You are on page 1of 74






JULY 2015
The thesis of Sopheak Pen has been approved by the examining committee
to be partial fulfillment of the requirement for the Master Degree of Science Program
in Geoinformatics of Burapha University

Advisory Committee

Principal advisor
(Dr.Supan Karnchanasutham)

(Associate Professor Dr.Kaew Nualchawee)

(Dr.Narong Pleerux)

This thesis has been approved by the Faculty of Geoinformatics to be partial

fulfillment of the requirements for the Master Degree of Science Program in
Geoinformatics of Burapha University

Dean of the Faculty of Geoinformatics

(Dr.Supan Karnchanasutham)
July, 2015

First of all, I would like to express my sincere gratitude and deepest

appreciation to Her Royal Highness Princess Mahachakri Sirindhorn, who provided
the scholarship for me to study at Burapha University.
I would like to express my special gratitude and deep appreciation to
Dr. Supan Karnchanasutham, Assoc. Prof. Dr. Kaew Nualchawee, Dr. Narong
Pleerux, who always help me to minimize the difficulties with friendly for their
guidance, and valuable advice through this study.
I would like to thank Mr. Preecha Boonkhaw and Nut'Narathip Phengphit
who helped and facilitated in the process of data collection for valuable advice and
enlighten on the technique.
I am very grateful to all professors and members of Faculty of
Geoinformatics who helped and supported me while I was studying here.
Finally, I would like to express my deepest gratitude to my beloved mother,
Nget Eng and father Pen Chhoern, who always gives me the warmest and greatest
love and support. Unforgettable, I offer special thanks to all my relatives and friends
for your help and encouragement.

Sopheak Pen

Cassava is currently the most important upland crop of Cambodia. It's an

agricultural product that can be process into various other products such as ethanol,
animal feed and cassava starch or flour for human consumptions. The object of this
research were to (1) classify the cassava plantation areas in Banteay Meanchey
Province, Cambodia using LANDSAT 8 (OLI) and SMMS (HJ-A1) images and (2)
compare the cassava plantation areas between LANDSAT 8 (OLI) and SMMS (HJ-
A1). The Maximum likelihood classification technique was applied to this research.
Land use types were evaluated into seven classes: cassava, field crop, forest, water,
perennial tree/fruit tree, rice and urban.
The result found that the cassava areas from LANDSAT 8 (OLI) were
83757.37 hectares or 13.54 % of study areas. Meanwhile the cassava areas from
MSSM (HJ-A1) were 97,215.33 hectares or 15.72 % of study areas. The overall
accuracy of LANDSAT 8 (OLI) and SMMS (HJ-A1) was 81.48 % and 75.56 %
respectively. Therefore LANDSAT 8 (OLI) can be used to classify land use/land
cover with higher accuracy than SMMS (HJ-A1).


Background to the study 2
Statement of the Problem 3
Objectives of the Study 4
The Study Area 4
Benefits of the Study 5
Cassava 7
Cassava production in Cambodia 8
Cultivation Practices 9
Geographic Information System and Remote Sensing 10
Geographic Information System 10
Components of Geographic Information System 11
Spatial data 12
Attribute data 13
Electronic maps and images 14
Working GIS 14
Remote sensing 15
Huanjing-1A (HJ-1A) 17
Orbit Characteristic of HJ-1A 19
Landsat 8 20
Band combinations for Landsat 8 21
Global Positioning System 22
Literature Review 24
Geocorrection land of image 30
Data exploration and preprocessing 31
Data exploration 31
Stretching of band data 32
Creating a multiband image 32
Collecting training samples 32
Evaluating training samples 33
Creating the signature file 34
Examining the signature file 34
Editing the signature file 34
Application classify 34
Post-classification processing 35
Filtering the classified output 35
Smoothing class boundaries 35
Generalizing output by removing small isolated regions 35
Equipment of Analysis 36
Land use from SMMS (Hj-A1) 38
Land use from image satellite Landsat 8 (OLI) 41
Land use comparing of SMMS (Hj-a1) Landsat 8 (OLI) 45
Values accuracy of model 46
Recommendation 51
Appendix cassava plantation in Banteay Meanchey Province 57

Tables Page
2-1. Payload Parameters (Rahman, 2012) 19
2-2 Orbit parameters (Rahman, 2012) 19
2-3 Landsat8 spectral band/wavelengths 21
2-4 Composition Band 22
3-1 The using software 37
4-1 Land use of Banteay Meanchey province from SMMS (HJ-A1) 2015 39
4-2 Land use of Banteay Meanchey province from image satellite Landsat 8
(OLI) 42
4-3 Comparison of land use Banteay Meanchey province image satellite
Landsat 8 (OLI) and SMMS (Hj-A1) 45
4-4 Dislocation evaluate the accuracy of the classification land use in SMMS
(Hj-A1) Satellite 47
4-5 Dislocation evaluate the accuracy of the classification land use in image
satellite Landsat 8 (OLI) 48

