological model established in this study can evalu-
a
Indicates a scenario excluding land use changes from the baseline scenario. ate the impacts of agricultural wastewater reuse for small-scale
b
Indicates a standard scenario, including changes in population and land use, that socio-hydrological systems by considering urbanization impacts
assumes the initial conditions. and simulating the responses to systemic changes by using policy
scenarios including instream flow regulation. It can be seen that
socio-hydrological modeling can be used as a decision support sys-
Table 6 tem for designing water resource policies contributing to successful
Impacts of instream flow regulation on the future trends in stream flow, groundwa- implementation of IWRM through an improved understanding of
ter irrigation, amount of effluent reuse, and social benefits for the 50-year simulation the changing processes of systems and their elements.
period (2010–2059). The results from the IFR scenario show the impacts of instream
flow regulation.
The main advantage of socio-hydrological modeling lies in
understanding the ongoing processes of change in complex systems
Classification Scenario Decade rather than the mere prediction of hydrological dynamics at a given
2010s 2020s 2030s 2040s 2050s time, which traditional modeling approaches have predominately
Annual groundwater MWRa 1.44 1.44 1.12 1.04 1.13
focused on so far. As Sivapalan et al. (2012) pointed out, a SD-based
irrigation (106 m3 ) IFRb 1.45 1.47 1.20 1.11 1.18 socio-hydrological model can serve as a tool to explore processes
Annual reuse amount MWR 0.35 0.98 1.28 1.22 1.11 and help improve our understanding of general trends and the root
of effluent (106 m3 ) IFR 0.35 0.95 1.22 1.18 1.08 causes of problems, and predict possible future systemic trends
Annual social benefits MWR 131 446 566 580 505
through more detailed insights into causality. Understanding the
from wastewater reuse IFR 131 432 543 553 502
dynamics of a system is especially important in preparing for unex-
a ($1000)
and b Indicates the maximum use of reclaimed wastewater for paddy irrigation
pected system behaviors. In this study, urbanization was found
either considering (IFM) or not considering (MWR) the instream flow, respectively.
to increase the groundwater level by decreasing groundwater
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
irrigation, and instream flow regulation was demonstrated to cause
a decrease in stream flow due to decreased base flow resulting
from compensative groundwater irrigation, as well as a decrease in
social benefits from the decline in wastewater reuse. These results
could be only partially derived by existing hydrological modeling
techniques.
The model presented in this study is not a general socio-
hydrological model, it can neither be employed at all scales, nor
does it address all issues related to hydrological dynamics. How-
ever, it operates as a small-scale model, which can capture the
features of wastewater reuse practices for paddy irrigation at the
watershed scale. Some results derived from this study cannot be
generalized because the model structure was designed to deal with
site-specific wastewater reuse issues. Modeling using SD requires
interdisciplinary insights, such that modeling anthropogenic sys-
tems is limited by the fact that the quantification of socio-economic
and political components is challenging (Madani and Marino,
2009). Notwithstanding the fact that socio-hydrological frame-
works are still in their infancy (Troy et al., 2015), the fact that no
single disciplinary field can design useful solutions to the global
water crisis (Srinivasan et al., 2012), research based on interdisci-
plinary or transdisciplinary approaches such as socio-hydrological
modeling are promising tools to understand hydrologic processes.
In this regard socio-hydrological modeling using a SD approach is a
starting point towards a better understanding of changing hydro-
logical systems.
5. Conclusions
Through the present study we gained an increased understand-
ing of socio-hydrological systems, and how predictions of system
behavior can be used to develop sustainable water resource man-
agement strategies targeted to achieving long-term water security
for people and their environment. A socio-hydrological model that
combined a deterministic conceptual hydrological model and a
social model, which considered population, land use, economics,
technology, and wastewater reuse policies, was developed using
SD. The model was applied to a central South Korean watershed
which reuses wastewater for paddy irrigation. The model showed
its structural validity by showing sensitivity within an expected
range and reasonable results from behavior sensitivity and extreme
conditions tests. The model also exhibited its behavioral validity
through comparisons of predicted hydrological processes with the
results from a site-validated SWAT model and observed data. The
verified SD model showed that urbanization in the study water-
shed increases stream flow, groundwater level, and the quantity
of wastewater reuse by engendering an increase in effluent dis-
charge tied to population growth, while changes in land use led to a
decrease in the infiltration rate and agricultural water demand. The
groundwater level was found to be increased by decreasing ground-
water irrigation. Instream flow regulation was demonstrated to
ensure stream flow during the spring drought but also to cause
a decrease in stream flow at other times due to a decreased base
flow, which occurs as a consequence of compensative groundwater
irrigation.
