Download Application of System Dynamics and DPSIR framework for

Application of System Dynamics and DPSIR framework for
sustainability assessment of urban residential areas
Xu Zhao
PhD student
DITAG- land Environment and
Geo-Engineering Department,
Politecnico di Torino
Italy
[email protected]
Marta Bottero
Dr.
DITAG- land Environment and
Geo-Engineering Department,
Politecnico di Torino
Italy
[email protected]
Summary
Constructing and improving urban residential areas is an eternal critical subject in the process of
the whole territorial development which is connected with a series of challenges and problems
such as population pressure, environmental pollution, public safety, etc.
In this paper, DPSIR (Driving Forces-Pressure-State-Impact-Response) framework has been
employed for better systematizing the sustainability indicators of urban residential areas. Due to
the urban activities cause impacts not only on local level but also a broader scale, a simulation
model, using system dynamics (SD) methodology, is structured to quantitatively investigate the
developmental tendency of the indicators. The integration of system dynamics (SD) and DPSIR
framework can better explain the interaction and the variation with time of the sustainable
indicators of urban residential areas. Hence it’s able to support the Decision Maker to view the
sustainable level of urban residential areas more comprehensively. At last the paper will consider
an application to a real case concerning the development of a residential area in China.
Key words: DPSIR framework, System Dynamics, urban residential areas, sustainable
development
1. Introduction
Urban residential areas show increasing signs of sustainable problems. Major concerns for cities,
especially in the emerging countries where residential areas are expending very quickly, are the
quality of air, traffic jam, water conditions, limited land resources and related social-economy
problems. Furthermore, with the rapid economic growth, people are demanding more and more in
the quality of their residence and living level.
The concept of sustainable development has put in evidence the impossibility of separating
environment and development; for this reason, sustainability assessment should be based on
multidisciplinary approaches that reflect the complex network of interactions between human and
environmental systems.
Many researchers have studied the issue of sustainability assessment from different perspectives
with various evaluation methodologies. For example, Nessa and Montserrat (2007) summarized
the conceptualization of sustainable development and housing from the international statements
and contended that housing is an underdeveloped indicator and called for more attention to be
paid to the importance of aspects of housing for sustainable development and the measurement of
progress towards it via social indicators. Bidone and Lacerda (2003) developed a Driving ForcesPressure-State-Impact-Response (DPSIR) framework in a practical context to integrate natural and
social-economic indicators. They also discussed the strengths and limitations of cost-benefit
analysis (CBA) included in DPSIR which is used to evaluate losses and benefits resulting from
policies. Shen et al., (2009) applied a system dynamics model for the sustainable land use and
urban development in HongKong. They utilized the model to test the outcomes of development
policy scenarios and make forecasts.
In this paper, we try to integrate the System Dynamics (SD) methodology and DPSIR framework to
evaluate the sustainable development of urban residential areas. System dynamics model can be
used to explore the complex and inter-dependent relationships between the various indicators
within social-economic environment and DPSIR framework can been employed for better
systematizing the sustainability indicators. The study is aimed at identifying and calculating the
quantitative criteria of sustainable level of urban residential areas. And the results analysis and
discussion are expected to give some suggestions to support the Decision Maker to view the
sustainable level of urban residential areas more comprehensively.
2. Sustainable development indicators and urban residential areas
Since the Earth Summit in 1991, many countries and international organizations have been
working on indicators about sustainable development. And sustainable development is increasingly
linked with the concepts of quality of life, well-being and livability (Michalos 1997; Moore and Scott
2005; Low Choy, 2004, 2005). However, housing is one of the more neglected aspects of
sustainability and the availability of housing indicators in international sustainable development
indicator sets is extremely limited, despite relatively advanced sustainable development policies
which refer to the importance of housing. (Nessa, W., & Montserrat, P.E. 2008)
• The Organization for Economic Cooperation and Development (OECD) has been working on
environmental indicators since 1989. The work of OECD mainly focuses on indicators that have
to be used in national, international and global decision making; furthermore the approach may
also be used to develop indicators at a sub-national or ecosystem level (OECD, 2003);
• The United Nations Commission on Sustainable Development (UNCSD) produces some
indicator sets in the field of sustainability assessment. UNCSD provides a very useful and timely
forum for the discussion of national-level indicators with the involvement of governments,
international organizations and various stakeholders. UNCSD indicators play an important role
in “helping countries make informed decisions concerning sustainable development”, and they
are “applied and used in many countries as the basis for the development of national indicators
of sustainable development” (United Nations, 2007);
• The European System of Social Indicators (EUSI) is part of a cross-national European project
which began in the late 1990s and which aims at monitoring and assessing welfare
development and social change in Europe (Berger-Schmitt & Noll, 2000). The EUSI framework
links sustainability to other welfare concepts, such as social cohesion, social exclusion, social
capital and quality of life;
• The Chinese Academy of Sciences, through the Chinese Urban Development Centre (CUDC),
has proposed an indicator system for urban sustainability assessment (CUDC, 2002). This
system concentrates on the strategic context, strategy objective, strategy mission and the
strategy design of sustainable urban development in China.