Figures Page
1-1 Map of Banteay Meanchey Province 5
2-1 Components of GIS (Mapsofindia, 2012) 11
2-2 GIS Working 15
3-1 Classification of workflow 28
3-2 SMMS (Hj-A1) 29
3-3 Enhanced Landsat operational land imager (OLI) and the thermal
infrared sensor (TIRS) image of 2015 30
3-4 Collecting training sample 32
3-5 Evaluation training sample 33
3-6 Editing class 33
3-7 Creating the signature file 34
4-1 Percentage of land use Banteay Meanchey SMMS (HJ-A1) 39
4-2 Area of land use Banteay Meanchey SMMS (HJ-A1) 40
4-3 Map SMMS (HJ-A1) of Banteay Meanchey showing location of cassava
plantation fields of studies 40
4-4 Map SMMS (HJ-A1) classification of Banteay Meanchey showing
location fields of studies 41
4-5 Percentage of land use Banteay Meanchey image satellite Landsat 8
(OLI) 43
4-6 Area of land use Banteay Meanchey image satellite Landsat 8 (OLI) 43
4-7 The map of satellite Landsat 8 (OLI) at Banteay Meanchey showing
location of cassava plantation fields of studies 44
4-8 The map of image satellite Landsat 8 (OLI) of Banteay Meanchey
province showing location fields of studies 44
4-9 Comparison of land use Banteay Meanchey province image satellite
Landsat 8 (OLI) and SMMS (Hj-A1) 46
4-10 The map shows the sample location to determine accuracy 47

GPS Global Positioning System

GIS Geographic Information Systems
RS Remote Sensing
FAO Food and Agriculture organization
HJ-1A Huan Jing 1A
MAFF Cambodian Ministry of Agriculture Fish.
SMMS Small Multi-Mission Satellite
USGS United States Geological Survey
UTM Universal Transverse Mercator

Background to the study

Cambodia is a country in Southeastern Asia, bordering the Gulf of Thailand,
between Thailand, Vietnam, and Laos. It has twenty four provinces and capital city.
Cambodia occupies 181,035 square kilometers in the southwestern part of the
Indochina peninsula, and is located between 10° and 15° latitude north and 102° and
108° longitude east. The country shares 803-kilometers border with Thailand on north
and west, 541kilometers border with Laos on northeast, 1,228-kilometers border with
Vietnam on east and southeast, for a total of 2,572 kilometers of land borders;
coastline along Gulf of Thailand about 443 square kilometers. The population of
Cambodia is 15,184,116 million (Cambodia, 2014) Cambodia has a tropical climate
with two distinct monsoon seasons the rainy season starts in mid-April and continues
to October. Average annual rainfall is 1,250-1,750 millimeter (MAFF). Sihanouk
Ville has the highest average annual rainfall of 2,996 millimeter and the average
rainfall in Banteay Meanchey Province is 1,000 millimeter, with peak rainfall
occurring in September/October and the lowest rainfall in February. As for the
temperature, it is lowest in December/January with average minimum temperature of
21 degrees celsius and the highest in April with average minimum of 36 degrees
celsius (Yem, 2010). The Cambodia is located tropical area is suitable for several kind
of economic plants such as soybeans, green bean, rice, cassava, corn (maize) and
sugarcane. Agriculture has been the first priority of the government’s development
strategies since1993. Agriculture is the fundamental sector of Cambodian economy
and the small farmers dominate the agricultural sector of the country.
The territory of Banteay Meanchey is a Cambodian province in the
northwest of the country, in 1988 the province Banteay Meanchey was split off from
Battambang, and its capital is named Sisophon, approximately 359 kilometers from
Phnom Penh by National Road Number 5. The district is subdivided into 7 and 2 city,
communes 64 and 649 villages (National Institute of Statistice, 2013). One of the
most known places in that province is Poi Pet, a town on the Thailand/Cambodia
border. It’s the key crossing point between the two countries. The total population is
745,618or (5.242%) of the total population 14,363,519 in Cambodia (2011, provincial
government data) with a growth rate of 5.93 %, which is consisting of 402,201 male
person (49.11%), and 414,181 female person (50.89%). The above number also
consists of 654, 033 person (93%), who are farmers, 8,228 person (1.17%), who are
fishermen, 35,162 person (5%), who are traders, and 5,814 person (0.83%), who work
as government officers. (Tourism of Cambodia , 2015)
Cassava is currently the most important upland crop of Cambodia, it’s an
agricultural product that can be process into various other products such as ethanol,
animal feed or cassava starch or flour for human consumptions. Cassava (Manihot
esculenta Crantz), a native to South America (Allem, 2002), is an important storage
root crop worldwide (Ceballos et al., 2004; El-Sharkawy, 2012). It is a key
component of the diet of over 800 million people across several continents (El
Sharkawy, 2012). The crop is a high starch producer with levels of up to 90% of its
total storage root dry mass (Jansson et al., 2009), Cassava is the third most important
source of calories in the tropics after rice and maize (FAO). Cassava is one of the
most important upland crops of Cambodia that a farmer plants it after rice. More than
85 % of cultivated area is planted to Rice, Maize, Moonbeam, Soybean, Sesame,
Peanut and Cassava. Agricultural is also the most important sector for employment,
employing more than half of the country’s total labor force. Agriculture is more
important for the rural poor as it provide their most important source of income
(World Bank , 2009).
Land cover is the physical material at the surface of the earth. Land covers
include grass, asphalt, trees, bare ground, water, etc. These study our uses Remote
Sensing data from Landsat imagery covering the study area. Remote sensing (RS) and
Geographic Information Systems (GIS) provided secure and established foundations
for measurement, mapping and analysis of natural resources in the world. There are
various ways to classify land cover based on remote sensing and GIS such as
supervised classification, unsupervised classification or combination of supervised
and unsupervised procedures with other sources like economic, social and historical
information as hybrid classification methods which are well known and established
today (Nguyen, 2015).
This research thesis is concerned with the application of recent technology
Remote Sensing and the Geographical Information System in the evaluation cassava
plantation areas in Banteay Meanchey Province in Cambodia.
The capability of remote sensing to capture up to date information for large area and
the ability of GIS to store and manager data from various sources mean the
technology are an important tool, in the field of agricultural and another field.