A SD-based socio-hydrological model represents a useful
methodology for process socio-hydrology by providing a deeper
understanding of the interactions of small-scale socio-hydrological
systems. By using this model as a tool for thinking, an under-
standing of general trends can be gained and the root causes of
problems identified, thus contributing to the implementation of
IWRM. Further studies are required to quantify social systems in a
more scientific manner and to develop more sophisticated model
evaluation procedures for social systems in verifying the SD-based
socio-hydrological model, as well as to identify more common
105
causal relationships between water and human systems for broader
applications of the SD-based socio-hydrological framework.
Acknowledgements
This work was supported by the Korea Institute of Planning and
Evaluation for Technology in Food, Agriculture, Forestry and Fish-
eries (IPET) through the Agri-Bio industry Technology Development
Program, funded by the Ministry of Agriculture, Food and Rural
Affairs (MAFRA) (grant number 114060-3). Partial funding was also
provide by an NSERC Discovery Grant held by Jan Adamowski.
References
Adamowski, J., Adamowski, K., Bougadis, J., 2010. Influence of trend on short
duration design storms. Water Resour. Manage. 24, 401–413, http://dx.doi.org/
10.1007/s11269-009-9452-z.
Ahmad, S., Prashar, D., 2010. Evaluating municipal water conservation policies
using a dynamic simulation model. Water Resour. Manage. 24 (13),
3371–3395, http://dx.doi.org/10.1007/s11269-010-9611-2.
Ahmad, S., Simonovic, S.P., 2004. Spatial system dynamics: new approach for
simulation of water resources systems. J. Comput. Civ. Eng. 18 (4), 331–340,
http://dx.doi.org/10.1061/(ASCE)0887-3801(2004)18:4(331).
Alcamo, J., Doll, P., Kaspar, F., Siebert, S., 1997. Global Change and Global Scenarios
of Water Use and Availability: An Application of Water GAP 1.0. University of
Kassel, Kassel Germany.
Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. FAO Irrigation and Drainage
Paper 56, Crop Evapotranspiration: Guidelines for Computing Crop Water
Requirements. Food and Agriculture Organization of the United Nations, Rome.
Barlas, Y., 1996. Formal aspects of model validity and validation in system
dynamics. Syst. Dynam. Rev. 12 (3), 183–210, http://dx.doi.org/10.1002/
(SICI)1099-1727(199623)12:3<183:AID-SDR103>3.0.CO;2-4.
Belayneh, A., Adamowski, J., Khalil, B., Ozga-Zielinski, B., 2014. Long-term SPI
drought forecasting in the Awash River Basin in Ethiopia using
wavelet-support vector regression models. J. Hydrol. 508, 418–429, http://dx.
doi.org/10.1016/j.jhydrol.2013.10.052.
Butler, C., Adamowski, J., 2015. Empowering marginalized communities in water
resources management: addressing inequitable practices in Participatory
Model Building. J. Environ. Manage. 153, 153–162, http://dx.doi.org/10.1016/j.
jenvman.2015.02.010.
Chang, H., Thiers, P., Netusil, N., Yeakley, J.A., Rollwagon-Bollens, G., Bollens, S.M.,
2014. Relationship between environmental governance and water quality in a
growing metropolitan area of the Pacific Northwest, USA. Hydro. Earth Syst.
Sci. 18, 1383–1395, http://dx.doi.org/10.5194/hess-18-1383-2014.
