Research Journal of Environmental and Earth Sciences 4(1): 48-54, 2012
ISSN: 2041-0492
© Maxwell Scientific Organization, 2012
Submitted: June 15, 2011
Accepted: July 20, 2011
Published: January 01, 2012
The Contribution Values of Energy Consumptions from the Different Industrial
Sectors on the Air Quality in Ningbo
Pingsha Huang, Lihua Teng and Pei Shi
Department of Environmental Science, College of Biological and Environmental Science,
Zhejiang Wanli University, Zhejiang 315100, P.R. China
Abstract: The aim of this study was to prived an overview of energy consumption in different industrial sectors
in Ningbo for furture adjusting the local industry construction and relateive layout in terms of energy saving
and air quality improvment. With the Bayesian network theory, combined the energy consumptions of unit
production and indicators of air quality, energy consumption on the contribution to air quality from the various
industrial sectors were analyzed in Ningbo. The results showed that the energy consumptions from five
industrial sectors and GDP/capital directly affected on comprehensive air pollution index, while there was a
direct causal relationship between GDP/capital and energy consumptions in five industrial sectors. The results
implied that, for comprehensive air pollution index, the contribution from the petroleum, coking and nuclear
fuel processes was larger. The sectors of the petroleum, coking and nuclear fuel processes and NOX from
vehicle made more contributions on the NO2. Meanwhile, the petroleum, coking and nuclear fuel processes and
chemical raw materials and chemical products sectors gave more contributions on the SO2. For the TSP and acid
rain rate, the sector of the petroleum, coking and nuclear fuel processes was key contributor. It could be said
the sector of the petroleum, coking and nuclear fuel processes was the most important industrial sector which
impacts on local air quality in Ningbo.
Key words: Air quality, Bayesian network, comprehensive air pollution index, contribution value, energy
consumption, GDP/capital
INTRODUCTION
edible oil, and aquatic products in China, there are over
100,000 private enterprises in Ningbo. The main
traditional industries of Ningbo are light industry and
textiles, garments, machinery, daily used household
electrical appliances, plastic injection machines, dye
making and motor parts, and coal generated power plant
and industrial metals production. Moreover, building
materials, pharmaceutical industries, and large-scale
harbor-based industries, represented by petrochemicals,
iron and steel, power generation and papermaking, etc.
have quickly developed continiously. The private
economic sector appeared earlyer in Ningbo than in other
cities in China, so the market-oriented economy has been
booming in Ningbo for last three decades.
A brief introduction of Ningbo: As the second largest
city in Zhejiang Province, located in coast of the East
China Sea, the south of Yangtze River Delta and NingShao Plain, Ningbo has a long coastline, 1562 km, with
deep-water seaport and 531 scattered islands (Ningbo
Municipal Bureau of Statistics, 2010-2009). The total
land area of Ningbo is 9,816 km2, with the urban area of
2,462 km2. The population of Ningbo is 5.7102 million
with 2.218 million living in the urban area. Due to its
specific geographic location and natural environment,
Ningbo has become a key investment city for long
time in history.
Being one of the earliest birthplaces of Chinese
modern industry, Ningbo has been listed as a key
investment area of the province and of the country since
reforming and opening up to the outside world. In recent
years, the economic development of Ningbo is very
rapidly. All of the industrial added value achieved 200.7
×109 Yuan (RMB), an increase of 4.1% has been archived
compared with the data in 2009 (Ningbo Municipal
Bureau of Statistics, 2010).
The industry in Ningbo has developed rapidly since
the 1990s. As a major production base of grain, cotton,
Energy consumption in Ningbo: The energy in Ningbo
is short. There is "no coal and no oil" on the land. More
than 99% of primary energy is inported from the outside
market. As a majoy ecnomic area, Ningbo is an energy
consumption city in China and an important base of the
secondary energy production in East China as well. In
recent years, as one of the main bottlenecks, the overall
supply of energy has become shorter. It has restricted the
social and economic sustainable development of Ningbo.
Based on the energy Annual Report 2008 of Ningbo the
Corresponding Author: Pingsha Huang, Department of Environmental Science, College of Biological and Environmental
Science, Zhejiang Wanli University, Zhejiang 315100, P.R. China
48
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
research (Lee and Abbott, 2003). However, there were
only a few reported applications of Bayesian network in
the field of environmental protection (Varis, 1995; Varis
and Kuikka, 1997; Borsuk et al., 2004; Henriksen and
Barlebo, 2008; Henriksen et al., 2007; Farmani et al.,
2009; Martin de Santa Olalla et al., 2005, 2007).