Statement of the Problem

Cambodia’s agriculture sector is growing, but suffers from constraints such
as low labor productivity, low yields, variable water resources, inefficient land titling,
inactive technology transfer, limited access to credit and inadequate rural
infrastructure (CDC, 2010), which make it very difficult to identify cassava plantation
There are many question that could be ask with may lead to some solution to
the above problem. The purpose of this thesis is to identify the land surface and
approach of economic cassava plant have planted in Banteay Meanchey province of
Cambodia country which is a member of Association of South-East Asian Nations
(ASEAN) and shares border to Thailand. The basic hypothesis of this research is a
block of land in Banteay Meanchey province (the north high land of Cambodia),
tropical climate and the gate of import (the seed) and export the cassava products
which is very important in sharing the economic benefit to together with Thailand
also. It is proposed the well measurement of how much the cassava plants land used
is, which area for the specific plant as cassava is, and economic benefit are provided
to farmers in this province. The analysis is divided into two main parts. The first
section of this report introduces the cause of the choosing this area to be the objective
of the thesis. It will explain the land used view by using Geographic Information
System and Remote Sensing satellite imagery. The second section is a case evaluation
of cassava plantation area to the farmers – increasing the side land plating and going
to the income of them. The effects, implementation strategies, and specific strengths
and weaknesses associated with each approach are described, including a comparison
from the previous work. This study entails an examination of the land used developed
in increasing total area of cassava plantation. In general the associated finding of my
research is located within this final section.

Objectives of the Study

The main objective of this study is to Classification of cassava planted area
using Landsat data in Banteay Meanchey Province by using likelihood supervisor
classification by gathered the satelliteHuanjing-1A (HJ-1A) 2015 and Landsat OLI
satellite image data using the year 2015 and field observation.
1. To classify cassava area using satellite image Landsat 8 and satellite
Huanjing-1A (HJ-1A) or Small Multi Mission Spacecraft (SMMS)
2. To comparing cassava area between satellite Landsat 8 and SMMS (HJ-

The Study Area

This study was carried out in Banteay Meanchey Province, having area
6185.84 square kilometers is located between 13°45′- 13°75′ N latitude and 101°06-
105°21′E longitude, this province was chosen because it has the largest area cassava
planted in Cambodia after Kampong Cham province and Battambang province that
have cassava planted approximately 62151 hectares in 2014 and shares a border with
Sa Kaeo Province of Thailand. The west and North, with Oddar Meancheay to the
North, with Siem Reap to the East and with Battambang to the South. The town of
Sisophorn is about 359 kilometers from Phnom Penh via national road number 5. The
climate of the area can be characterized as the tropical warm, wet and dry climate
with extreme low and high temperatures of 23 degrees celsius to 36 degrees celsius
and rainfall in this province 885.30 millimeters in per year (Cambodia T. o., 2015)

Figure 1.1 Map of Banteay Meanchey Province

Benefits of the Study

1. The output of this study is very importance for farmer and residents in the
area first to get useful knowledge related to plant cassava of surface land use and land
suitable for agriculture by using geographic information system and remote sensing
satellite imagery.
2. To know about area cassava planted at Banteay Meanchey
3. The farmers can divide the time to cultivate crops other very well and to
realize that crop kind can grow combination with cassava that give farmers can
expand the area of the left and these crops turning more and more cassava as well as
get a higher yield, improve their livelihoods. These maps could serve as input or
guides in the planning and formulation of sustainable management strategies.
4. To get Comparing cassava area between satellite Landsat 8 and SMMS
5. To get map area of cassava area planted after classification at Banteay
Meanchey, Cambodia

Agriculture is the fundamental sector of Cambodia economy. Small farmer

dominate the agriculture sector of country and the most of farmer in Banteay
Meanchey province like planted cassava more than rice. Cassava is a source of
income high for people live at there that yield and production were high especially
near border Thailand. Easy to manage to grow cassava and divide farmland to better
bring technology remote sensing and the Geographical Information System (GIS) for
analysis in the evaluation cassava plantation. Remote sensing and GIS are effective
tools to generate, analyze, and display these multi-disciplined spatially correlated
data. In the last several decades, methodologies have been developed and studies have
been conducted for effective analysis by using remote sensing and GIS in
classification of cassava plantation.