Chavez, A., Rodas, K., Prado, B., Thompson, R., Jimenez, B., 2012. An evaluation of
the effects of changing wastewater irrigation regime for the production of
alfalfa (Medicago sativa). Agric. Water Manage. 113, 76–84, http://dx.doi.org/
10.1016/j.agwat.2012.06.021.
Davies, E.G.R., Simonovic, S.P., 2011. Global water resources modeling with an
integrated model of the social-economic-environmental system. Adv. Water
Resour. 34 (6), 684–700, http://dx.doi.org/10.1016/j.advwatres.2011.02.010.
DeFries, R., Eshleman, K.N., 2004. Land-use change and hydrologic processes: a
major focus for the future. Hydrol. Process. 18, 2183–2186, http://dx.doi.org/
10.1002/hyp.5584.
Di Baldassarre, G., Viglione, A., Carr, G., Kuil, L., Salinas, J.L., Bloschl, G., 2013.
Socio-hydrology: conceptualising human-flood interactions. Hydrol. Earth
Syst. Sci. 17, 3295–3303, http://dx.doi.org/10.5194/hess-17-3295-2013.
Elshafei, Y., Sivapalan, M., Tonts, M., Hipsey, M.R., 2014. A prototype framework for
models of socio-hydrology: identification of key feedback loops and
parameterisation approach. Hydro. Earth Syst. Sci. 18, 2141–2166, http://dx.
doi.org/10.5194/hess-18-2141-2014.
Elshorbagy, A., Ormsbee, L., 2006. Object-oriented modeling approach to surface
water quality management. Environ. Model. Softw. 21, 689–698, http://dx.doi.
org/10.1016/j.envsoft.2005.02.001.
Foley, J.A., Ramankutty, N., Brauman, K.A., Cassidy, E.S., Gerber, J.S., Johnston, M.,
Mueller, N.D., O’Connell, C., Ray, D.K., West, P.C., Balzer, C., Bennett, E.M.,
Carpenter, S.R., Hill, J., Monfreda, C., Polasky, S., Rockstrom, J., Sheehan, J.,
Siebert, S., Tilman, G.D., Zaks, D.P.M., 2011. Solutions for a cultivated planet.
Nature 478, 337–342, http://dx.doi.org/10.1038/nature10452.
Ford, A., 2009. Modeling the Environment. Island Press, Washington, D.C.
Forrester, J., 1961. Industrial Dynamics. The MIT Press, Cambridge, MA and Wiley,
New York.
Ghashghaei, M., Bagheri, A., Morid, S., 2013. Rainfall-runoff modeling in a
watershed scale using an object oriented approach based on the concepts of
system dynamics. Water Resour. Manage. 27, 5119–5141, http://dx.doi.org/10.
1007/s11269-013-0457-2.
Guo, H.C., Liu, L., Huang, G.H., Fuller, G.A., Zou, R., Yin, Y.Y., 2001. A system
dynamics approach for regional environmental planning and management: a
study for the Lake Erhai Basin. J. Environ. Manage. 61, 93–111, http://dx.doi.
org/10.1006/jema.2000.0400.
Haidary, A., Amiri, B.J., Adamowski, J., Fohrer, N., Nakane, K., 2013. Assessing the
impacts of four land use types on the water quality of wetlands in Japan. Water
106 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Resour. Manage. 27, 2217–2229, http://dx.doi.org/10.1007/s11269-013-0284-
5.
Halbe, J., Pahl-Wostl, C., Sendzimir, J., Adamowski, J., 2013. Towards adaptive and
integrated management paradigms to meet the challenges of water
governance. Water Sci. Technol. 67, 2651–2660, http://dx.doi.org/10.2166/wst.
2013.146.
Halbe, J., Adamowski, J., Bennett, E., Pahl-Wostl, C., Farahbakhsh, K., 2014.
Functional organization analysis for the design of sustainable engineering
systems. Ecol. Eng. 73, 80–91, http://dx.doi.org/10.1016/j.ecoleng.2014.08.011.
Hamilton, A.J., Stagnitti, F., Xiong, X., Kreidil, S.L., Benke, K.K., Maher, P., 2007.