As far as residential urban areas are concerned, the above mentioned sustainability indicator
systems focus on different issues, such as environmental quality, well-being of the population,
economic aspects, etc. We could select a representation of the indicators that are available for the
assessment of the sustainable development of residential urban areas, according to the main
dimensions that have been identified: housing, economy, environment and society.
3. Methodology
Environmental indicators and evaluation methods should be incorporated in sustainability
assessment. Qualitative analysis of environmental indicators was based on an index system
established in this study, while evaluation methods was discussed based on DPSIR framework and
System dynamics (Fig. 1). The two aspects were combined to obtain a comprehensive evaluation
system.
3.1
System dynamics (SD) model
This paper applies system dynamics theories to
formulate, simulate and validate the sustainable
development of urban residential areas.
System dynamics is a methodology and computer
simulation modeling technique for framing, understanding
and discussing complex issues and problems [1]. And it
is widely used to gain understanding of a system with
complex, dynamic and nonlinearly interacting variables
[2].
System dynamics has four elements defined within the
system: (a) stock, (b) flow, (c) converter, and (d)
connector (HPS, 1997; Mohapatra, 1994), as shown in
Fig. 2. A stock collects all those in-flows and also serves
as the source from where out-flows come. A flow serves
as a vehicle to deliver information to or drain information
from the stock. The connector is the information
transmitter connecting elements [2].
This methodology has been applied into various fields,
including but not confined to, global system of mobile
telecommunication (Tung & Claudia, 1996), the
interaction among various water cycle components
(Shahbaz et al., 2007), urban transportation system and
its application (Wang et al., 2008), forecasting demand
and evaluating policy scenarios (Erma et al., 2010),
building design and operation (Benjamin & Lawrence,
2009), the sustainable land use planning and
development (Shen et al., 2009), sustainable performance of construction projects (Shen et al.,
2004).
3.2
DPSIR framework
The DPSIR framework is able to illustrate the complexities of the system interactions in sustainable
development and urban residential areas. The framework can be summarized as follows:
• The Driving Forces are processes and anthropogenic activities that able to cause pressures
during the development of urban residential areas; in other words, they are the reasons that
cause the changes in the development process.
• The Pressures are the direct stresses, deriving from the anthropogenic activities and affecting
the environment, economic and society.
• The State reflects the actual conditions of urban residential areas;
• The Impact is the measure of the environmental effects due to the development of urban
residential areas;
• The Response refers to specific actions oriented to reduce the pressure and promote
development in terms of economic or administrative measures.
4. The integration of System Dynamics and DPSIR framework in
sustainability assessment
The procedure which aims at integrating System Dynamics and DPSIR indicators can be can be
divided into 6 steps, as follows:
Step1. Selection of indicators
The availability and reliability of data, the usability of the available data within DPSIR framework,
and the sensitivity of indicators to reflect the underlying social and economical processes were
used as the criteria to establish the indicators system proposed in this work. The indicators system
contains different kinds of indicators including social, economic, environmental and housing
indicators. From indicators identified in the four indicator-systems, we selected a synthetical and
concise indicators system which is suitable for dealing with sustainability assessment of urban
residential areas. Table 1 represents the 24 indicators selected for the application of the
sustainability assessment of urban residential areas, according to the DPSIR framework.
In this way, the overall system is described as different layers: categories of the DPSIR framework,
thematic areas and indicators. The structure here illustrated is represented in Fig 3.
Table 1 Assessment indicators for sustainable development of urban residential areas
Thematic Area
Housing
Economy
Environment
Society
Indicator
• The floor space completed of residential housing;
• Living space per capita;
• Dwellings in deficient state of repair;
DPSIR Category
S
S
S
•
•
•
•
The total demand of the residential houses;
The total supply of the residential houses;
GDP;
Disposable income per capita;
D
R
D
D
•
Land price per floor area
P
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Average urban housing price;
Average rent price;
Exploitation and investment of real estate;
Crime in residential area;
Environmental pollution abatement and control expenditure;
Urban air emissions (SOx,NOx,VOC);
Ambient water conditions in urban areas;
Generation of waste;
Green coverage ratio;
Share of renewable energy sources in total energy use.
Urban population;
Urban infrastructure;
Dependency rate;
Road traffic volumes.
Proportion of population living below national poverty level;
Stock of road vehicles.
