Carbon Footprint of Vietnam’s Small Urban Areas (A Case Study of Ha Dong District, Hanoi)45268

Carbon Footprint of Vietnam’s Small Urban Areas A Case Study of Ha Dong District, Hanoi Nguyen An Thinh 1* 1 VNU University of Economics and Business, Vietnam National University, Hanoi

Carbon Footprint of Vietnam’s Small Urban Areas (A Case Study of Ha Dong District, Hanoi)

Nguyen An Thinh (1)*

(1) VNU University of Economics and Business, Vietnam National University, Hanoi, Vietnam

Correspondence: anthinhhus@gmail.com

Abstract: Increasing urbanization advocates the compact city This study classifies small urban areas

in the Ha Dong district (Hanoi, Vietnam) based on indicators for a compact city, and calculates the carbon footprint of urban clusters based on data household of lifestyle surveys A step use approach combining exploratory factor analysis, hierarchy cluster analysis, and carbon footprint is used The results sort out four main factors characterizing the compact city as “quality of infrastructure”,

“density of open space”, “transportation pattern”, and “public transportation and urban green space” Small urban areas in Ha Dong are grouped into 4 urban clusters based on the factor “quality

is higher than this of the average world, about 3 times over the GHGs target, and nearly 6 times than that of GHGs in Vietnam The urban cluster C3 shows the highest carbon footprint, whereas the C4 has the lowest one Recommendations based on the study result include: have been raised for study area as increase the number of high-density cities (compact city); develop the urban green spaces and public transport system; and improve tools for urban planning based on the criteria of sustainable development and green growth

Keywords: Urban small area; urban clusters; compact city; carbon footprint; factor analysis; Hanoi city

1 Introduction

Sustainable cities offer a considerable challenge for contemporary land use planning During recent years, Vietnam urbanized at a rate of approximately 3.4 percent per year (WB, 2012); by 2013 about 33.47 percent of the land was urban (MOC Vietnam, 2013) Urban planning until now contributed to environmental pollution, which hampered the economic growth and the sustainable development of the country The urban population is predicted

to double in the next ten years, which necessitates agriculture land transformed to urban land by four times (WB, 2012) Urban areas developed without planning could lead to the decrease of infrastructure qualities, impacting on environment and urban landscape During the United Nations Conference on Sustainable Development (Rio 20+) in 2012, decision makers committed to promote an integrated approach to planning and building sustainable cities and urban settlements It would not create liveable habitats for urban people but deal with other ecological and environmental problems

Climate change, population growth and urbanization are considered significant issues that need to be considering in the process of land use planning and urban planning ensuring sustainable development Vietnam has witnessed urban sprawling faster and faster during different periods of socioeconomic development (WB, 2012) Consequences of unplanned urban are inefficient land use as well as environmental pollution Sustainable

urban planning requires both improving the land quality and decreasing the emission of green house gases (GHGs) In European country, compact cities have been showed as good examples of sustainable urban planning and climate change responses in terms of creating liveable habitats for urban people, and promoting relatively high residential density with mixed land uses and based on an efficient public transport system (Dantzig and Saaty, 1973)

In the international literature, there were many urban areas developed from 1950 to

1970, resulting from the increased housing demand after wars As a result, different problems arising from the process of urbanization: the irrational land use, the overuse and exploitation of natural resources, environment pollution, traffic jam, and green house gas emission In 1973, the term “compact city” was invented and defined as a way to efficient urban planning, which allowed to limit the process of urban sprawl and to reduce energy consumption (Dantzig and Saaty, 1973) Brundtland commitment in 1987 first introduced the term “sustainable development” and stated that “transforming the old urban architecture into the compact city was considered as a good practice to archive sustainable

development Compact city was the structure of urban planning which enabled citizens to

cycle and walk easily as well as use public transportation effectively (Elkin, 1991) Compact city was regarded as the more sustained urban architecture in compared with urban sprawl because its structure helped to reduce using private cars and to increase the effect of infrastructure (Williams, 2000)

While urban sprawl became ineffective and lead to inequality society (Bourne, 1992), compact cities offered good conditions for society development (Garcia and Riera, 2003) Different research on the relationship between the compact city and sustainable development showed that the compactness and the sustainability had interactive effects (Neuman, 2005) Transportation was one of major resources of carbon dioxide emission One solution that could be implemented was to adjust land use planning (Baron, 1990) Through efficiency urban planning and effective land use, the amount of fossil fuel decreased by