Wastewater irrigation: the state of play. Vadose Zone J. 6, 823–840, http://dx.
doi.org/10.2136/vzj2007.0026.
Hjorth, P., Bagheri, A., 2006. Navigating towards sustainable development: a
system dynamics approach. Futures 38, 74–92, http://dx.doi.org/10.1016/j.
futures.2005.04.005.
Inam, A., Adamowski, J., Halbe, J., Prasher, S., 2015. Using causal loop diagrams for
the initialization of stakeholder engagement in soil salinity management in
agricultural watersheds in developing countries: a case study in the Rechna
Doab watershed, Pakistan. J. Environ. Manage. 152, 251–267, http://dx.doi.org/
10.1016/j.jenvman.2015.01.052.
Jang, T.I., Lee, S.B., Sung, C.H., Lee, H.P., Park, S.W., 2010. Safe application of
reclaimed water reuse for agriculture in Korea. Paddy Water Environ. 8,
227–233, http://dx.doi.org/10.1007/s10333-010-0203-9.
Jang, T.I., Kim, H.K., Seong, C.H., Lee, E.J., Park, S.W., 2012. Assessing nutrient losses
of reclaimed wastewater irrigation in paddy fields for sustainable agriculture.
Agric. Water Manage. 104, 235–243, http://dx.doi.org/10.1016/j.agwat.2011.
12.022.
Jeong, H., Jang, T., Seong, C., Park, S., 2014. Assessing nitrogen fertilizer rates and
split applications using the DSSAT model for rice irrigated with urban
wastewater. Agric. Water Manage. 141, 1–9, http://dx.doi.org/10.1016/j.agwat.
2014.04.009.
Jeong, H., Kim, H., Jang, T., Park, S., 2016. Assessing the effects of indirect
wastewater reuse on paddy irrigation in the Osan River watershed in Korea
using the SWAT model. Agric. Water Manage. 163, 393–402, http://dx.doi.org/
10.1016/j.agwat.2015.08.018.
Jimenez, B., Drechsel, P., Kone, D., Bahri, A., Raschid-Sally, L., Qadir, M., 2010.
Wastewater sludge and excreta use in developing countries: an overview. In:
Drechsel, P., Scott, C.A., Raschid-Sally, L., Redwood, M., Bahri, A. (Eds.),
Wastewater Irrigation and Health: Assessing and Mitigating Risk in
Low-Income Countries. Earthscan, International Development Research Centre
(IDRC), International Water Management Institute (IWMI), London,
UK/Ottawa, Canada/Colombo, Sri Lanka, pp. 3–27, Available at: http://www.
iwmi.cgiar.org/Publications/Books/PDF/Wastewater Irrigation and Health
book.pdf. (accessed 04.01.16.).
Kandasamy, J., Sounthararajah, D., Sivabalan, P., Chanan, A., Vigneswaran, S.,
Sivapalan, M., 2014. Socio-hydrologic drivers of the pendulum swing between
agricultural development and environmental health: a case study from
Murrumbidgee River basin, Australia. Hydro. Earth Syst. Sci. 18, 1027–1041,
http://dx.doi.org/10.5194/hess-18-1027-2014.
Khan, S., Yufeng, L., Ahmad, A., 2009. Analysing complex behaviour of hydrological
systems through a system dynamics approach. Environ. Model. Softw. 24,
1363–1372, http://dx.doi.org/10.1016/j.envsoft.2007.06.006.
Kim, S.M., Im, S.J., Park, S.W., Lee, J.J., Benham, B.L., Jang, T.I., 2008. Assessment of
wastewater reuse effects on nutrient loads from paddy field using field-scale
water quality model. Environ. Model. Assess. 13, 305–313, http://dx.doi.org/10.
1007/s10666-007-9093-7.
Kim, H.K., Jang, T.I., Im, S.J., Park, S.W., 2009. Estimation of irrigation return flow
from paddy fields considering the soil moisture. Agric. Water Manage. 96,
875–882, http://dx.doi.org/10.1016/j.agwat.2008.11.009.
Kolinjivadi, V., Adamowski, J., Kosoy, N., 2014. Recasting payments for ecosystem
services (PES) in water resource management: a novel institutional approach.