P
P
R
P
R
I
I
I
I
R
D
D
P
P
S
S
Step2. Determination weight of indicators
The weight of each indicator was determined using Multicriteria Analysis (MCA). The Analytic
Hierarchy Process (AHP) method (Saaty, 1980) has been developed. On the basis of the
knowledge about sustainability assessment of urban residential areas, it is possible, using the AHP,
to discuss the weight of each indicator by consulting the opinions from experts. Judgment matrixes
have been established. For example, Table 2 represents the judgment matrix for the comparison of
the importance of the different DPSIR categories for the assessment of the sustainability of urban
areas. The process has been repeated for the whole assessment system and a relative stable
weight result has been reached. Table 3 represents the indicator system structure; the values in
the square brackets reflect the weight of the element with reference to the overall system.
Fig. 3 The structure of DPSIR framework
Table 2 Pairwise comparison matrix for DPSIR category assessment.
D
P
S
I
R
D
1
1/2
1/3
1/3
2
P
2
1
1/2
1/2
3
S
3
2
1
1
5
I
3
2
1
1
5
R
1/2
1/3
1/5
1/5
1
Weight
0.25
0.15
0.08
0.08
0.44
Table 3 The sustainable indicator system in the DPSIR framework.
System
DPSIR categories
Thematic areas
Housing [0.25]
Driving forces
[0.25]
Economy [0.5]
Society [0.25]
Environment [0.33]
Society [0.33]
Pressure [0.15]
Sustainable
development
of urban
residential
areas[1]
Economy [0.34]
Housing [0.75]
State [0.08]
Society [0.25]
Impact [0.08]
Environment [1]
Environment [0.34]
Responses [0.44]
Economy [0.33]
Housing [0.33]
Indicators
1: The total demand of the residential houses [1]
2: GDP [0.5]
3: Urban disposable income per capita [0.5]
4: Urban population [0.5]
5: Urban infrastructures [0.5]
6: Crime in residential areas [1]
7: Dependency ratio [0.5]
8: Road traffic volumes [0.5]
9: Land price per floor area [0.25]
10: Average urban housing price [0.5]
11: Average rent price [0.25]
12: The floor space completed of residential housing [0.24]
13: Living space per capita [0.53]
14: Dwellings in deficient state of repair [0.23]
15: Stock of road vehicles [0.25]
16: Proportion of population living below national poverty level [0.75]
17: Urban air emissions [0.28]
18: Generation of waste [0.28]
19: Ambient water conditions in urban areas [0.28]
20: Green coverage ratio [0.16]
21: Share of renewable energy sources in total energy use [1]
22: Environmental pollution abatement and control expenditure [0.4]
23: Exploitation and investment of real estate [0.6]
24: The total supply of the residential houses [1]
Step3. Calculation of the indicator value
The assessment model should include two basic function parts: evaluation and prediction. The two
basic function parts should be related, respectively, to historical data (previous indicators) and
estimate values (future indicators). As far as the environmental indicators are concerned, it is
possible to obtain the information required by referring to historical data derived from statistical
yearbooks, social-economic development reports, environmental observations, recordings, etc. As
far as the housing-economy-society indicators are concerned, it is possible to estimate the values
in System Dynamics model.
In this paper, the System Dynamics model has four level variables, Urban population, Urban GDP,
the total demand of the residential houses and the total supply of the residential houses; Six rate
variables, the population growth, the economy growth, the demand growth, the supply growth, the
actual demand and the supply completed.
We can establish the System Dynamics model by the software “vensim” (Fig 4).
Fig. 4 Stock-flow diagram of housing-economy-society indicators in “vensim”
Step4. Standardization of the indicator value
In order to take into consideration the positive and negative values of the indicators, it is necessary
to calculate the standard value of each indicator as in equation (1):
(1)
Where xi is the standard value of indicator i in the temporal period considered for the analysis, x is
the actual value of indicator i in each part of the considered period, x is the average value of
indicator i in the period and s is the standard deviation of the indicator i in the period.
Step5. Calculation results
On the basis of steps 2 and 3, it is possible to obtain the value of the subsystems layer related to
the DPSIR categories, as in equation (2):
n
m
j =1
i =1
Yk = ∑ σ j (∑ wi xi )
(2)
where Yk is the value of the subsystem layer related to category k of the DPISR framework, σj is
the weight of thematic area j corresponding to Yk, n is the number of the thematic areas under Yk,
wi is the weight of indicator i, m is the number of indicators under thematic area j and xi is the
standard value of indicator i.
The final value of the sustainability level of the system is derived from the weighted sum of the five
subsystems k, as in equation (3):
5
Z = ∑ μ kYk
(3)
k =1
Where Z is the final value of sustainability of the system, μk is the weight of category k of the
DPSIR framework and Yk is the value of category k.
It is possible to observe that Z is a composite index that results from the value of the Driving
Forces, Pressure, State, Impact and Response categories; furthermore, the value of Z is included
in the (0, 1) domain.