10-15 percent, which means the amount of CO2 released also dropped by 10-15 percent through the changes in using transportation (Rickaby, 1992) Compact cities were developed by mixing land use in a reasonable scale for the purpose of effective energy consumption (Owen, 1992) because the effect is higher in high density cities (WB, 2010) Obviously, the urban structure and energy consumption have close linkage

This study deals with classifying small urban areas and calculating carbon footprint

in a case study of Ha Dong district (Hanoi city) Ha Dong district is selected as a case study area It is located along two sides of the National Road no 6, is 13 kilometer away from city center, and covers 4,833.66 square kilometers Ha Dong is an urban district which contiguous to 3 rural districts (Thanh Oai district in the south, to Thanh Tri district in the east, and Hoai Duc district in the west) Thus, it becomes the west gate to the suburb of Hanoi In the history of development, Ha Dong became a district of Hanoi because of the urban sprawl In addition, the speed of urbanization accelerates, which results in more and more effect showing clearly in the urban landscape of Ha Dong Meanwhile, Ha Dong affected by population growth and urban resettlement In general, Ha Dong’s infrastructure

is not synchronous The center areas have good infrastructure, in contrast to the suburb

areas Therefore, practical solutions need to be considered to solve problems related to urban planning

The rest of paper consists of four parts: the part 1 shows an introduction and literature review; the part 2 introduces material and methodology of factor analysis and carbon footprint calculation; the part 3 focuses on the results of small urban area classification and calculating carbon footprint for these areas; lastly, conclusion and discussion are pointed out in the part 4

2 Methodology

2.1 Conceptual study model

Figure 1 indicates flow-chart of steps for studying carbon footprint of urban small areas Once land use map in the year 2010 had been used to conduct an urban small area map, the factor analysis and hierarchy cluster analysis was applied to group small urban areas into a reduced number of urban clusters The analysis is based on data collected by an urban architecture questionnaire Then responses of citizens’ lifestyle questionnaire were used to create input data for carbon footprint estimation for different small urban areas Lastly, an urban planning recommendation was conducted based on analyzed results

Figure 1 Flow-chart of steps for studying carbon footprint of urban small areas

2.2 Questionnaire

Two questionnaires were conducted for survey as follow:

(i) Urban architecture questionnaire: as shown in the Table 1 and Table 2, this questionnaire was designed based on 30 indicators belonging to 7 criteria of compact city The 3-point-Likert scale is used to quantify and standardized the properties of small urban areas Data conducted from this questionnaire then to be used in factor analysis and hierarchy cluster analysis

(ii) Carbon footprint questionnaire: enables to collects data from 3 sectors for carbon footprint estimation (as shown in Table 3)

Table 1: Criteria of a compact city

Mixed land use (C1) Density of open spaces; the number of

commercial zones/ services places Take advantage of compact building design (C2) Housing architecture

Provide a variety of transportation choices (C3) The quality of transportation Preserve open space, farmland, natural beauty, and critical

environmental areas (C4)

The quality of public places (parks, schools, hospitals, playgrounds) Strengthen and direct development toward existing

communities (C5)

The environmental quality

Make development decisions predictable, fair, and cost

effective and create a range of housing opportunities and

choices (C6)

The effect of infrastructure planning

Develop ground space (C7) The number of ground spaces

Table 2: The interpretation of criteria and indicators for urban architecture questionnaire

C1

I1 Open space density

low (1), average (2), high (3) I2 Water body density

I3 Number of commercial sites (markets,

shopping malls, etc.)

poor (1), average (2), good (3) I4 Number of service sites (restaurants, cinemas,

etc.)

C2 I5 Housing architectural style houses (1), houses and high buildings (2),

high building (3)

C3

I6 Number of public transport none (1), small (2), reasonable (3) I7 Quality of roads low (1), average (2), good (3) I8 Wide of roads narrow (1), average (2), wide (3) I9 Wide of footpaths none (1), narrow (2), wide (3) I10 The number of over bridge none (1), small (2), reasonable (3)

I11 Bus lane

none (1), small (2), reasonable (3) I12 Vehicle lane

I13 Type of road network radial road network (1), ring road network

(3), both (2) I14 Number of lighting

none (1), little (2), reasonable (3) I15 Number of traffic sign

I16 Number of lanes (in main roads) 2 lanes (1), 3-4 lanes (2), >4 lanes (3)

C4

I17 The quality of public space

low quality (1), average quality (2), high

quality (3)