Cassava (Manihot esculenta Crantz) was introduced from Brazil, its country
of origin, to the tropical areas of Africa, the Far East and the Caribbean Islands by the
Portuguese during the 16th and 17th centuries. In the Gold Coast (now Ghana), the
Portuguese grew the crop around their trading ports, forts and castles and it was a
principal food eaten by both Portuguese and slaves. By the second half of the 18th
century, cassava had become the most widely grown and used crop of the people of
the coastal plains. The Akan name for cassava 'Bankye' could most probably be a
contraction of 'Aban Kye' - Gift from the Castle (Korang-Amoakoh et al., 1987).
Cassava is called “yuca” in Spanish, “mandioca” in Portuguese, “cassava” in
Haitian Creole, and “Manioc” in French. It is consumed in a variety of ways, but only
after some form of processing. It is consumed in a variety of ways, but only after
some form of processing. Cultivars are classified into two groups based on the
amounts of hydrogen cyanide present. Sweet types contain less than 50 mg kg-1HCN
(fresh weight) and are generally sold as fresh roots, whereas bitter types have higher
amounts of HCN along with higher yields and starch content (Conceição, 1981).
Cassava (Manihot esculenta) is the fourth most important source of food
energy in the tropics. More than two-thirds of the total production of this crop is used
as food for humans, with lesser amounts being used for animal feed and industrial
purposes. The ingestion of high levels of cassava has been associated with chronic
cyanide toxicity in parts of Africa, but this appears to be related to inadequate
processing of the root and poor overall nutrition. Although cassava is not a complete
food it is important as a cheap source of calories. (Cock, 1982)
Cassava (manioc, yuca, or mandioca; Manihot esculenta Crantz,
Euphorbiaceae) is an important cash crop and food crop of resource-limited farmers in
Africa, Asia, and Latin America and the Caribbean. The storage roots are utilized
either fresh, as in the case of sweet cultivars low in cyanogen glycosides, or after
processing into dry products such as flour, starch, and animal feed in the case of bitter
cultivars high in cyanogen glycosides (Emmanuel Okogbenin, Tim L. Setter, Morag
Ferguson, Rose Mutegi, Hernan Ceballos, Bunmi Olasanmi, and Martin Fregene,

Cassava production in Cambodia

Cambodian cassava is mainly grown in the central and southeastern part of
the country. Especially in Kampong Cham and Kampong Thom province, while some
is also grown along the Mekong River and, Siem Reap, Kampong Speu, Kampong
Thom, Battambang, and Banteay Meanchey. There are two main local varieties in
Cambodia, one is sweet, the other is bitter. Mong Reththy Tapioca (MRT) plantation,
located in Sihanouk ville, in southwest Cambodia, introduced RAYONG 60 and
KASETSART 50 in 2000. In areas near the border, the farmers introduced some
Vietnamese varieties (bitter) in Kampong Cham province, and a Thai company
introduced some Thai varieties (bitter) in Battambang province Banteay Meanchey
Province. Because of a lack of extension, and farmers in many provinces having
difficulty finding cassava markets for animal feed and industrial raw material, they
generally don‟t like planting the new bitter varieties; they just want to sell in the local
market sweet roots for fresh human consumption. The new varieties are not widely
grown yet (MAFF).

Cultivation Practices
Cassava is adaptable to diverse climates and can be grown in soil with low
fertility. It is planted either as a single crop or intercropped with maize, legumes,
vegetables, rubber or other plants. Cassava is normally planted during February–April
and harvested in eight to 12 months depending on market price and the availability of
labor for harvesting.
Cultivation practices in western and eastern Cambodia are similar, with a
few notable differences due to different soil and climate conditions. In Banteay
Meanchey province who growing cassava planted area is cover about 62151 hectares
(Department of Agriculture Banteay Meanchey 2014). The yield and production were
high especially near Thailand border. There are some Thai cassava varieties in
Cambodia such as Rayong 60 and Karsetsart 50 or other varieties from buyers‟ trader
Thai. The farmers in Banteay Meanchey near Thailand border introduced high
yielding varieties, cassava is mono-cropped and usually planted in December, and the
earliest planting is the middle of November and the last in February. The first
ploughing starts in early December before the forecast rain, followed by a second
ploughing and row making in the middle of January. Most farmers hire a local tractor
owner to plough and hire laborers to make rows for planting. Most have their land
ploughed twice, which results in a greater yield, while about 5 percent do it only once
to lack of financial resource. Planting seeds usually takes place in March. The
majority of farmers use their own cassava seeds from the previous harvest (Hing
vutha, Thun Vathana, 2009).The majority of farmer use their own cassava seeds from
the previous harvest. Herbicide is napery in Malai and needs to be applied at least
twice because weeds grow high and thick. The first application is the made in the
middle of May and the second a month and a half later. A third application of
herbicide might be made, depending on weed condition and farmers financial
resource. Finally, some branches are normally cut a month or so before harvesting to
admit enough sunlight for the root grow bigger. Cassava is mostly planted with other
crops, especially rubber etc. Farmers mostly use more a tractor instead labor for land
preparing in order not to disturb the other crops in western and north western areas.

Geographic Information System and Remote Sensing

Geographic Information System
Geographic Information System (GIS) is a powerful tool for collecting,
storing, retrieving, transforming and displaying spatial data from the real world
(Burrough., 1986). Many people offer definitions of GIS. In the range of definitions
presented below, different emphases are placed on various aspects of GIS. Some miss
the true power of GIS, its ability to integrate information and to help in making
decisions, but all include the essential features of spatial references and data analysis.
Geographic Information System (GIS) integrates hardware, software and data for
capturing, managing, analyzing, and displaying all forms of geographically referenced
data. GIS allows the viewing, understanding, questioning, interpreting, and
visualizing data in many ways that reveals relationships, patterns, and trends in the
form of maps, globes, reports, and charts. It answers questions and solve problems by
looking at data in a way that is quickly understood and easily shared. GIS technology
can be integrated into any enterprise information system framework. GIS can be used
to map the changes in an area to anticipate future conditions, decide on a course of
action, or to evaluate the results of an action or policy. By mapping where and how
things move over a period of time, one can gain insight into how they behave. Dana
Tomlin's definition, from Geographic Information Systems and Cartographic
Modeling (Englewood Cliffs, NJ: Prentice-Hall, 1990).This is a broad definition a
considerably narrower definition, however, is more often employed. In common
parlance, a geographic information system or GIS is a configuration of computer
hardware and software specifically designed for the acquisition, maintenance, and use
of cartographic data. From Jeffrey Star and John Estes, in Geographic Information
Systems. (Englewood Cliffs, NJ: Prentice-Hall, 1990) An Introduction. A geographic
information system (GIS) is an information system that is designed to work with data
referenced by spatial or geographic coordinates. In other words, a GIS is both a
database system with specific capabilities for spatially-reference data, as well as a set
of operations for working with data. In a sense, a GIS may be thought of as a higher-
order map (Understanding GIS: The ARC/INFO Method, 1990). A GIS is an
organized collection of computer hardware, software, geographic data, and personnel
designed to efficiently capture, store, update, manipulate, analyze, and display all
forms of geographically referenced information.