Step6. Results analysis
After the calculations made in step5, it is possible to obtain a set of values of the composite index
years of the period Z and of the Driving forces, Pressures, State, Impact or Response subsystems
for the different considered in the analysis, as illustrated in Table 4 where n represents the single
year in the considered period.
Table 4 Assessment results of the sustainability level in the considered period (year 1 – year n+2).
Driving forces
Pressure
State
Impact
Response
Z
1
YD1
YP1
YS1
YI1
YR1
Z1
2
YD2
YP2
YS2
YI2
YR2
Z2
…
…
…
…
…
…
…
n
YDn
YPn
YSn
YIn
YRn
Zn
It is possible to observe that when Z is very low, it will be difficult to continue a sustainable urban
development, because some indicators indicate a bad performance (for example GDP or urban
infrastructures). However, the previous consideration does not mean that the higher the value of Z,
the higher the sustainability level of the system. According to the environmental impact
assessment classification (Chen and Chen, 1992), a five-grade classification is utilized to classify
the sustainability of residential urban areas (Table 5).
Table 5 A five-grade classification of the sustainability of residential urban areas.
In other words:
• Z is very large in grade I. This shows that in the social, economic and environmental aspects the
residential urban areas are developing excessively. For example, the excessive emphasis on
the environmental quality of residential areas can lead the actual demands of the inhabitants
being ignored; this is particularly true in emerging countries, where the need to raise income
and welfare is higher that the need to have green spaces.
• Z has a reasonable value in grade II and III. This means that the residential urban areas are
developing at a steady rate (grade III) or quickly (grade II). This is the optimum condition for
sustainable development.
• Z has a smaller value in grade IV than in grades II and III. In this case, the sustainable
development of residential urban areas is able to bear the pressure. The situation can be
ameliorated by a sequence of special measures adopted to strengthen environmental, social
and economic development.
• Z has the smallest value in grade V. At this point, the sustainable development of residential
urban areas is bearing great pressure. It can be observed that the social-economic
sustainability and environmental sustainability confront a major obstacle.
5. Application to a Chinese case study
The considered study case refers to the city of Nanjing which is the provincial capital of JiangSu in
China. The urban area of the Nanjing megalopolis has been growing rapidly, from 2599 km2 in
2001 to 4723 km2 in 2008. The population has registered a great variation over the years, from
3.71 million inhabitants in 2001 to 5.41 million inhabitants in 2008. Sustainability assessment of the
Nanjing urban residential areas has been performed using the 24 indicators which have been
observed through a period of 10 years (2001-2010).
5.1 Model Simulation
Table 6 gives a short description of the 24 indicators used in the application.
Table 6 Description of the quantitative indicators used in the application
1
2
3
Indicators
The total demand of the
residential houses
GDP
4
Urban disposable income
per capita
Natural population
5
6
Urban infrastructures
Crime in residential area
7
Dependency ratio
8
Road traffic volumes
9
10
Land price per floor area
Average urban housing
price
Description
It indicates the total floor areas which people demand in one year.
Unit of measure
million m2
GDP stands for Gross domestic product and it reflects the sum of
private consumptions, gross investments, government spending
and exports, while the imports are subtracted.
Urban disposable income per capita is calculated by adding up all
sources of income and subtracting current taxes.
This represents the births and deaths in the population of a country
or city. It takes into account migration.
This represents the investments in urban infrastructures in a year.
This is indicated by the number of criminal registered cases per unit
of 10000 people per one year.
This represents an age-population ratio who are usually not in the
labor force who registered at an employment agency and those
who are usually in the labor force.
This aims at measuring the urban traffic condition and it is
represented by the number of public transportation vehicles per unit
of 10000 people.
It is measured by the ratio of basic land price and average plot ratio
This is the ratio of housing prices and the basic price in 2001.
billion
yuan(RMB)
million
billion
n/y
%
n/10000 p
yuan/ m2
yuan/ m2
11
Average rent price
12
The floor space
completed of residential
housing
Living space per capita
Dwellings in deficient
state of repair
Stock of road vehicles
13
14
15
16
Proportion of population
living below national
poverty level
17
Urban air emissions
18
19
Generation of waste
Ambient water conditions
in urban areas
Green coverage ratio
20
21
Share of renewable
energy sources in the
total energy use
22
Environmental pollution
abatement and control
expenditure
Exploitation and
investment of real estate
The total supply of the
residential houses
23
24
This is indicated by the growth rate of housing rent. It considers the
rent in 2001 as a basic price.
This is indicated by the floor area completed in one year.
%
This reflects the average level of housing per capita.
This indicates the households or units relocated due to building
demolition.
This is represented by the quantity of possessed family cars per
100 urban households.
This represents the ratio of the population living below the national
poverty and the full city town population. Low-income families are
urban residents whose average family income is lower than the
minimum living standard of NanJing city.