In this study, taking air pollution indicators, the
energy consumption of the five industrial sectors,
GDP/capital and comprehensive air pollution index as
variables. A Bayesian network model was established in
light of the monitoring data of air pollutants from 1998 to
2008.
overall efficiency of energy utilization in Ningbo was
close to 40%, higher than the national average level and
the province average level of Zhejiang Province. The high
energy-consuming industries in Ningbo, such as in oil
refining, power generation, caustic soda, stainless steel
and paper production, has decreased their energy
consumption to lower level. In 2008, the energy
consumption for producing 104 Yuan (RMB) GDP was
0.77 tons of coal equivalent (hereinafter referred to as tce)
droped down 0.5% compared to that in 2007, and a total
of 1485×103 tce has been saved in 2008. At the same time
the energy consumption for per unit product in these
industrial sectors has become lowwer and lower as well.
Due to industrial structure and relative constraints the
energy consumption is still facing great challenges in
Ningbo. It has been pointed out that compared with 2007
the industrial energy consumption in 2008 increased
9.7%. It reached 73.1% of total social energy
consumption. In the most energy consumption industrial
sectors, electricity and heat industry, petroleum
processing industry, ferrous metal smelting and rolling
processing industry, textiles and paper and paper products
industry played very important roles.
The absolute values of energy consumptions in
various industries sectors could be figured out, but for
different processes and various types of energy sources,
the amount of greenhouse gases emission from the
different industrial sectors were different. So, it is
necessary to assess the impact factors of main energy
consumption industrial sector by means of statistical
methods, to know how these factors impact the air quality.
It is to understand the industrial pollution and therefore,
to adjust of the industrial structure.
MATERIALS AND METHODS
The data collection: The data used in this study included
energy consumption in industrial sectors, GDP/capital, air
pollution indicators and comprehensive air pollution
index. All of these data were from the annual reports of
environmental quality based on routine mornitoring and
the statistics bulletins by Ningbo Government from 1999
to 2008. The five main industrial sectors in Ningbo were
the Production and Supply of Electric Power and Heat
Power (PSEPHP), the Processing of Petroleum, Coking
and Procssing of Nuclear Fuel (PPCPNF), the
manufacture of textile (MT), the Manufacture of Paper
and Paper Products (MPPP) and the Manufacture of Raw
Chemical Materials and Chemical Products (MRCMCP).
Air pollution indicators included sulphur dioxide (SO2),
Nitrogen Dioxide (NO2), Total Suspended Particles
(TSP), dust, rate of acid rain and nitrogen oxides from
vehicle (NOX). Comprehensive Air Pollution Index
(CAPI) means the sum of individual air Pollution
Indicator divided by its relative national air quality
standard of the secondary class.
Bayesian network for the scientific studies: A Bayesian
network is a probabilistic graphical model that represents
the causality and correlation of a set of data, capable of
inferring and representing what is under uncertainty. It is
popular in the filed of artificial intelligence in recent
years. Since 1988 when a clear definition was given by
Pearl (1988), the Bayesian network has attracted much
attention. The network is graphically represented as a
Directed Acyclic Graph (DAG) consisting of nodes and
arcs where the nodes stand for random variables and the
arcs show direct links between variables. A detailed
description of the Bayesian network inference theory was
well discussed by Jensen (2001), Neapolitan (2003),
Spiegelhalter et al. (1993), Korb and Nicholson (2004).
The Bayesian network has been applied widely in the
domain of epidemiology, such as disease diagnosis
(Aronsky and Haug, 2000; Burnside et al., 2000; Liao
et al., 2010), risk analysis (Maglogiannis et al., 2006;
Maskery et al., 2008; Dlamini, 2010), management
decisions (Said, 2006), analysis for stochastic dependence
(Agostinelli and Rotondi, 2003), classification of
cytological findings (Hamilton et al., 1995), nursing
Calculation methods: According to the references of
Varis (1995), Henriksen and Barlebo (2008) and Borsuk
et al. (2004), with the BNT toolkit matlab self parameter
calculation procedure, the data were used to calculated.