Components of Geographic Information System

A working Geographic Information System seamlessly integrates five key
components: hardware, software, data, people, and methods. (Ebook, 2014)




Software Hardware

Figure 2.1 Components of GIS (Mapsofindia, 2012)

1. Hardware
Hardware includes the computer on which a GIS operates, the monitor on
which results are displayed, and a printer for making hard copies of the results.
Today, GIS software runs on a wide range of hardware types, from centralized
computer servers to desktop. Computers used in stand-alone or networked
configurations. The data files used in GIS are relatively large, so the computer must
have a fast processing speed and a large hard drive capable of saving many files.
Because a GIS outputs visual results, a large, high resolution monitor and a high-
quality printer are recommended.

2. Software
GIS software provides the functions and tools needed to store, analyze,
and display geographic information. Key software components include tools for the
input and manipulation of geographic information, a database management system
(DBMS), tools that support geographic query, analysis, and visualization, and a
graphical user interface (GUI) for easy access to tools. The industry leader is
ARC/INFO, produced by Environmental Systems Research, Inc. The same company
produces a more accessible product, ArcView that is similar to ARCINFO in many
3. Data
Possibly the most important component of a GIS is the data. A GIS will
integrate spatial data with other data resources and can even use a database
management system, used by most organizations to organize and maintain their data,
to manage spatial data. There are three ways to obtain the data to be used in a GIS.
Geographic data and related tabular data can be collected in-house or produced by
digitizing images from photographs or published maps. Data can also be purchased
from commercial data provider. Finally, data can be obtained from the federal
government at no cost.
4. People
GIS users range from technical specialists who design and maintain the
system to those who use it to help them perform their everyday work. The basic
techniques of GIS are simple enough to master that even students in elementary
schools are learning to use GIS. Because the technology is used in so many ways,
experienced GIS users have a tremendous advantage in today‟s job market.
5. Methods
A successful GIS operates according to a well-designed plan and business
rules, which are the models and operating practices unique to each organization.

Spatial data
It should be noted that spatial data is at the heart of every GIS application.
Spatial data stores the geographic location of particular features, along with
information describing what these features represent. The location is usually specified
according to some geographic referencing system (e.g., latitude, longitude) or simply
by an address. Spatial data may define some physical characteristics, such as location
or position, or it may also define a property such as the area of a forest and
agricultures (which results from defining the various positions of its boundaries).
(Davis, 1996) In GIS, spatial data is classified as three main types: point, line, and
1. A point is a convenient visual symbol (an X, dot or other graphic), but
it does not reflect the real dimensions of the feature. Points may indicate specific
locations (such as a given address, or the occurrence of an event) or which are
usually too small to depict properly at the chosen scale features (such as a building)
2. A line is a one-dimensional feature with a starting and an ending point.
Lines represent linear features, either real (e.g., roads or streams)
3. A polygon is an enclosed area, a two-dimensional feature with at least
three sides (and therefore with an area). For example, it may represent a parcel of
land, agricultural fields, or a political district

Attribute data
Since the data collected and stored in the database determine the kind of
questions that can be asked of the data, it is necessary to understand the scales of
measurement in which data are recorded. The measurement scales normally used are
nominal, ordinal, interval and ratio.
Nominal Scale – The nominal scale is the lowest level of measurement
which is used to distinguish among features. Nominal data could be a name or a
description of features. For instance, a lake could be differentiated from a sand dune.
In a tropical area, there could be regions identified with sugar cane fields or rice
paddy fields. Basically, each name or description is distinct.
Ordinal Scale – Ordinal scales allows for data to be ranked in either an
ascending or descending order. A hierarchy of rank could be established depending
on the features under consideration. For example, a country could have cities ranked
as small, medium and large. In addition, the country may have parks that are ranked
as being minor, intermediate and major. Although the ordinal scale permits
differentiation on the basis of rank, it does not show or specify the magnitude of
Interval Scale – With the interval scale of measurement the distance
between the ranks is known. To employ an interval scale an arbitrary starting point is
used. The widely used example of the Celsius temperature scale explains the interval
scale. For example, it cannot be said that 38 degrees Celsius are twice as hot as 19
degrees Celsius, because 0 degrees Celsius is arbitrary.
Ratio Scale – A ratio scale is more advanced than the interval scale because there can
be an absolute starting point. For example, 78 miles is twice as far as 39 miles.
(Lakhan, 1996)

Electronic maps and images

An essential component of any GIS is some kind of map or image of an area.
This can be a digital map, satellite image or aerial photograph. Many GIS will use a
full range of map data and images as a base to link information against. The maps are
produced either as raster or vector data.
Raster maps
These are images made by a series of colored dots on a screen (pixels), just
like high quality digital photographs. Raster maps can be thought of as „unintelligent‟
as you can only obtain information that is visually represented on them. Raster maps
take up a lot of computer storage space but can be very useful as background maps to
other information.
Vector maps
Each vector map feature is recorded using XY coordinates. These may be a
single point (like a trig point), lines (such as roads) or polygons (such as buildings or
woods). The referenced features of vector maps make it possible for a GIS to link
information from spreadsheets and databases to the maps. Vector data is stored in
themed layers such as Roads, water, settlement.