This considers the Air Pollution Index (API). The index has 5
grades, where: grade I (API< 50): the air quality is excellent; Grade
II (50<API<100): the air quality is good; Grade III (100<API<200):
light air pollution exists; Grade IV (200<API<300): medium air
pollution exists; Grade V (API> 300): heavy air pollution exists.
Here the indicator is obtained from the number of days in which the
pollution index attains Grade I and Grade II in a year.
This reflects the domestic waste in a whole year.
This indicates the total urban domestic water consumption volume.
m2
n.
million/m2
n. /100 p
%
d/y
million ton
million m3
This represents the ratio between the green areas in the city and
the overall urban area.
This indicates the energy consumption (standard coal) for every ten
thousand Chinese yuan (CNY) worth of the gross domestic product
(GDP). This is an index on the energy utilization efficiency to reflect
the consumption level and the saving energy and reducing
consumption conditions.
This indicates the complete investment concerning pollution-control
projects.
%
This indicates the amount of investment in real estate development.
billion
It indicates the total floor areas supplied for all people in one year.
million m2
million m2
million
The historical values of the environmental indicators have been derived from specific reports of the
city of Nanjing. With reference to the values of housing-economy-social indicators, these have
been estimated using System Dynamics model. Fig 5 and Fig 6 represents the line chart of level
variables and rate variables.
Table 7 represents the values for the 24 indicators considered in the analysis over the years 20012010. The core formulates of using the software “vensim” in this application are listed in the
Appendix A.
Fig. 5 Line chart of level variables
Fig. 6 Line chart of rate variables
On the basis of the methodology described in section 4, it is possible to calculate the sustainability
level of the subsystems and of the overall system from the values of the indicators of Table 7, for
each year of the considered period (Table 8).
Table 7 Indicator values of the overall system
I1
I2
I3
I4
I5
I6
I7
I8
I9
I10
I11
I12
I13
I14
I15
I16
I17
I18
I19
I20
I21
I22
I23
I24
2001
7.12
121.85
14089.3
3.72
2.92
70
3.59
10.85
1020.83
2550.35
100
3.05
18.93
13057
0.09
0.95
247
1.33
149.42
40
1.8
176.39
10.97
9.72
2002
7.46
134.04
14042.6
3.82
3.08
77
4.13
9.84
1096.08
2737.18
99.4
3.26
19.28
20032
0.3
1.6
215
1
242.77
42.9
1.5
133.12
12.06
10.28
2003
7.88
152.4
10093.3
3.91
3.81
71
4.18
10
1161.2
3132.27
103.8
3.87
19.9
21308
1.6
1.99
297
1.52
138.62
43.51
1.43
162.6
16.61
11.15
2004
7.8
186.08
12209.6
4.02
4.65
68
4.03
9.63
1514.68
3911.78
109
5.17
21.18
13500
2
2.12
295
1.66
144.17
44.46
1.37
233.03
26.24
12.98
2005
7.44
227.57
21891
4.16
6.37
95
3.35
11.4
1531.2
3913.88
109
5.93
22.89
15000
4.88
2.4
304
1.69
154.09
44.94
1.36
205.93
30.04
14.06
2006
9.2
246.69
22912.8
4.31
6.91
90
3.33
13.8
1525.5
3938.54
109.4
6.16
23.45
15000
6.38
2
305
1.62
415.29
45.49
1.31
527.12
31.08
16.04
2007
9.57
305.16
27319.6
4.47
6.71
70
3.3
14.2
1628.4
4148.95
111.4
7.62
26.1
16000
6.63
2
312
16.2
398.14
45.92
1.25
585.91
41.5
18.16
2008
12.03
361.3
31110.3
4.65
10.12
74
3.16
15
1656.75
4758.44
125.3
9.53
28.7
25000
13
2
322
1.66
409.17
46.5
1.18
836.8
48.78
22.6
2009
14.21
415.5
34372.8
4.84
11.63
69
2.65
17
1728
5256.52
127.2
12.51
31.16
24381
13.3
1.48
320
1.59
565.3
46.05
1.15
1100.4
58.17
28.03
2010
15.2
458.3
36355.1
5.04
12.83
61
2.25
20
1749.3
5911.69
133.7
14.41
32.76
27965
17.3
0.98
330
1.5
689.41
45.67
1.09
1406.3
63.24
34.67
Table 8 Sustainability level of the subsystems and the overall system
Year
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Driving forces
0.08
0.08
0.09
0.09
0.09
0.11
0.13
0.17
0.20
0.22
Subsystems
Pressures
State
0.06
0.05
0.05
0.03
0.06
0.02
0.06
0.03
0.06
0.03
0.07
0.03
0.08
0.04
0.08
0.04
0.11
0.06
0.13
0.08
Impact
0.05
0.08
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.04
Response
0.12
0.12
0.13
0.13
0.15
0.19
0.25
0.32
0.36
0.43
System
Sustainable development of urban residential areas
0.36
Grade Ⅳ
0.36
Grade Ⅳ
0.33
Grade Ⅴ
0.34
Grade Ⅴ
0.36
Grade Ⅳ
0.43
Grade Ⅳ
0.53
Grade Ⅲ
0.64
Grade Ⅲ
0.76
Grade Ⅱ
0.90
Grade Ⅰ
Fig.7 shows the line chart of five subsystems while Fig.8 shows the line chart of the sustainable
development situation of residential urban areas in the city of Nanjing for the period 2001-2010.