Ecosyst. Serv. 10, 144–154.
Lambin, E.F., Turner, B.L., Geist, H.J., Agbola, S.B., Angelsen, A., Bruce, J.W., Coomes,
O.T., Dirzo, R., Fischer, G., Folke, C., George, P.S., Homewood, K., Imbernon, J.,
Leemans, R., Li, X., Moran, E.F., Mortimore, M., Ramakrishnan, P.S., Richards,
J.F., Skanes, H., Steffen, W., Stone, G.D., Svedin, U., Veldkamp, T.A., Vogel, C., Xu,
J., 2001. The causes of land-use and land-cover change: moving beyond the
myths. Global Environ. Chang. 11 (4), 261–269, http://dx.doi.org/10.1016/
S0959-3780(01)00007-3.
Lee, J., Park, G., Yoo, C., Kim, S., Yoon, J., 2010. Effects of land use change and water
reuse options on urban water cycle. J. Environ. Sci. 22 (6), 923–928, http://dx.
doi.org/10.1016/S1001-0742(09)60199-6.
Li, L., Simonovic, S.P., 2002. System dynamics model for predicting floods from
snowmelt in North American prairie watersheds. Hydrol. Process. 16,
2645–2666, http://dx.doi.org/10.1002/hyp.1064.
Liu, Y., Tian, F., Hu, H., Sivapalan, M., 2014. Socio-hydrologic perspectives of the
co-evolution of humans and water in the Tarim River basin, Western China:
the Taiji-Tire model. Hydro. Earth Syst. Sci. 18, 1289–1303, http://dx.doi.org/
10.5194/hess-18-1289-2014.
Ministry of Land, Transport and Maritime Affairs (MLTM), 2011. Long-term Plans
for Water Resources (2011–2020), Gwacheon, Korea.
Madani, K., Marino, M.A., 2009. System dynamics analysis for managing Iran’s
Zayandeh-Rud River basin. Water Resour. Manage. 23, 2163–2187, http://dx.
doi.org/10.1007/s11269-008-9376-z.
Mirchi, A., Madani, K., Watkins, D., Ahmad, S., 2012. Synthesis of system dynamics
tools for holistic conceptualization of water resources problems. Water Resour.
Manage. 26 (9), 2421–2442, http://dx.doi.org/10.1007/s11269-012-0024-2.
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L.,
2007. Model evaluation guidelines for systematic quantification of accuracy in
watershed simulations. Trans. ASABE 50 (3), 885–900.
Nalley, D., Adamowski, J., Khalil, B., 2012. Using discrete wavelet transforms to
analyze trends in streamflow and precipitation in Quebec and Ontario
(1954–2008). Journal of Hydrology 475, 204–228.
Nalley, D., Adamowski, J., Khalil, B., Ozga-Zielinski, B., 2013. Trend detection in
surface air temperature in Ontario and Quebec, Canada during 1967–2006
using the discrete wavelet transform. Atmos. Res. 132/133, 375–398, http://dx.
doi.org/10.1016/j.atmosres.2013.06.011.
Nash, J.E., Sutcliffe, J.V., 1970. River flow forecasting through conceptual models:
part 1. a discussion of principles. J. Hydrol. 10 (3), 282–290, http://dx.doi.org/
10.1016/0022-1694(70)90255-6.
Nasiri, F., Savage, T., Wang, R., Barawid, N., Zimmerman, J.B., 2013. A system
dynamics approach for urban water reuse planning: a case study from the
Great Lakes region. Stoch. Environ. Res. Risk Assess. 27 (3), 675–691, http://dx.
doi.org/10.1007/s00477-012-0631-8.
National Statistical Office (NSO), 2012. Population Projection. National Statistical
Office, Daejeon, Korea, Available at: http://kostat.go.kr/portal/korea/kor nw/2/
2/6/index.board. (accessed 14.02.14.).
Noory, H., van der Zee, S.E.A.T.M., Liaqhat, A.M., Parsinejad, M., van Dam, J.C., 2011.