RESULTS AND DISCUSSION
The GDP, energy consumptions and air qualities in
Ningbo from 1999 to 2008: With the rapid economic
development and the technology improvement, the
GDP/capital increased rapidly year by year, from 16,929
Yuan (RMB) in 1998 to 69,997 Yuan (RMB) in 2008 in
Ningbo. And meanwhile, the energy consumption has
showed a continuous decline obviously except for the
consumption in the production and supply of electric
power and heat power sector. The energy consumption of
unit prodution in the production and supply of electric
power and heat power was double (Table 1).
In Table 1, it showed that the air pollution indicators,
such as TSP, SO2 and NO2, were lower than the national
49
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
Table 1: The energy consumptions of the unit production in five industrial sectors, GDP/capital, CAPI and air pollution indicators
Energy consumption of unit production
GDP/
NO2
TSP
Dust
Rate of
In five main industrial sectors (tce/104 Yuan)
(mg
(mg/
(t/km2 acid
------------------------------------------------------------------ Capital
SO2
3
3
3
m)
.M)
Rain (%)
Year PSEPHP PPCPNF MT
MPPP
MRCMCP (Yuan)
CAPI (mg/m ) /m )
National air quality standard of the secondary class
0.06
0.04
0.20
1998 1.0800
1.8200
0.4700 0.8700
0.7183
16929
2.54
0.017
0.029
0.167
8.105
74.5
1999 0.9500
1.5700
0.4969 0.8900
0.74511
18342
2.31
0.021
0.031
0.141
7.627
70.5
2000 0.8483
0.8421
0.4838 0.8383
0.6219
19935
2.14
0.015
0.034
0.137
8.364
81.2
2001 0.8733
0.9245
0.5393 0.7883
0.6134
22078
1.82
0.023
0.031
0.110
7.978
84.9
2002 1.2671
0.9759
0.4724 0.7764
0.5724
24121
1.93
0.017
0.039
0.108
7.925
94.8
2003 1.2737
0.8411
0.4591 0.8687
0.5591
27570
1.31
0.026
0.044
0.129
7.976
91.5
2004 0.6500
0.7700
0.3300 0.7500
0.4300
33200
1.25
0.027
0.044
0.109
6.991
90.7
2005 2.2413
0.6879
0.2104 0.9892
0.3104
38733
1.28
0.034
0.045
0.083
7.029
96.7
2006 2.3705
0.4365
0.2087 0.7982
0.3087
44842
1.54
0.041
0.043
0.078
5.926
95.5
2007 2.5391
0.4078
0.1919 0.6728
0.1506
51285
1.83
0.042
0.043
0.078
5.688
100
2008 2.2120
0.3220
0.1767 0.6429
0.1767
69997
1.85
0.043
0.044
0.084
5.558
97.9
Table 2. Normalized results of energy consumptions of the five industries, GDP/capital, CAPI and air pollution indicators
1998 - 0.5958
2.1417
0.7471 0.6588
1.2245 - 1.0403
1.8168 - 10.7129 - 1.6375
1.9752
0.9177
1999 - 0.7884
1.5766
0.9428 0.8703
1.3586 - 0.9509
1.2521 - 6.7525 - 1.3040
1.0532
0.4346
2000 - 0.9390 - 0.0687
0.8474 0.3236
0.7425 - 0.8501
0.8348 - 12.6931 - 0.8037
0.9113
1.1795
2001 - 0.9019
0.1175
1.2508 - 0.2042
0.7001 - 0.7144
0.0491 - 4.7723 - 1.3040 - 0.0461
0.7894
2002 - 0.3187
0.2337
0.7645 - 0.3310
0.4953 - 0.5851
0.3192 - 10.7129 0.0300 - 0.1170
0.7358
2003 - 0.3089 - 0.0710
0.6679 0.6450
0.4288 - 0.3669 - 1.2030 - 1.8020
0.8638
0.6277
0.7873
2004 - 1.2327 - 0.2317 - 0.2703 - 0.6101 - 0.2164 - 0.0105 - 1.3504 - 0.8119
0.8638 - 0.0816 - 0.2082
2005
1.1241 - 0.4173 - 1.1395 1.9193 - 0.8141
0.3396 - 1.2767 6.1188
1.0305 - 1.0035 - 0.1698
2006
1.3155 - 0.9855 - 1.1519 - 0.1005 - 0.8226
0.7263 - 0.6384 13.0495 0.6970 - 1.1809 - 1.2846
2007
1.5652 - 1.0504 - 1.2738 - 1.4261 - 1.6129
1.1340