Fig. 7 Line chart of the 5 subsystems
Fig. 8 Line chart of the overall system
5.2 Verification with historical data
The developed model is verified using the historical data of 2001, 2002, 2005, 2007 and 2008. The
examined variables include 4 indicators: “Urban GDP”, “The floor space completed of residential
housing”, “Living space per capita” and “Urban infrastructures”.
The method of verification is to make a comparison of the simulation results and historical data,
and calculate the relative error and mean square deviation of the variables.
ei = yˆit − yit / yit (4)
MSEi =
n
∑e
2
it
/ n (5)
1
Here, yit represents the historical data of the indicator “i” in the year of “t”, yˆit represents the
simulation result of the indicator “i” in the year of “t”. “n” is the number of the selected indicators in
this model verification. ei is the relative error of the indicator “i” and MSEi is the mean square error
of the indicator “i”.
Generally, we take the simulation results as an equitable and reasonable answer, when MSEi <5%.
According to function (4), (5) and Table 9, we could get MSEi of the 4 indicators is: 0.04, 0.048,
0.044 and 0.045< 0.05 (Table 10), so this model meets the accuracy requirements theoretically.
Table 9 The comparison table of historical data and simulation result
2001
2002
2005
2007
2008
The floor space completed of
residential housing
Urban GDP
year
Historical
data
Simulation
result
Relative
error
121.85
138.51
224.11
328.37
377.50
121.85
134.04
227.57
305.16
361.30
0.0%
3.2%
1.5%
7.1%
4.3%
Historical
data
3.09
3.47
5.65
7.47
8.91
Simulation
result
3.05
3.26
5.93
7.62
9.53
Living space per capita
Urban infrastructures
Relative
error
Historical
data
Simulation
result
Relative
error
Historical
data
Simulation
result
Relative
error
1.3%
6.1%
4.9%
2.0%
6.9%
19.00
20.10
24.3
26.08
30.84
18.91
19.28
22.89
26.10
28.70
0.5%
4.1%
5.8%
0.1%
6.9%
2.98
3.14
6.33
7.32
10.63
2.92
3.08
6.37
6.71
10.12
2.0%
1.9%
0.6%
8.3%
4.8%
Table 10 The mean square error of the 4 indicators
year
2001
2002
2005
2007
2008
MSEi
Urban GDP
2
1
e1
0.000
0.032
0.015
0.071
0.043
e
0
0.001024
0.000225
0.005041
0.001849
4.0%
The floor space completed
of residential housing
2
e2
e2
0.013
0.000169
0.061
0.003721
0.049
0.002401
0.020
0.000400
0.069
0.004761
4.8%
Living space per capita
2
3
e3
0.005
0.041
0.058
0.001
0.069
e
0.000025
0.001681
0.003364
0.000001
0.004761
4.4%
Urban infrastructures
2
e4
0.020
0.019
0.006
0.083
0.048
e4
0.000400
0.000361
0.000036
0.006889
0.002304
4.5%
6. Discussion
Taking into consideration the five subsystems (Fig.7), it is possible to observe that the State and
Impact categories develop smoothly in the considered period, while notable variations appear in
the line chart of the Driving Forces, Pressure and Response categories. The Driving Forces and
Response in particular rise steadily over the years; this is due to the increase in the values of
specific indicators, such as “GDP” (Driving Forces category), “Environmental pollution abatement
and control expenditure”, “Exploitation and investment of real estate” and “Exploitation and
investment of real estate” (Response category). With reference to the overall system under
examination (Fig.8), the model shows that the sustainability level of the Nanjing residential urban
area has an ascending trend, with a good performance (Grade II and Grade III) for the latter part of
the considered period (from 2007 to 2009). It is possible to notice that the bad performance of the
system in the first part of the period (Grade IV in 2001, 2002 and Grade V in 2003) is pushed up
towards more sustainable levels by the ascending trend of the Driving Forces and Response
categories. Especially, due to this trend of the sustainability level keeping on a rise in these years,
it will get into a bad performance (Grade I) again in the period of 2010.
