LAND COVER CHANGE AND THE CO2 STOCK IN THE CO2 STOCK IN THE PALEMBANG CITY, INDONESIA a STUDY USING REMOTE SENSING GIS TECHNIQUE AND LUMENS

THAI NGUYEN UNIVERSITY UNIVERSITY OF AGRICULTURE AND FORESTRY THINH DUC NGUYEN TOPIC TITLE: LAND COVER CHANGE AND THE CO 2 STOCK IN THE PALEMBANG CITY, INDONESIA: A STUDY USING REMOTE

THAI NGUYEN UNIVERSITY

UNIVERSITY OF AGRICULTURE AND FORESTRY

THINH DUC NGUYEN

TOPIC TITLE: LAND COVER CHANGE AND THE CO 2 STOCK IN THE PALEMBANG CITY, INDONESIA: A STUDY USING REMOTE SENSING,

GIS TECHNIQUE AND LUMENS.

BACHELOR THESIS

Study mode: Full-time

Major: Environmental Science and Management

Faculty: International Training and Development Center

Batch: 2010 – 2015

DOCUMENTATION PAGE WITH ABSTRACT Thai Nguyen University of Agriculture and Forestry (TUAF)

Degree Program Bachelor of Environmental Science and Management

Thesis Title Land Cover Change and The CO2 Stock in the Palembang City,

Indonesia: A study using Remote Sensing, GIS technique and LUMENs

Supervisor (s) 1 Dr Iskhaq Iskandar1

2 Dr Ho Ngoc Son2

ABSTRACT

A study of the land cover change and the CO2 stock in the Palembang City, Indonesia was conducted by using remote sensing and GIS data and processed by the LUMENs software The study has successfully compared the land transformation during period of 1989 – 2000 and 2000 – 2013 It reveals that the settlement area increased much faster during 2000 – 2013 compared to that of 1989 – 2000 During

2000 – 2013, a rapid development of the Palembang City after hosting the National Olympic Games in 2004 has a significant impact on the land cover change mainly from the plantation area into the settlement area In addition, the rate of CO2 emission per unit-area during the 2000 – 2013 period is also higher than that of the 1989 - 2000 period due to the increase of transformation of plantation into settlement

ACKNOWLEDGEMENT

In order to complete this research, first, I am thankful to the corporation between the Thai Nguyen University of Agriculture and Forestry (TUAF) and the University of Sriwijaya (UNSRI), Palembang, Indonesia I appreciate the hospitality of the University of Sriwijaya, which provided a suitable environment for my research and also an accommodation during my stay in UNSRI from September 2014 to January

2015

Second, I would like to express my deep gratitude to my warm-hearted supervisors, Dr Iskhaq Iskandar and Dr Ho Ngoc Son It is clearly that I could never complete my research without the whole-hearted supervision of my supervisors In addition, I want to express my special thanks to Dr Iskhaq Iskandar for sharing and helping me to overcome difficulties in life during my stay in UNSRI

Last, I would like to thank all friends who have supported me to overcome my difficulties while I was conducting my research in Palembang, Indonesia