Working GIS
A GIS stores information about the world as a collection of thematic layers
that can be linked together by geography. This simple but extremely powerful and
versatile concept has proven invaluable for solving many real-world problems from
modeling global atmospheric circulation, to predicting rural land use, and monitoring
changes in rainforest ecosystems.
Geographic information contains either an explicit geographic reference such as a
latitude and longitude or national grid coordinate, or an implicit reference such as an
address, postal code, census tract name, forest stand identifier, or road name. An
automated process called geocoding is used to create explicit geographic references
(multiple locations) from implicit references (descriptions such as addresses). These
geographic references can then be used to locate features, such as a business or forest
stand, and events, such as an earthquake, on the Earth's surface for analysis.

Figure 2.2 GIS Working

Remote sensing
Remote Sensing is a powerful tool that could be used to address the problem
of thematic maps which are out of date and have to be updated. The capabilities of
Remote Sensing to map and extract information about earth resources for various
applications are well documented. Among those prominently used is land cover
mapping, considered as one of the most important, most direct and well established
applications of remote sensing (Cambell, 1987).
Remote sensing provides spatial coverage by measurement of reflected and
emitted electromagnetic radiation, across a wide range of wavebands, from the earth‟s
surface and surrounding atmosphere. The improvement in technical tools of
meteorological observation, during the last twenty years, has created a favorable
substratum for research and monitoring in many applications of sciences of great
economic relevance, such as agriculture and forestry. Each waveband provides
different information about the atmosphere and land surface: surface temperature,
clouds, solar radiation, processes of photosynthesis and evaporation, which can affect
the reflected and emitted radiation, detected by satellites (Saha, 2003). The challenge
for research therefore is to develop new systems extracting this information from
remotely sensed data, giving to the final users, near-real-time information. Over the
last two decades, the development of space technology has led to a substantial
increase in satellite earth observation systems. Simultaneously, the Information and
Communication Technology (ICT) revolution has rendered increasingly effective the
processing of data for specific uses and their instantaneous distribution on the World
Wide Web (WWW). The meteorological community and associated environmental
disciplines such as climatology including global change, hydrology and oceanography
all over the world are now able to take advantage of a wealth of observational data,
products and services flowing from specially equipped and highly sophisticated
environmental observation satellites. An environmental observation satellite is an
artificial Earth satellite providing data on the Earth system and a Meteorological
satellite is a type of environmental satellite providing meteorological observations.
Several factors make environmental satellite data unique compared with data from
other sources, and it is worthy to note a few of the most important:
1. Because of its high vantage point and broad field of view, an
environmental satellite can provide a regular supply of data from those areas of the
globe yielding very few conventional observations.
2. The atmosphere is broadly scanned from satellite altitude and enables
large scale environmental features to be seen in a single view.
3. The ability of certain satellites to view a major portion of the atmosphere
continually from space makes them particularly well suited for the monitoring and
warning of short-lived meteorological phenomena; and
4. The advanced communication systems developed as an integral part of
the satellite technology permit the rapid transmission of data from the satellite, or
their relay from automatic stations on earth and in the atmosphere, to operational
users. These factors are incorporated in the design of meteorological satellites to
provide data, products and services through three major functions.
5. Remote sensing of spectral radiation which can be converted into
meteorological measurements such as cloud cover, cloud motion vectors, surface
temperature, vertical profiles of atmospheric temperature, humidity and atmospheric
constituents such as ozone, snow and ice cover, ozone and various radiation
6. Collection of data from in situ sensors on remote fixed or mobile
platforms located on the earth‟s surface or in the atmosphere; and
7. Direct broadcast to provide cloud-cover images and other meteorological
information to users through a user-operated direct readout station.
The first views of earth from space were not obtained from satellites but
from converted military rockets in the early 1950s. It was not until 1 April 1960 that
the first operational meteorological satellite, TIROS-I, was launched by the USA and
began to transmit basic, but very useful, cloud imagery. This satellite was such an
effective proof of concept that by 1966 the USA had launched a long line of
operational polar satellites and its first geostationary meteorological satellite. In 1969
the USSR launched the first of a series of polar satellites.