0.0737 14.0396 0.6970 - 1.1809 - 1.5252
2008
1.0807 - 1.2443 - 1.3844 - 1.7427 - 1.4823
2.3183
0.1228 15.0297 0.8638 - 0.9681 - 1.6566
Nox
from
vehicle (t)
753.4
3644.6
6538.6
9435.5
12441.5
15262.6
18465.4
20258.7
23596.7
27951.4
29432.0
- 1.5327
- 1.9577
- 0.8209
- 0.4278
0.6239
0.2733
0.1883
0.8257
0.6983
1.1763
0.9532
- 1.5769
- 1.2625
- 0.9477
- 0.6327
- 0.3057
0.0011
0.3494
0.5444
0.9075
1.3811
1.5421
Table 3: The results of discretization of the energy consumptions of the unit production in five industrial sectors, GDP/capital, CAPI and air pollution
indicators
Energy consumption of unit production
in five industrial sectors
Rate of NOx
---------------------------------------------------------------- GDP/
acid
from
NO2
TSP
Dust
rain
vehicle
Year
PSEPHP PPCPNF MT
MPPP
MRCMCP Capital CAPI
SO2
1998
low
high
mid
mid
mid
low
high
high
high
high
mid
high
high
1999
low
high
mid
mid
mid
low
mid
high
low
mid
mid
high
low
2000
low
mid
mid
mid
mid
low
mid
high
low
mid
mid
low
low
2001
low
mid
mid
mid
mid
low
mid
high
low
mid
mid
mid
low
2002
mid
mid
mid
mid
mid
low
mid
high
mid
mid
mid
mid
mid
2003
mid
mid
mid
mid
mid
mid
low
high
mid
mid
mid
mid
mid
2004
low
mid
mid
low
mid
mid
low
low
mid
mid
mid
mid
mid
2005
mid
mid
low
high
low
mid
low
high
mid
low
mid
mid
mid
2006
mid
low
low
mid
low
mid
low
high
mid
low
low
mid
mid
2007
high
low
low
low
high
mid
mid
high
mid
low
high
mid
mid
2008
mid
low
low
high
low
high
mid
high
mid
low
high
mid
high
standard of the secondary class, and TSP and dust
decreased, but SO2 and NO2 had the trend of deterioration
year by year. In recent years the amount of NOx increased
with the dramatically increase the number of car,
especially the private car. So the vehicle emissions had
become an important factor resulted in poorer air quality.
The rate of acid rain increased to almost 100% currently.
The models based in Bayesian network was built by
the EM algorithm. The results were shown in Fig. 1 to 6,
respectively.
Figure 1 showed that the five industrial sectors and
GDP/capital have a direct impact on the comprehensive
air pollution index; that there was a direct correlation
between the five industrial sectors and GDP/capital; and
that there was a direct cause-and-effect relationship
between the electricity and heat supply industry and the
other four sectors. Judging from the diagnostic function of
the Bayesian network, the probability of high energy
consumption increased in the fields of the petroleum,
coking, and nuclear fuel processes when the
comprehensive air pollution index was high, while the
probability of low energy consumption appeared in the
fields of the paper and textile industries when the
comprehensive air pollution index was higher. Therefore,
The contribution to the value of air quality by the
different industrial sectors in Ningbo: The data in
Table 1 was normalized by standard methods. The results
were shown in Table 2. Bayesian networks allow the
prediction of a discrete outcome from a set of variables
that may be continuous, discrete, and dichotomous or a
mix of any of these. The algorithm of mean discretization
was used to discrete data in Table 2. The results were
shown in Table 3.