Distributed agro-hydrological modeling with SWAP to improve water and salt
management of the Voshimghir Irrigation and Drainage. Agric. Water Manage.
98, 1062–1070, http://dx.doi.org/10.1016/j.agwat.2011.01.013.
Osan, 2012. A Master Plan for Improving the Sewerage Systems, Osan-si,
Gyeonggi-do Korea.
Park, D., Roesner, L.A., 2012. Evaluation of pollutant loads from stormwater BMPs
to receiving water using load frequency curves with uncertainty analysis.
Water Res. 46, 6881–6890, http://dx.doi.org/10.1016/j.watres.2012.04.023.
Qi, C., Chang, N.B., 2011. System dynamics modeling for municipal water demand
estimation in an urban region under uncertain economic impacts. J Environ.
Manage. 92, 1628–1641, http://dx.doi.org/10.1016/j.jenvman.2011.01.020.
Qudrat-Ullah, H., 2012. On the validation of system dynamics type simulation
models. Telecommun. Syst. 51, 159–166, http://dx.doi.org/10.1007/s11235-
011-9425-4.
Rathinasamy, M., Adamowski, J., Khosa, R., 2013. Multiscale streamflow forecasting
using a new Bayesian model average based ensemble multi-wavelet Volterra
nonlinear method. J. Hydrol. 507, 186–200, http://dx.doi.org/10.1016/j.jhydrol.
2013.09.025.
Rose, S., Peters, N.E., 2001. Effects of urbanization on streamflow in the Atlanta
area (Georgia, USA): a comparative hydrological approach. Hydrol. Process. 15
(8), 1441–1457, http://dx.doi.org/10.1002/hyp.218.
Santhi, C., Arnold, J.G., Williams, J.R., Dugas, W.A., Srinivasan, R., Hauck, L.M., 2001.
Validation of the SWAT model on a large river basin with point and nonpoint
sources. J. Am. Water Resour. Assoc. 37 (5), 1169–1188, http://dx.doi.org/10.
1111/j.1752-1688.2001.tb03630.x.
Saysel, A.K., Barlas, Y., 2001. A dynamic model of salinization on irrigated lands.
Ecol. Model. 139, 177–199, http://dx.doi.org/10.1016/S0304-3800(01)00242-3.
Simonovic, S.P., Fahmy, H., 1999. A new modeling approach for water resources
policy analysis. Water Resour. Res. 35 (1), 295–304, http://dx.doi.org/10.1029/
1998WR900023.
Simonovic, S.P., Fahmy, H., El-Shorbagy, A., 1997. The use of object-oriented
modeling for water resources planning in Egypt. Water Resour. Manage. 11,
243–261, http://dx.doi.org/10.1023/A:1007988424353.
Singh, J., Knapp, H.V., Demissie, M., 2004. Hydrologic modeling of the Iroquois
River watershed using HSPF and SWAT, ISWS CR 2004-08, Champaign, Ill.:
Illinois State Water Survey, Available at: www.sws.uiuc.edu/pubdoc/CR/
ISWSCR2004-08.pdf. (accessed 05.02.14.).
Sivapalan, M., Savenije, H.H.G., Bloschl, G., 2012. Socio-hydrology: a new science of
people and water. Hydrol. Process. 26 (8), 1270–1276, http://dx.doi.org/10.
1002/hyp.8426.
Srinivasan, V., Lambin, E., Gorelick, S., Thompson, B., Rozelle, S., 2012. The nature
and causes of the global water crisis: syndromes from a meta-analysis of
coupled human-water studies. Water Resour. Res. 48, W10516, http://dx.doi.
org/10.1029/2011WR011087.
Stave, K.A., 2003. A system dynamics model to facilitate public understanding of
water management options in Las Vegas. Nevada. J. Environ. Manage. 67,
303–313, http://dx.doi.org/10.1016/S0301-4797(02)00205-0.
Straith, D., Adamowski, J., Reilly, K., 2014. Exploring the attributes, strategies and
contextual knowledge of champions of change in the Canadian water sector.
Can. Water Resour. J. 39 (3), 255–269, http://dx.doi.org/10.1080/07011784.