7. Conclusion
In this paper, a system dynamics model that comprises an integrated housing-economyenvironment-society system is developed to assist evaluation of the development trend of the
urban residential areas in NanJing city, China. The model examines interactions among four
subsystems (“Driving forces”, “Pressures”, “State” and “Responses” category) within a time frame
of 10 years, and then the result of this model is discussed with the values of environment indicators
of “impact” category in DPSIR framework. It can be also concluded that the integration of System
Dynamics theory and DPSIR framework is a useful strategy to study the development of urban
residential areas in terms of sustainability. A validation is undertaken to verify the model in the end,
through comparison with historical data and mean square error. By this study, we can find the
development process of our residential areas, and put a further concern on the development of the
urban residential areas from various angles.
However, there are still a number of opportunities for expanding the study and for validating the
results obtained herein. Firstly, only core-indicators were considered in this work. It would be of
scientific interest to add other indicators resulting from policies and strategies. Secondly, further
research would be required to collect more historical data and optimize the structure of system
dynamics model. Finally, in this paper we only considered the historical values of environment
indicators of “impact” category in DPSIR framework and didn’t put them in system dynamics model.
In system dynamics model, creating equations to describe and quantize the relationship of
environment indicators with urban economy, social development and housing needs more work on
the monitoring and statistics of environment impacts.
Reference
[1]
BENJAMIN P.T., and LAWRENCE C.B., “Use of system dynamics as a decision-making tool
in building design and operation”, Building and Environment, 2010(45): 1006-1015.
[2] BERGER-SCHMITT, R., & NOLL, H, “Conceptual framework and structure of a European
system of social indicators”, EuReporting Working Paper 2000, No. 9. Mannheim: Centre for
survey research and methodology.
[3] BIDONE.E.D., and LACERDA L.D., “The use of DPSIR framework to evaluate sustainability
in coastal areas. Case study: Guanabara Bay basin,Rio de Janeiro, Brazil”, Regional
Environmental Change, 2004, Vol 4, No 1, pp 5-16.
[4] CHEN, Z.J and CHEN, G.J., “Study on warning system of environmental impact assessment”,
Environment Science, 1992, 13(4), 20-23.
[5] CUDC (2002). “China urban sustainable development indicators system”, Retrieved February
19, 1010, from http://www.china.com.cn/zhuanti2005/txt/2005-03/20/content_5815984.htm.
[6] ERMA S., CHOU S.Y., RUDI H., CHEN C.H., “Demand scenario analysis and planned
capacity expansion: A system dynamics framework”, Simulation Modelling Practice and
Theory, 2010(18): 732-751.
[7] HPS (High Performance Systems, Inc.), “lthink Technical Documentation”, High
Performance Systems, Inc. 1997.
[8] LOW C.D., “The regional open space paradox”, Queensland Planner, 2004, 44(3), 12-15.
[9] LOW C.D., “Achieving the vision: the role of environmental infrastructure in the achievement
of a sustainable and liveable region”, Queensland Planner, 2005, 45(2), 4-17.
[10] MICHALOS A., “Combining social, economic and environmental indicators to measure
sustainable human well-being”, Social Indicators Research, 41, 1-4.
[11] MOORE N., and SCOTT M., “Renewing urban communities: Environment, citizenship and
sustainability in Ireland”, Aldershor: Ashgate, 2005
[12] NESSA W., and MONTSERRAT P.E., “Sustainable Housing in the Urban Context:
International Sustainable Development Indicator Sets and Housing”, Social Indicators
Research, 2008,87, 211-221.
[13] MOHAPATRA, P.K.J., “Introduction to System Dynamics modeling”, Sangam Books, London,
1994, p.25-43.
[14] OECD Organization of Economic Co-operation and Development (2003). Environmental
indicators development, measurement and use. Paris: OECD Environment Directorate
Environmental Performance and Information Division. Retrieved February 19, 2010, from
http://www.oecd.org/dataoecd/7/47/24993546.pdf/
[15] ROBERT A., TAYLOR., “Origin of System Dynamics: Jay W. Forrester and the History of
System Dynamics”. US. Department of Energy’s Introduction to System Dynamics. Retrieved
23/10/2008
[16] SAATY, T.L., “The Analytic Hierarchy Process”, New York: McGraw Hill, 1998.
[17] SHAHBAZ K., LUO Y.F., and AFTAB A., “Analysing complex behavior of hydrological
systems through a system dynamics approach”, Environmental Modelling & Softeware,
2009(24): 1363-1372.
[18] SHEN L.Y., WU Y.Z., CHAN E.H.W., and HAO J.L., “ Application of system dynamics for
assessment of sustainable performance of construction projects”, Journal of Zhejiang
University Science.2005 6A(4):339-349
[19] SHEN Q.P., CHEN Q., TANG B.S., YEUNG S., HU Y.C., & CHEUNG G., 2009. “A system
dynamics model for the sustainable land use planning and development”, Habitat
International, 2009(33): 15-25.