50
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
MRCMCP
High 25.9
Low 62.9
Mid 11.2
20.9
55.7
Mid 23.4
MPPP
High 24.8
Low 74.9
Mid 0.33
GDP/Capital
High 11.8
Low 64.9
Mid 23.3
MT
Low 69.6
Mid 30.4
High
Low
Mid
CAPI
28.3
55.3
16.5
PSEPHP
High 15.0
Low 63.0
Mid 22.0
Fig. 1: The Bayesian network of CAPI, five industrial sectors and GDP/capital in Ningbo
High
Low
Mid
MRCMCP
38.1
34.6
27.2
MPPP
High 7.66
Low 92.4
High 49.1
Low 31.2
Mid 19.7
MT
High 4.96
Low 46.8
Mid 48.2
GDP/capital
High 8.71
Low 58.7
Mid 32.6
NO
High 66.4
Low 34.6
High
Low
Mid
PSEPHP
High 8.24
Low 84.1
Mid 7.62
NO2
100
0
0
Fig. 2: The Bayesian network of NO2, five industrial sectors, GDP/capital and NOx from vehicle in Ningbo
High
Low
Mid
High
Low
MRCMCP
8.76
31.3
59.9
High
Low
Mid
PPCPNF
21.1
32.1
46.8
MPPP
56.6
43.4
High
Low
Mid
High
Low
Mid
PSEPHP
10.8
42.6
46.6
MT
0.17
38.4
61.5
GDP/Capital
High 12.9
Low
50.8
Mid
36.3
SO
High 100
Low
0
Fig. 3: The Bayesian network of SO2, five industries and GDP/capital in Ningbo
it can be concluded that industrial sectors like the
petroleum, coking, and nuclear fuel processes have a
stronger impact on the comprehensive pollution index
while industrial sectors such as paper and textile
manufacturing have a weaker impact.
Figure 2 showed that the nodes having a direct
connection with the indicator NO2 were NOx emissions
from the automobile exhaust and industries from the other
4 industrial sectors except for the chemical industry. It
could be seen from the probability table that, among the
51
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
High
Low
Mid
High
Low
Mid
MRCMCP
16.4
23.9
59.7
High
Low
Mid
PSEPHP
9.04
High
45.5
Low
45.5
Mid
MPPP
35.5
34.0
30.0
GDP/Capital
High
25.0
Low
50.0
Mid
25.0
PPCPNF
25.5
24.7
49.7
MT
Low 32.2
Mid 67.8
High
Low
Mid
TSP
100
0
0
Fig. 4: The Bayesian network of TSP, five industrial sectors and GDP/capital in Ningbo
MRCMCP
High 0.28
Low
4.41
Mid
95.3
High
Low
Mid
Low
Mid
High
Low
Mid
Mt
4.70
95.3
MPPP
0.26
9.35
90.4
PPCPNF
94.3
0.46
5.26
GDP/Capital
High 4.63
Low 90.4
Mid 4.97
PSEPHP
High 3.97
Low 93.6
Mid 2.42
Rate of Acid Rain
High 100
Low
0
0
Mid
Fig. 5: The Bayesian network of rate of acid rain, five industrial sectors and GDP/capital in Ningbo
High
Low
Mid
High
Low
Mid
MRCMCP
46.8
50.2
2.97
MT
Low 97.0
Mid 2.97
High
Low
Mid
PPCPNF
2.97
93.6
3.47
MPPP
31.2
49.8
19.1
GDP/Capital
High 33.3
Low
2.15
Mid
64.5
High
Low
Mid
PSEPHP
50.0
1.73
48.3
Dust
High 100
Low
0
Mid
0
Fig. 6: The Bayesian network of dust, five industrial sectors and GDP/capital in Ningbo
direct factors, the probability value of NOx emissions
from the automobile exhaust, petroleum, coking, and
nuclear fuel processes were higher when the NO2 in the
air was higher, while the probability value was lower in
the fields of paper industry, electric power and heat power
sectors. This indicated that NOx emissions from the
automobile exhaust and the petroleum, coking, and
nuclear fuel processes contributed more to monitored
52
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
NO2, while the relative lower value were came from the
contribution of paper industry, electric power and heat
power supply.