2014.942576.
Takeda, I., Fukushima, A., 2006. Long-term changes in pollutant load outflows and
purification function in a paddy field watershed using a circular irrigation
system. Water Res. 40 (30), 569–578, http://dx.doi.org/10.1016/j.watres.2005.
08.034.
Troy, T.J., Pavao-Zuckerman, M., Evans, T.P., 2015. Debates—perspectives on
socio-hydrology: socio-hydrologic modeling: tradeoffs, hypothesis testing, and
validation. Water Resour. Res. 51, 4806–4814, http://dx.doi.org/10.1002/
2015WR017046.
Tuppad, P., Kannan, N., Srinivasan, R., Rossi, C.G., Arnold, J.G., 2010. Simulation of
agricultural management alternatives for watershed protection. Water Resour.
Manage. 24 (12), 3115–3144, http://dx.doi.org/10.1007/s11269-010-9598-8.
 H. Jeong, J. Adamowski / Agricultural Water Management 171 (2016) 89–107
Valipour, M., 2014a. Future of the area equipped for irrigation. Arch. Agron. Soil Sci.
60 (12), 1641–1660, http://dx.doi.org/10.1080/03650340.2014.905675.
Valipour, M., 2014b. Drainage, waterlogging, and salinity. Arch. Agron. Soil Sci. 60
(12), 1625–1640, http://dx.doi.org/10.1080/03650340.2014.905676.
Valipour, M., 2015. Future of agricultural water management in Africa. Arch. Agron.
Soil Sci. 61 (7), 907–927, http://dx.doi.org/10.1080/03650340.2014.961433.
Venkatesan, A.K., Ahmad, S., Johnson, W., Batista, J.R., 2011. Salinity reduction and
energy conservation in direct and indirect potable water reuse. Desalination
272, 120–127, http://dx.doi.org/10.1016/j.desal.2011.01.007.
Viglione, A., Di Baldassarre, G., Brandimarte, L., Kuil, L., Carr, G., Salinas, J.L.,
Scolobig, A., Bloschl, G., 2014. Insights from socio-hydrology modelling on
dealing with flood risk -roles of collective memory, risk-taking attitude and
trust. J. Hydrol. 518 (Part A), 71–82, http://dx.doi.org/10.1016/j.jhydrol.2014.
01.018.
Wagener, T., Sivapalan, M., Troch, P.A., McGlynn, B.L., Harman, C.J., Gupta, H.V.,
Kumar, P., Rao, S.C., Basu, N.B., Wilson, J.S., 2010. The future of hydrology: an
evolving science for a changing world. Water Resour. Res. 46, W05301, http://
dx.doi.org/10.1029/2009wr008906.
107
Wei, S., Yang, H., Song, J., Abbaspour, K.C., Xu, Z., 2012. System dynamics
simulation model for assessing socio-economic impacts of different levels of
environmental flow allocation in the Weihe River Basin. China. Eur. J. Oper.
Res. 221 (1), 248–262, http://dx.doi.org/10.1016/j.ejor.2012.03.014.
Wescoat, J.L., 2013. Reconstructing the duty of water: a study of emergent norms
in socio-hydrology. Hydrol. Earth Syst. Sci. 17, 4759–4768, http://dx.doi.org/
10.5194/hess-17-4759-2013.
World Health Organization (WHO), 2006. Guidelines for the Safe Use of
Wastewater, Excreta and Greywater. World Health Organization, Geneva.
Yaeger, M.A., Housh, M., Cai, X., Sivapalan, M., 2014. An integrated modeling
framework for exploring flow regime and water quality changes with
increasing biofuel crop production in the U.S. Corn Belt. Water Resour. Res. 50
(12), 9385–9404, http://dx.doi.org/10.1002/2014WR015700.
Zlinszky, A., Timar, G., 2013. Historic maps as a data source for socio-hydrology: a
case study of the Lake Balaton wetland system, Hungary. Hydro. Earth Syst. Sci.
17, 4589–4606, http://dx.doi.org/10.5194/hess-17-4589-2013.