Thai Nguyen, 22/01/2015

Author

Thinh Duc Nguyen

TABLE OF CONTENT

LIST OF FIGURES iv

LIST OF TABLES v

LIST OF ABBREVIATIONS vi

PART I INTRODUCTION 1

1.1 Problem statement .1

1.2 Study area 1

1.3 Objectives 4

1.4 Research questions 4

1.5 Limitations .5

PART II LITERATURE REVIEW 6

2.1 The necessary of land cover change study 6

2.2 CO2 emissions from land cover change 11

2.3 GIS, Remote sensing and LUMENS 12

2.3.1 GIS and Remote Sensing 12

2.3.2 LUMENS 13

2.4 Land cover change study by remote sensing data 14

PART III METHODS 15

3.1 Materials 15

3.2 Methodology 15

3.2.1 Data acquisition 15

3.2.2 Geo-referencing and visual interpretation classification 16

3.2.3 Ground Truthing Verification 18

3.2.4 Land cover change analysis and carbon stock 21

PART IV RESULTS 23

4.1 The land cover change in the Palembang city 23

4.2 Carbon stock 32

PART V DISCUSSION AND CONCLUSION 38

5.1 Discussion 38

5.2 Conclusion 38

REFERENCES 40

LIST OF FIGURES

Figure

1.1 Map of the study area 3

3.1 Map of the Waypoints of the ground truthing in the Palembang city 21

3.2 The flow chart of methodology 22

4.1 Map of the land cover change during 1989 – 2013 25

4.2 Map of the land cover change during 1989 – 2000 28

4.3 Map of the land cover change during 2000 – 2013 31

4.4 Map of the CO2 emission during 1989 - 2013 33

4.5 Map of the CO2 emission during 1989 – 2000 35

4.6 Map of the CO2 emission during 2000 – 2013 37

LIST OF TABLES

Table

2.1 Definition of land cover type (Ministry of Forestry of the Republic of

Indonesia) 7

2.2 The top 10 countries with largest annual net loss of forest area 1990 – 2000 (U.N Food and Agriculture Organization Global Forest Resources Assessment 2010) .9

2.3 History of the Indonesian population 1990 – 2013 (Countrymeter, Indonesia Population Clock, 2014) 10

2.4 The carbon stock of the land cover in Palembang city (World Agroforestry Center, 2014) 11

3.1 Satellite imageries used in this study 15

3.2 Type of Land Cover in the Palembang City based on the Satellite Imagery 17

3.3 Waypoints in each land cover types 19

4.1 The statistics of the land cover change during 1989 – 2013 23

4.2 The matrix of the land cover change during 1989 – 2013 24

4.3 The statistics of the land cover change during 1989 – 2000 26

4.4 The matrix of the land cover change during 1989 – 2000 27

4.5 The statistics of the land cover change during 2000 – 2013 29

4.6 The matrix of the land cover change during 2000 – 2013 30

4.7 Statistics of CO2 emission during 1989 – 2013 32

4.8 Statistics of the CO2 emission during 1989-2000 32

4.9 Statistic of the CO2 emission during 2000 – 2013 36

Land Sat 7 SLC on Land Sat 7 Scan Line Corrector

Imager sensor

Environmental Services

PART I INTRODUCTION

1.1 Problem statement

Since 1989, a rapid development on economics has been observed in the Palembang city This rapid development leads to problems in plan and management of the city As the economics has grown up very fast, the migration of population from countryside to the city has also significantly increases In turn, it causes a heavy pressure on the infrastructure, environment, and community This situation is a common attached consequence of the urbanization in the world and the Palembang city is not excluded In order to provide potential information for managing and planning of the administration of the city, the land cover change of the Palembang city from 1989 to 2000 and from 2000 to 2013 are studied It is believed that the vegetation change is not only affect environment and community locally but also globally The deforestation due to urbanization is believed to contribute to the global climate change Therefore, in this study, an estimation of the CO2 emission from land cover change in Palembang city will also be conducted

1.2 Study area

Palembang city is the capital city of the South Sumatra Province, Indonesia It is the second largest city in the Sumatra Island The city is located between 104°36'38.479"E, 104°51'50.753"E and 2°52'4.159"S, 3°5'26.534"S with an area of about 400.61 km2 (Putra et al., 2011) The altitude of the city ranges between 8 to 27 meters above sea level The population in Palembang City is the highest in the South Sumatra Province with density of 3,758 people per square km (Putra et al., 2011)

Located in the tropical zone, the Palembang city’s climate has only two seasons, namely: the rainy and dry seasons The hydrology of the Palembang is characterized

by the Musi River that divides the Palembang City into two major areas; the Palembang Ulu (upstream) and the Palembang Ilir (downstream) Note that the Musi River is one of the biggest rivers in the Indonesia

In addition, most of the Palembang area is low land Therefore, flood is a major problem of the Palembang City during the rainy season This requires an integrated management of the city to avoid those impacts in the future development plans

Figure 1.1 Map of the study area

1.3 Objectives

The research is designed to provide the detail information about the land cover change in the Palembang City that can be used for the management and future development plan of the city We approach the goal through two research objectives:

a Assess to what extent the land cover change in the Palembang City, in particular for period from 1989 to 2000 and from 2000 to 2013

b Quantity the amount of CO2 emission during two time periods: 1989 – 2000, and 2000 – 2013

c Evaluation of how the land covers change (e.g vegetation change) influences the CO2 emission change

1.4 Research questions

This study is designed to address the following questions:

a What is the area of vegetation change from the land cover change map of 1989

to 2000 and 2000 to 2013?

b What is the total amount of CO2 emitted by the vegetation change?

c Is the CO2 emission on the increase due to the land covers change during 1989 – 2000 and 2000 – 2013?