Figure 3 displayed direct impacts on the indicator
SO2 were from the production and supply of electric
power and heat power, the processing of petroleum,
coking and prcessing of nuclear fuel, the manufacture of
raw chemical materials and chemical products, and
GDP/capital. When SO2 was higher, the contribution of
the processes of petroleum, coking and nuclear fuel and
the manufacture of raw chemical materials and chemical
products were larger to SO2 than that of GDP/capital and
the production and supply of electric power and heat
power to SO2. Figure 4 illustrated the nodes having a
connection with TSP included those from the manufacture
of textile, the processes of petroleum, coking and nuclear
fuel, the manufacture of raw chemical materials and
chemical products and the production and supply of
electric power and heat power. The paper industry had an
indirect impact on the TSP node. In addition, the
GDP/capital had a direct correlation with the five
industrial sectors. It can be seen with the probability
distribution table that, among the direct impacts on TSP,
the probability of the energy consumption of the power
and heat supply sector was larger when monitored TSP
was higher, while the probability value in the fields of the
chemical industry, the petroleum, coking and nuclear fuel
processes were smaller when the monitored TSP was
higher. It indicated that industrial sectors, including the
petroleum, coking, nuclear fuel processes and the
chemical processing had more contributions to TSP than
that of the textile industry and the power and heat supply
sector to TSP.It could be seen from Fig. 5 that the energy
consumptions in industrial sectors, including the
processes of petroleum, coking and processing of nuclear
fuel, the manufacture of raw chemical materials and
chemical products and the production and supply of
electric power and heat power, had a direct correlation
with the rate of acid rain. When the rate of acid rain was
higher, the probability of a higher value increased in the
fields of the processes of petroleum, coking and
processing of nuclear fuel. This implied that the
petroleum, coking, and nuclear fuel processes had a
stronger impact on the rate of acid rain, the chemical
processing had a stronger impact on the rate of acid rain,
and production and the supply of electric power and heat
power sector had a comparative impact on the rate of acid
rain. Figure 6 showed that the node of the dust directly
related to that of the supply of electric power and heat
power, the petroleum, coking and nuclear fuel processes
and the manufacture of raw chemical materials and
chemical products. It displayed that the GDP/capital had
a direct cause-and-effect relationship with the five
industrial sectors. The textile industry had an impact on
the dust through the supply of electric power and heat
power sector. According to the conditional probability
distribution, when monitored dust was higher, there was
a larger probability in the fields of the supply of electric
power and heat power, the manufacture of raw chemical
materials and chemical products. But compared with the
supply of electric power and heat power, the probability
value was lower in the fields of the manufacture of raw
chemical materials and chemical products. It indicated
that the supply of electric power and heat power had a
stronger impact on the dust than that of the manufacture
of raw chemical materials and chemical products on the
dust, and the petroleum, coking and nuclear fuel processes
had a greater probability of lower value, indicating the
impact of their contributions to dust is smaller.
CONCLUSION
With multivariate statistical analysis and Bayesian
network model, the cross impacts and contribution values
of energy consumptions from the main industrial sectors
and GDP/capital in Ningbo were analyzed. It showed that:
energy consumptions from five main industrial sectors
and GDP/capital can affect directly on the comprehensive
air pollution index, while there was a direct cause-andeffect relationship between the energy consumption of
five main industrial sectors and GDP/capital. Among the
industrial sectors, the petroleum, coking and nuclear fuel
processes sector made more contribution on the
comprehensive pollution index, and the paper and paper
products and the textile sectors gave a little impact on the
CAPI. For the NO2, more contributions were from
automobile exhaust and the petroleum, coking and nuclear
fuel processes. Although many factors impacted on the
SO2, the sector of the petroleum, coking and nuclear fuel
processes made more contribution. More TSP could come
from the sectors of the petroleum, coking and nuclear fuel
processes and the manufacture of raw chemical materials
and chemical products. For the acid rain, the petroleum,
coking and nuclear fuel processes was the most important.
If considered energy consumption in various
industrial sectors and their impact on air quality, it could
be said that the petroleum, coking and nuclear fuel
processes sector was the most important industrial sector
effected the air quality in Ningbo.
ACKNOWLEDGMENT
This study was supported by the funding to the
Research Project on Air Quality Impacted by Energy
Consumption in Ningbo from the Environment Protection
Science Research and Design Institute of Ningbo. The
authors would like to give grateful to the Mr. Zhu
Yonggang for his help in the data processing.
REFERENCES
Agostinelli, C. and R. Rotondi, 2003. Using Bayesian
belief networks to analyse the stochastic
dependencebetween interevent time and size of
earthquakes. J. Seismol., 7: 281-299.