Huanjing-1A (HJ-1A)
HJ-1A (Huan Jing = Environment) satellites are small Chinese Earth
observation satellites. The main application fields for China are environmental
monitoring and prediction, solid waste monitoring, disaster monitoring and prediction
(flood, drought, typhoon and wind damage, sand storm, earthquake, land creep, frost
and grassland fires, coal fires, crop pest monitoring, ocean disaster monitoring)
(eoPortail Directory, 2010). The first two satellites, HJ-1A and HJ-1B, was
successfully launched in China on September 6, 2008. Both were manufactured by the
China Spaces at Company. And use the China Aerospace Science and Industry
Corporation (CAST) 968 satellite bus with a designed service life of three years. The
HJ-1A was equipped with an electro-optical imager with a 30-meter resolution and a
700-km swath, and a hyper spectral imager with a 100-meter resolution and a 50-km
swath. (Kevin Pollpeter, 2014).
HJ-1A the first micro-satellite constellation for Environment and Disaster Monitoring
of China. The same multispectral imagers named HJ-1/CCD with four bands (R, G, B,
Nir) and large swath are installed on both HJ-1A and HJ-1B. The HJ-1/CCD is the
main sensor of the constellation. The satellite constellation is composed of a number
of small satellites, the ground system, and the application system. It provides all-
weather (3 to 100 meter) imagery. The Huan Jing constellation consists of two small
optical satellites, the HJ-1A and the HJ-1B.
HJ-1A is also the Small Multi-Mission Satellite (SMMS) of Asia Pacific
Space Cooperation Organization (APSCO) and it will be the important data resources
of APSCO space application. SMMS is a joint venture payload between China, Iran,
South Korea, Mongolia, Pakistan, Thailand and Bangladesh under the Asian-Pacific
organization (Global Master Change Directory, 2012). China is establishing its
disaster and environment monitoring capability mainly depending on the ordinary
technologies, and meanwhile, China will also launch the small satellites to monitor
the earth environment, and apply the satellite remote sensing technology to conduct
all weather, around-the-clock and high time-resolution disaster and environment
monitoring. The small satellite constellation for disaster and environment monitor is
composed of 4 optical small satellites and 4 small satellites with synthetic aperture
radar. During the period of the Tenth Five-Year Plan, China will launch 2 optical
satellites and 1 radar satellite. The resolutions of CCD camera, the infrared camera
and the synthetic aperture radar are 30m, 150m, and 20m, respectively.
The average re-visit time is 32 hours China, Thailand and Iran are working
on a joint Small Multi Mission Spacecraft (SMMS) devoted to civilian remote-
sensing and communications experiments. The SMMS satellite will carry a low-
resolution charge-coupled device (CCD) camera and an experimental
telecommunications system. The SMMS will give Iran and Pakistan a semi-
autonomous space-imaging capability. The 470-kg (1,034-lb) spacecraft is set for
launch on a Chinese booster by 2004-05 into a 650-km. (400-mi.) Sun-synchronous
polar orbit (Zulu, 2011).
Table 2.1 Payload parameters (Rahman, 2012)

Side- Data
Spatial Swath Repetition
Band Spectral lookin transmis
Satellite Payload resolution width cycle
no. range (µm) g sion rate
(m) (km) (days)
ability (Mbps)
1 0.43~0.52 30

2 0.52~0.60 30
700 83,757 4
3 0.63~0.69 30
HJ-1A 120
4 0.76~0.9 30

- (110-128 100 50 ±30 4
ral Imager

Orbit characteristic of HJ-1A

Sun-synchronous circular orbit, altitude = 649 km, inclination = 97.95º,
LTDN (Local Time of Descending Node) equator crossing at 10:45 hours. HJ-1A are
in a coplanar orbit with a phasing of 180º. In the final stage of the constellation, 4
satellites will be distributed in the same orbital plane at phase angles of 90.Table 2
below describe some of Orbital characteristics of HJ-1A

Table 2.2 Orbit parameters (Rahman, 2012)

Satellite HJ-1A
Orbit Sun synchronous recurrent frozen orbit
Altitude 650 km
Inclination 97.95
Repetition cycle 31 days
Descending node (Local time) 10:30 AM
On-board capacity 16 Gbits

Landsat 8 (OLI)
Landsat 8 consist of two major segments the observatory and the ground
system. The observatory consists of the spacecraft bus and its payload of two earth
observing sensors, the operational land imager (OLI) and the thermal infrared sensor
(TIRS). OLI and TIRS collect Landsat 8 science data. The two sensors will
coincidently collect multispectral digital images of the global land surface including
coastal regions, polar ice, islands, and the continental areas. The spacecraft bus stores
the OLI and TIRS data on an onboard solid-state recorder and then transmits the data
to ground receiving stations, these two sensors provide seasonal coverage of the
global landmass at a spatial resolution of 30 meters (visible, NIR, SWIR), 100 meters
(thermal), and 15 meters (panchromatic). Landsat 8 (OLI) is a joint initiative of
NASA and the U.S. Geological Survey to maintain a robust archive of Landsat data
and imagery, which provides an uninterrupted multispectral record of the earth‟s land
surface and it let us analyze everything from terrain types to crop growth to natural
disasters all around the world. Table 3 below, describe some of the spectral band and
wavelengths within satellite image.
OLI and TIRS sensors mounted on Landsat Data Continuity Mission
spacecraft (LDCM). Operational Land Imager (OLI) and Thermal Infrared Sensor
(TIRS) images consist of nine spectral bands with a spatial resolution of 30 meters for
Bands 1 to 7 and 9. New band 1 (ultra-blue) is useful for coastal and aerosol studies.
New band 9 is useful for cirrus cloud detection. The resolution for Band 8
(panchromatic) is 15 meters. Thermal bands 10 and 11 are useful in providing more
accurate surface temperatures and are collected at 100 meters. Approximate scene
size is 170 km north-south by 183 km east west (106 miles by 114 miles).
Table 2.3 Landsat 8 spectral band/wavelengths (Source: Credit U.S. Geological
Survey Department of the Interior USGS)

Wavelength Resolution
(µm) (m)

Band 1 – Coastal aerosol 0.43 - 0.45 30

Band 2 – Blue 0.45 – 0.51 30

Landsat Data
Band 3 – Green 0.53 - 0.59 30
Band 4 – Red 0.64 - 0.67 30
Lunch Band 5 - Near Infrared (NIR) 0.85 - 0.88 30