1.5 Limitations

The study area is limited to the Palembang city, which is very small and there is not much change in vegetation In addition, the horizontal as well as the temporal resolutions of the satellite used in this study are low Therefore, it might not precisely represent the land cover as well as vegetation changes

PART II LITERATURE REVIEW

2.1 The necessary of land cover change study

Land cover is the natural vegetation which is on the earth surface Due to human activities such as development of settlement, the land cover now is not only naturally but also artificially In addition, the process of land use change causes the change in land cover Therefore land cover also describes how land has been used

The land cover is defined as the assemblage of biotic and abiotic components on the earth surface which is one of the most crucial properties of the ecosystem on the earth (Pandian et al., 2014) The component of “land cover” in Palembang city includes secondary forest, grassland, open area (bare soil), water body, plantation, settlement, shrub and swamp (Ministry of Forestry of the Republic Indonesia) The definition of each land cover type in this study is described in the following Table 2.1

Table 2.1 Definition of land cover type

1 Water body Waters, including sea, river, lake, irrigation, etc Fishpond

appearance, rice field and swamps were classified

2 Settlement The area which uses as a residential, including urban, rural,

industrial, public area etc It’s seen clearly

3 Shrub Dry forest which was growth back (has been succession) but it’s

not optimal, or the dry land with sparse tree coverage (nature) or dry land with dominance of low vegetation (nature) There is logged

4 Swamp Swamp area that has not been forested In the study for land cover,

swamp also includes rice field in wetland

5 Plantation Plantation is planted trees area which have proprietary or other

right with coverage dominated by plantation or industry plant Plantation in Palembang city is mainly palm oil and rubber tree

8 Open Area Open area without vegetation (mountaintop rock outcrop, snowy

mountaintop, volcanic crater, burned sand, beach sand, stream sediment), and burned of open area The appearance of open area for mining is classed as mining class, meanwhile, open area that is the example of land clearing classed as open area class Open area

in the rice plant rotation and fishpond are still classed as swamp

Source: Ministry of Forestry of the Republic of Indonesia

The change in land cover is believed due to accessibility and history of urban growth which are the main factors in an arrangement of land use patterns (Bijender and Joginder, 2014) Sreenivasulu et al (2013) has suggested that the advancement of vegetation mapping has greatly increased and research on land use/land cover change thus providing an accurate evaluation of the spread and health of the world’s forest, grassland, and agricultural land use has become an important priority

Due to the urbanization and human activities, the land cover has been changed all over the world In details, the area of forest coverage has decreased 83 million hectares from 1990 to 2000 and 52 million hectares between 2000 and 2010 (Adams, 2012) These changes significantly affect the earth system functioning (Lambin et al., 2000) They influence on the biotic diversity, habitats degradation, and environmental pollution In addition, as a consequence of deforestation, green houses gases emission contributes to the global climate change

Indonesia is the largest archipelago in the world and it is now with a fast growth

in economic Located in tropical zone, Indonesia is dominated by tropical rainforests and jungles, as well as swampy mangrove areas However, in recent decades these dense biodiversity habitats are now devastating by new infrastructure systems for urbanization Furthermore, the economics of Indonesia is mainly based on agriculture and palm oil plantation with approximately 90% of the population engaged in agriculture It is claimed that palm oil production is one of the large driver of deforestation in Indonesia (Indonesia Facts, 2014) Almost half of the world output for palm oil production is accounted for Indonesia As a consequence of palm oil largely planted on land, forest has been logged or burned and threatened remain forest

According to the data from FAO, Indonesia is on the top ten countries of forest area loss during two periods of 1990 to 2000 and 2000 to 2010 as shown in Table 2.2

Table 2.2 The top 10 countries with largest annual net loss of forest area 1990 – 2000 Country Annual Change 1990-2000 Country Annual Change 2000-

Table 2.2 shows that in the period of 1990-2000, the forest loss rate in Indonesia

is the second in the world In the next 10 years, the rate was decreasing and it became

in the world Nevertheless, the forest area loss needs to be controlled Therefore in order to be able to assess the risk of land use change as well as to reduce the contribution of land use change to global, Indonesia has started an investigation since