53
Res. J. Environ Earth Sci., 4(1): 48-54, 2012
Aronsky, D. and P.J. Haug, 2000. Automatic
identification of patients eligible for a pneumonia
guideline. Proc/AMIA Annu. Symp., 12-16.
Borsuk, M., C. Stow and K. Reckhow, 2004. A Bayesian
network of eutrophication models for synthesis,
predict ion, and uncertainty analysis. Ecol. Model.,
173: 219-239.
Burnside, E., D. Rubin and R. Shachter, 2000. ABayesian
network for mammography. Proc/AMIA Annu.
Symp., Vol: 106-110.
Dlamini, W.M., 2010. Application of Bayesian networks
for fire risk mapping using GIS and remote sensing
data. Geo J., DOI: 10.1007/s10708-010-9362-x.
Farmani, R., H.J. Henriksen and D. Savic, 2009. An
evolutionary Bayesian belief network methodology
for optimum management of groundwater
contamination. Environ. Model. Software, 24:
303-310.
Hamilton, P.W., R. Montironi, W. Abmayr, M. Bibbo,
N.Anderson, D. Thompson and P.H. Bartels, 1995.
Clinical applications of Bayesian belief networks in
pathology. Pathologica, 87: 237-245.
Henriksen, H.J. and H.C. Barlebo, 2008. Reflections on
the use of Bayesian belief networks for adaptive
management.J.Environ. Management, 88: 1025-1036.
Henriksen, H.J., P. Rasmussen, G. Brandt and D.V. BowJensen, 2007. Public participation modelling using
Bayesian networks in management of groundwater
contamination. Environ.
Model. Software,
22: 1101-1113.
Jensen, F.V., 2001. Bayesian Networks and Decision
Graphs. Springer, New York.
Korb, K. and A.E. Nicholson, 2004. Bayesian Artificial
Intelligence. Chapman and Hall, London.
Lee, S.M. and P.A. Abbott, 2003. Bayesian networks for
knowledge discovery in large databasets: Basics for
nurse researchers. J. Biomed. Inform., 36: 389-399.
Liao, Y.L., J.F. Wang, Y.Q. Guo and X.Y. Zheng, 2010.
Risk assessment of human neural tube defects using
a Bayesian belief network. Stoch. Environ. Res. Risk
Assess, 24: 93-100.
Maglogiannis, I., E. Zafiropoulos, A. Platis and
C. Lambrinoudakis, 2006. Risk analysis of a patient
monitoring system using Bayesian network
modeling. J. Biomed. Inform. 39: 637-647.
Martin de Santa Olalla, F.J., A. Dominguez, A. Artigao,
C. Fabeiro and J.F. Ortega, 2005. Integrated water
resources management of the Hydrogeological Unit
Eastern Mancha using Bayesian Belief Networks.
Agric. Water Manage., 77: 21-36.
Martin de Santa Olalla, F.J., A. Dominguez and
J.F.Ortega, 2007. Bayesian networks in planning a
large aquifer in Eastern Mancha, Spain. Environ.
Model. Software, 22: 1089-1100.
Maskery, S.M., H. Hu, J. Hooke, C.D. Shriver and
M.N.Liebman, 2008. A Bayesian derived network of
breast pathology co-occurrence. J. Biomed. Inform.,
41: 242-250.
Neapolitan, R., 2003. Learning Bayesian networks. Upper
Saddle River, Prentice Hall.
Ningbo Municipal Bureau of Statistics, 2010. Statistics
Report in 2009. Beijing: Statistics press.
Pearl, J., 1988. Probabilistic Reasoning In Intelligent
Systems: Networks of Plausible Inference.
Kaufmann, San Mateo.
Said, A., 2006. The implementation of a bayesian
network for watershed management decisions. Water
Res. Manage., 20: 591-605.
Spiegelhalter, D.J., A.D. Dawid., S.L. Lauritzen and
R.G.Cowell, 1993. Bayesian analysis in expert
systems. Statistical Sci., 8: 219-283.
Varis, O., 1995. Belief networks for modelling and
assessment of environmental change. Environ., 6:
439-444.
Varis, O. and S. Kuikka, 1997. Bayesian approach t o
expert judgment elicitation with case studies on
climatic change impact assessment on surf ace
waters. Climatic Change, 37: 539-563.
54