I18 The quality of schools I19 The quality of hospitals I20 The quality of playgrounds C5 I21 Density of green spaces along main roads low (1), average (2), high (3)

I22 The waste treatment process

none (1), average (2), good (3) I23 The water waste treatment process

I24 Water quality (in rives, lakes) heavy polluted (1), slight polluted (2), good

quality (3)

I25 Air quality polluted (1), only polluted in rush hours (2),

good

C6

I26 Electric capable network disordered (1), reasonable (2), ground capable

(3) I27 Internet capable network

I28 Water consumption rain water (1), tap water (3), both (2) I29 Energy consumption coal (1), green energy (3), both (2) C7 I30 Number of round space none (1), low (2), high (3)

2.3 Factor analysis

Factor analysis is a method of multi-criteria analysis, which allows reducing from a large number of variances to a smaller number of factors that account for the most of variance among the original data Factors are nominated by applying principal component analysis to a standardized correlation matrix A table of factor loadings shows which variables are grouped together on which common factors, and the degree of correlation between individual variables and the factors The factors are interpreted as axes in state spaces, and the meanings of the axes are inferred from the variables which are most correlated with them

* Factor matrix:









k

i

j i ij

x

1

 (1)

Where: x i are observed variables; fi are the common factors; α ij are factor loading of

factor x i; and ej is measurement error for x i

For factor analysis, to calculate the covariance of any two observable variables:

km jm k

j k j

jk

r  1 1  2 2    (2)

* Principal components analysis was used to determine eigenvalue and eigenvector

of correlation matrix Eigenvalue i show the proportion of the variance of xj Eigenvector show the attribute of component i







m

i

j ij

j w x

x

1

(3)

Where: x j are observed variables; z j are matrix components; wij are variables j loading of component i

* Varimax Rotation: Once principal components analysis had been completed, a variable in the n dimensional space specified by the factors involved, factor loadings are the

cosine of the angle formed by a vector from the origin to that coordinate and the factor axis Varimax rotation function is as follow (γ = 1):































k

j p

k

k

j ij ij

r

p AR agr

R

2

1

2 4

2.4 Hierarchical cluster analysis

A hierarchical cluster analysis based on the Euclidean nearest-neighbor distance attempted at identifying relatively homogeneous groups of small urban areas in Ha Dong district based on factor scores Euclidean nearest-neighbor distance is defined using simple Euclidean geometry as the shortest straight-line distance between a commune and its nearest neighbor Theoretically, an algorithm that starts with each of small urban area in a separate cluster and combines clusters until only one is left was used for this analysis Hierarchical clustering, consequently, created a hierarchy of clusters, which may be represented in a tree structure called a dendrogram

This analysis is based on Euclidean distance metrics as following:





i

i

i b a b

2

 

d a,b :a ,A bB

min

Where: a and b belongs to two sets of observations A and B; and d is the chosen

metric

Results of a hierarchical cluster analysis showed that there are four groups of the small urban areas as urban clusters

2.5 Carbon footprint (CF) calculation

Carbon footprint is a measurement of total cumulated GHGs emitted direct and indirect over times resulting from different human activities There are different methods of GHGs estimation depending on the object (individuals, residents, nations or companies, factories, economic sectors) The result of carbon footprint estimation becomes the foundation of implementing policies and strategies for the purpose of GHG reduction To quantify the impact of human activities and lifestyle of residents, the study use the method

of GHG estimation (with particular Vietnamese CO2e index) presented by Carbon Footprint Ltd., US (retrieved from http://www.carbonfootprint.com in 2015):

CF = CFH + CFT + CFS (5)

Table 3: The interpretation of sectors and variables in the carbon footprint questionnaire for

households

CF H Total amount of electricity consumption (kWh/person/month)

Total amount of GHGs released

from household energy

consumption

Total amount of natural gas consumption

(kWh/person/month) Total amount of coal consumption (ton/person/month)

CF T Total amount of GHGs released

from transportation

Private transportation:

- The average distance driven per month

- Type of private transportation: US car, EU car, average

car/van, motorbike, others

- Type of fossil fuel consumed: petrol, diesel, others

- Capacity:

+ Car: Large, average, small + Motorbike: <125cc, 125-150cc, >150cc

Private Transportation:

- The average distance driven per month

CF s Total amount of secondary GHG

from residential lifestyle

Food and drink products:

- Vegetarian

- Mainly consume fishes

- Mainly consume red meats

- Mainly consume white meats

- Consume both red and white meats Consume organic products: often/sometimes/never Consume seasonal products: only/try/never consume