2011 and stopped providing new licenses to convert primary forests to palm oil or

other uses in two years (Adams, 2012) In addition, to reach the global goals for development, many international organizations such as World Agroforestry Center and GIZ as well as national organizations were born to help the developing countries These are inter-government organizations that conduct the research and exert interventions to government for development plans in the future

Due to immense agricultural and demographic pressure, land is becoming an increasingly scarce resource (Bijender and Joginder, 2014) In Indonesia, there is no exception when the country is over 250 million people, one of the fourth populous countries in the world Indonesia history population from last 20 years is shown in Table 2.3

Table 2.3 History of the Indonesian population 1990 – 2013

Source: Countrymeter, Indonesia Population Clock, 2014

As shown in Table 2.3, the growth rate of population is on the decline However the total population is on an increase which requires more space for the settlement Thus the study of land cover change provides useful information for the tendency of settlement distribution

2.2 CO 2 emissions from land cover change

The estimation of CO2 from land cover change still has a lot of uncertainties, not only uncertainty in rate of deforestation and reforestation, but also uncertainty on the carbon density of land undergoing change (Houghton et al., 2012) However, the estimation is necessary for the land management to reach the millennium goals The estimation for CO2 emission in land cover change study is possible because there is the steady state soil carbon stock for each land use For example, the secondary forest accumulates an amount of about 141.3 ton C/ha and the carbon stock of an open area

is about 3.35 ton C/ha The principle of CO2 emission estimation is that if one hectare

of the secondary forest is converted into one hectare of an open area, the carbon emission equivalents to the subtraction of these two types of carbon stock Thus, the amount of CO2 emission for a conversion of each hectare of secondary forest to an open area, the amount of emission is equivalent to 137.95 ton/ha The reference of carbon stock for each land cover is in the Table 2.4

Table 2.4.The carbon stock of the land cover in Palembang city

2.3 GIS, Remote sensing and LUMENS

2.3.1 GIS and Remote Sensing

Geographic Information System (GIS) is a system that is used for a management, analysis, and display the geographic information Geographic information is represented by a series of geographic datasets that models geography using a simple and generic data structures GIS includes a set of comprehensive tools for working with the geographic data

The spatial data for research are imageries from land sat Land sat is series of satellite for earth resource observation such as land and forest The first land sat had been launched in 1972 and finished mission in 1978 The latest one (Land sat 8) has been launched in 2013 with a significance in technology (USGS, 2013) With the help

of Geographic Information System (GIS) and Remote Sensing data, land use/land cover change study becomes more practical because the cost is cheaper and time consuming is shorter than manual method

The processing of remote sensing data facilitates a synoptic analysis of the earth system, and patterning a change at local, regional, and global scales over time Remote sensing data also provides an important links between intensive localized ecological research and regional, national and international conservation and management of biological diversity

In this study, the land cover change study in Palembang city was analyzed using remote sensing technique to monitor land use changes which is vital role in urban development (Singh and Dubey, 2012) From the land use change database, the n carbon stock and emission estimation were estimated by using LUMENS software

2.3.2 LUMENS

LUMENS stands for the land use planning for multiple environmental services

The software is released by World Agroforestry Center in 2014 LUMENS is revised

version of LUWES (Land use planning for low emission and development strategy),

which contains four modules, namely: (1) PUR – Planning Unit Reconciliation, (2) QUES- Quantification of Environmental Services, (3) TA – Trade-off Analysis, and (4) SCIENDO – Scenario simulation and development This study was used pre-QUES and QUES-C that are two sub-modules for land use change analysis and carbon accounting respectively The advantage of LUMENS is an open source and every reported analysis is reported fully information about land cover change LUMENS is a free-tool for sustainable management which helps the planners, managers and policy makers in decision-making This is a land use planning tool that helps in achieving sustainable landscape, simulating scenario of land cover changes and estimating the carbon storage in land cover which in turn mapping the carbon dioxide emission from change of land cover (World Agroforestry Center, 2014) The use of LUMENS is still

at the beginning, and thus no research has been reported so far Nevertheless, some projects from World Agroforestry Center were using LUMENS for the land planning purpose and the carbon stock in Indonesia The research of land cover change and carbon stock in Palembang city is a very first contribution for the development of LUMENS for new version in the future