[20] TUNG B., and CLAUDIA L., “Supporting cognitive feedback using system dynamics: A
demand model of the Global System of Mobile telecommunication”, Decision Support
System, 1996 (17): 83-98.
[21] United Nations (2007) Indicators of Sustainable Development: Guidelines and Methodologies.
New York: United Nations.
[22] WANG J.F., LU H.P., and PENG H., “System Dynamics Model of Urban Transportation
System and its Application”, Journal of transportation systems engineering and information
technology, Vol. 8, Issue. 3, 2008, pp. 83-89.
Appendix A
The core equations used in the software “vensim” to simulate the system.
• economic growth= economic growth rate(Time)*urban GDP
Units: billion
• exploitation and investment of real estate= proportion of real estate investment(Time)*urban
GDP Units: billion
• GDP per capita= urban GDP/urban population*1e+009
Units: yuan
• "housing price-to-income ratio"= living space per capita*urban housing price/urban
disposable income per capita
Units: Dmnl
• land price per floor area=0.75*base land price(Time)/average plot ratio*impact of the
proportion of real estate ivestment on land price(proportion of real estate investment(Time))
Units: yuan/m2
• living space per capita=1e-008*urban disposable income per capita*urban disposable
income per capita+0.0001*urban disposable income per capita+15.908 Units: m2
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
population growth=urban population*population growth rate(Time)/100+immigration
population(Time)-emmigration population(Time)
Units: people
sales area of dewelling houses=the total demand of residential houses*(0.7+0.1-0.3-STEP
(0.05,2001)-STEP(0.05,2002)-STEP(0.05,2003)+ STEP(0.2,2004)+ STEP(0.1,2005) +
STEP(0.05,2006))*"impact of housing price-to-income ratio"("housing price-to-income
ratio")*"impact of supply-to-demand ratio"("supply-to-demand ratio") Units: thousand m2
"supply-to-demand ratio"=the total supply of residential houses/the total demand of
residential houses
Units: Dmnl
the actual demand=sales area of dewelling houses
Units: thousand m2
the demand growth= "impact of housing price-to-income ratio"("housing price-to-income
ratio")*impact of loan rate on the demand growth(loan rate)*impact of dependency rate on
the demand growth(Dependency rate(Time))*impact of proportion of population living
below national poverty level(proportion of population living below national poverty
level(Time))*(population growth+dewellings in deficient state of repair(Time)*(0.5+0.15STEP(0.15,2004))*the urban family size+0.01*urban population*(0.6-0.15-STEP(0.15,2001)
+STEP(0.5,2002)-STEP(0.5,2003))+urban population*proportion of population married
*(0.85-0.02-STEP(0.03,2001)-STEP(0.2,2003)))*living space per capita/1000
Units: thousand m2
the floor space completed of residential housing=10*(0.03*exploitation and investment of
real estate*exploitation and investment of real estate+10.73*exploitation and investment of
real estate+191.7)*impact of urban infrastructure on the floor space completed(urban
infrastructure)
Units: thousand m2
the floor space of commercial house sold newly=0.233*exploitation and investment of real
estate*exploitation and investment of real estate+177*exploitation and investment of real
estate+1902.1
Units: thousand m2
the supply completed=sales area of dewelling houses
Units: thousand m2
the supply growth=the floor space of commercial house sold newly+the floor space
completed of residential housing*(0.15-STEP(0.1,2001) +STEP(0.05,2002) +STEP
(0.05,2003)-STEP(0.02,2004) +STEP(0.02,2005) +STEP(0.03,2006)+STEP(0.02,2007)
+STEP(0.02,2008)+STEP(0.02,2009))
Units: thousand m2
the total demand of residential houses= INTEG (the demand growth-the actual demand,
7120)
Units: thousand m2
the total supply of residential houses= INTEG (the supply growth-the supply completed,
9720)
Units: thousand m2
total cost=(construction cost(Time)+land price per floor area)*(1+0.7*2*loan rate/100)
Units: yuan/m2
urban disposable income per capita=GDP per capita*the distribution proportion of GDP
Units: yuan
urban GDP= INTEG ( economic growth,121.85)
Units: billion
urban housing price=total cost*(1+tax rate)*(1.25+STEP(0.02,2002) +STEP(0.03,2003)+
STEP(0.03,2004) +STEP(0.03,2008)+STEP(0.05,2009)+STEP(0.08,2010))*"impact of
supply-to-demand ratio"("supply-to-demand ratio")*impact of urban rent price on urban
housing price(the growth rate of urban rent price(Time))
Units: yuan/m2
urban infrastructure= urban GDP*proportion of urban infrastructure investment(Time)
Units: billion
urban population= INTEG (population growth, 3.71882e+006)
Units: people