Consume imported products:

- Not consume imported products

- Only using traditional products

- Most product are made in Vietnam

- Focus on the quality of product

- Not concern the origin of products Fashion: follow latest trend/only buy if necessary/only buy

old clothes Plastic bags: none use plastic bags/limit use plastic bags/hardly use plastic bags/ only buy if products are

wrapped by plastic bags Furniture and electric divides:

- Using latest products

- Bought new product and have used more than 5 years

- Only buy or change if necessary

- Only use old products Recycle products: All/Almost/Some/None products made

from recycled materials Recreation:

- Take part in activities which are none- GHG emission (e.g

cycling, walking, etc.)

- Sometime/often go shopping malls, restaurants, cinemas,

etc

- Take part in high GHG emission activities Total number of private transportation Using baking and finance services: Yes/No

(Source: http://www.carbonfootprint.com)

3 Results

3.1 Classifying urban small areas

As shown in the table 4, once factor analysis had completed, 19 out of 30 compact city variables were selected and quantified into 4 factors as follows:

 Factor 1: “the quality of infrastructure” includes 15 variables correlating and

showing the synchronization between infrastructure and urban landscape of individual residential areas

 Factor 2: “the density of open space” describes the appearance of open spaces

inside urban

 Factor 3: “transportation pattern” shows the distribution of two main road

types inside urban

 Factor 4: “public transportation and urban green space” is a dipole factor

showing the inverse correlation of two variables It indicates that public traffic spaces and urban green spaces are not planned logically

Table 4: Varimax rotated component matrix

Indicator Factor 1 Factor 2 Factor 3 Factor 4

I4 0.833 0.030 0.054 0.185 I8 0.828 0.270 0.256 0.006 I9 0.816 0.260 -0.018 -0.164 I16 0.801 0.276 0.141 -0.075 I24 0.799 -0.108 -0.363 -0.016 I3 0.796 0.185 0.185 -0.029 I30 0.778 0.168 0.012 0.067 I29 0.763 0.267 0.261 -0.053 I20 0.732 0.404 0.176 -0.099 I26 0.723 0.297 0.319 -0.088 I15 0.713 0.340 0.282 -0.112 I7 0.701 0.444 0.251 0.072 I5 0.682 0.186 0.320 0.188 I16 0.643 0.490 0.210 -0.184 I23 0.634 0.486 0.131 -0.066

I2 0.218 0.844 -0.125 0.132 I13 0.185 -0.057 0.890 0.079 I11 0.307 -0.006 0.108 0.858 I21 0.339 -0.067 0.019 -0.670

Table 5 Total Variance Explained

Squared Loadings Total Percentage

of variance

Cumulative Total Percentage

of variance

Cumulative

Based on the mathematical relationship between 4 factors and the 2010 land use map, four urban classification maps were created (as shown in figures 2, 3, 4, and 5)

Dancu by Fac1

1.23 to 2.23 (10) -0.55 to 0.12 (22) -1 to -0.55 (17)

Figure 2: Map of the quality of infrastructure (Factor 1)

Dancu by Fac2

0.7 to 2.23 (16) -0.16 to 0.7 (20) -0.72 to -0.16 (16) -1 to -0.72 (13)

Figure 3: Map of the density of open space (Factor 2)

Dancu by Fac3 2.62 to 3.5 (5) -0.02 to 2.62 (18) -0.49 to -0.02 (22) -1 to -0.49 (20)

Figure 4: Map of transportation pattern (Factor 3)

Dancu by Fac4 1.27 to 2.54 (8) -0.01 to 1.27 (22)

-1 to -0.57 (25)

Figure 5: Maps of public transportation and urban green space (Factor 4)

The total percentage of variability of factor 1 accounts for 75 percent Moreover, factor 1 describles a direct correlation Thus, small urban areas in Ha Dong are grouped into

4 urban clusters based on the quality of infrastructures (As shown in figure 6)

Urban cluster C1 includes new small urban areas having similar properties in

complex building each of them is a combination of housings, commercial zones, schools, hospitals, sport centers, playground, and etc In addition, there are several urban green spaces and underground spaces in cluster C1 Overall, these areas have good quality infrastructures, creating livable environment for residents as well as promoting socioeconomic development

Urban cluster C2 includes old small urban areas with the high diversity in its pattern:

complex buildings, containing housings, commercial zones, schools, hospitals,