2.4 Land cover change study by remote sensing data

Since the first land sat launch in 1972, the application of remote sensing data in the research activity and land resource management has been fostered The change of urban land in a lot of cities has been studied about the change of urban land to provide the information for sustainable management The study of land cover change in New Delhi from 1992 to 2004, for example, has used data from Landsat Thematic Mapper and IRS – P6 LISS III (Bijender and Joginder, 2014) The method for New Delhi land cover classification was visual classification and ground truthing The autho rs used visual classification for land cover type interpretation of Delhi city It is claimed that the visual classification of various classes has been improved with ground truthing which gave a better map for land cover in the study area In another study, Singh and Dubey (2012) have shown that land use and land cover change using an unsupervised classification is simpler than that using a supervised classification They suggested that the unsupervised classification automatically generates information by the ISODATA algorithm, in such that with the help of visual interpretation and standard interpretation such as tone or color, shape, size, texture, shadow, pattern site or aspect, and various associated features The study of land cover change with the use of visual interpretation, however, is limited when the remote sensing imagery contains haze or noise Therefore, imageries from Indian Remote Sensing Satellite (IRS) were visually and digitally interpreted by using the image interpretation elements (Sree nivasulu et al., 2013)

PART III METHODS

3.1 Materials

In order to evaluate the land cover change and to estimate the CO2 emission in Palembang City, this study mostly relies on the satellite remote sensing The following data were used in this study:

a Satellite imagery from Land Sat 5, 7 and 8 on the path 124 and row 062 for a period of 1989, 2000, and 2013 respectively

b Software Arc-GIS 10.2 and LUMENS 0.1;

c Geographic Information System (GIS) data;

d CO2 storage per hectare for each land covers type

3.2 Methodology

3.2.1 Data acquisition

In order to obtain good quality imageries from satellites, three imageries from Land Sat 5 TM, 7 SLC on and 8 OLI were chosen in this study Detailed of the satellite imagery data is presented in the Table 3.1:

Table 3.1 Satellite imageries used in this study

3.2.2 Geo-referencing and visual interpretation classification

The satellite imageries from the Land Sat cover a large area in the Sumatra Island However, this study only focused on the Palembang City The first step is a geo-referencing in which we register these imageries with the boundary and the water body vectors of the Palembang city This geo-referencing process is to register the boundary of Palembang city to be exact in the imageries In this study, we used the vector of the Musi River in the Palembang city to concur with the river in the Land sat imagery After registration, the study area was clipped study area by boundary vector The next step is the ground truthing in which we verified the defined land cover from satellite imagery with the real condition However, before ground truthing verification, we classified the land cover types of the Palembang city based on the difference from color, texture, pattern, tone, shape, etc The different appearance of each land cover types are defined in the Table 3.2:

Table 3.2 Type of Land Cover in the Palembang City based on the Satellite Imagery

Water body The object appearance is indicated by light blue area, and blue tend to

be white or black (in 543 band combination) Object covers an area large enough

Settlement That is indicated by set of closed building pattern in urban area and

closed street network Sparse residential in the rural area and there is road pattern which connect to other residential The appearance of objects is in light red color in 543 combination band

Shrub Object appearance is indicated by low vegetation and finely textured

until rather coarse, in the light green at 543 bands that indicate s shrubs and there is logged The light green tends to appear discontinuously and usually mix in the settlement

Swamp In wet condition, the appearance of swamp is very specifically There is

flooding that sometimes covers an area large enough and indicates by black in combination of 543 band land sat imagery On the other hand,

in dry condition, the flood will be seen red or brown in combination of

543 bands

Plantation Cocoa plantation appearance indicates with the light green object with

yellowish brown spotting (543 band) which tends to be light with finely textured The boundaries are clear and regularly shown that the object is

a plantation Secondary

Forest

The characteristics of secondary forest is characterized by dark green object (in the 543 band) tend to be dark and coarse textured with tree crown that appear clustered There is logged In the imagery, colors tend to be dark because of the position of the object is on a high mountain cliff so less sunlight

Grassland Finely textured vegetation which have dark green but little bit soft and

sometimes light yellow (in the 543 band) In wet condition, the marsh grass showed specifically, but in dry conditions quite difficult to distinguish with open area because of having same color in the 543 band of land sat imagery Therefore, the supporting data like field image is necessary

Open Area The object appearance ( in the 543 band land sat imagery) is indicated

by light red until dark red area, sometimes brown, depends on soil material content and white if the material is content lime At the dry land, there is red area which indicate open land or residential

Note that there is one difficulty from visual classification for open area, grassland and settlement because their appearances in satellite imagery are the same color in