February 11,2013 Band 6 – SWIR 1 1.57 - 1.65 30

Band 7 – SWIR 2 2.11 – 2. 29 30

Band 8 - Panchromatic 0.50 - 0.68 15

Band 9 – Cirrus 1.36 - 1.38 30

Band 10 – Thermal Infrared (TIRS)1 10.60 – 11.19 100

Band 11 - Thermal Infrared (TIRS) 2 11.50 -12.51 100

Band Combinations for Landsat 8 (OLI)

Landsat 8 (OLI) image look incredible now, while of the bands from
previous Landsat mission are still incorporated, there are a couple of new ones, such
as the coastal blue band water penetration/aerosol detection and the cirrus cloud band
for cloud masking and other application. Here‟s a rundown of some common band
combination applied to Landsat 8 (OLI), displayed as a red, green, blue (RGB).
Table 2.4 Composition Band (Source: Credit U.S. Geological Survey Department of
the Interior USGS)

Natural Color 432

False Color (Urban) 764
Color Infrared (Vegetation) 543
Agriculture 652
Atmospheric Penetration 765
Healthy Vegetation 562
Land and Water 564
Natural with Atmospheric Removal 753
Shortwave Infrared 754
Vegetation Analysis 654

Global Positioning System

The working of Global Positioning System (GPS) is a satellite navigation
system providing worldwide coverage. A group of 24 satellites, circling twice-daily
20,000 km above the earth's surface, transmit coded signals that are picked up by GPS
receivers. The constellation of navigation satellites around the earth enables position
to be determined anywhere at any time, and in any weather condition for free. By
recognizing the codes for each satellite, the receiver can determine the time taken for
the signal to be transmitted. The GPS uses this information to then calculate the
distance to each satellite. Once four or more satellites are located, the GPS
"triangulates" the distances to provide a location on the earth's surface, i.e. longitude,
latitude, and elevation. However, the signal is still prone to a number of errors that
can reduce the positional accuracy. These include atmospheric errors, multi-path
errors, satellite and receiver errors, and intentional errors.
Atmospheric errors are introduced as the signal passes through the
atmospheric layers.
Charged particles and moisture droplets delay the signal, leading to timing
inaccuracies. Atmospheric errors may range from 3 to 50 m, depending on the time of
day and the arrangement of satellites in the sky. A "dual-frequency" GPS minimizes
these errors through computer modeling or by comparing the relative speeds of two
different signals - but these receivers are costly. Multi-path errors occur when the
signal bounces off obstructions, such as buildings or sheds, before reaching the
receiver. Such errors may exceed 100 m in certain situations. Complex signal
rejection procedures - or simply using the GPS in wide-open spaces - should
minimize these errors. Satellite (or "ephemeris") errors result when the broadcast
orbit differs from the actual orbit. The US Department of Defense uses radar to
determine these errors, and any updated positional information can be added to the
satellite code to reduce this error. Receiver errors result largely from noise or the use
of inaccurate clocks inside the GPS unit - but can be minimized with more expensive
clocks.( Amod Ashok Salgaonkar, Trivesh Suresh Mayekar, Avinash Rambhau Rasal,
Kiran Rasal, Balkrishna Hotekar, Rakesh Jadhav, Amar Gaikwad) GPS is not a single
unit. It is a system and has a following three major components
1. Satellites: There are 24 satellites & 3 spare satellites. The exact location
of each of the satellites at any given moment is known. Very accurate clocks are
installed onboard these satellites. The satellites send radio signals continuously
towards earth. These signals contain several pieces of information such as satellite ID
number, time stamp, exact position of satellite etc.
2. Ground Control Stations: These are five control stations to monitor the
satellites. These stations unable the information on earth to be transmitted to the
satellites. Control stations track satellites & update the position of each satellites
continuously. These stations ensure accuracy of the system.
3. GPS receivers: GPS units are referred to as receivers. These units receive
radio signals from satellites, which contain important information such as time stamp,
satellite ID number, satellite position etc. The receiver knows exactly when the signal
leaves the satellite (time stamp) and when the signal arrives at the receiver. Hence, it
is possible to calculate the distance from satellites as distance → time × velocity of
light. The receiver also knows the exact position of satellite via the signal. The
receiver is therefore able to determine its exact distance from satellite.

Literature Review
Land suitability and crop substitution modeling for cassava (Tamkuan,
2013). Thailand is importing highly expansive energy sources such as petroleum
products. Moreover, cassava and sugarcane can be used as food and alternative energy
source. Therefore, this study suggests the use of biofuel crops to be utilized as source
of energy. Kampaeng-phet province was selected as research area in this research
because it has potential for many industries to support growing these crops and extent
the plantation. This study has three objectives. The first objective was land evaluation
for cassava considering 10 factors including soil texture, soil depth, soil drainage, soil
fertility, soil pH, surface water, irrigation, rainfall, temperature and slope. Then, the
weights of factors were investigated by analytic hierarchy process (AHP) and Fuzzy
AHP. It was found that Fuzzy AHP had approximately more accuracy than AHP to
evaluate the land suitability for cassava plantation. The second objective was to
classify agricultural area using HJ-1A satellite image. Two methods of classifications
in this study were pixel base classification (maximum likelihood) and object based
classification. For this objective, it was found that object based classification had
overall accuracy (76.27%) more than pixel based classification (64.55%).The last
objective was to make crop substitution model for extending cassava plantation
regarding land suitability, economics (revenue and profit). The different scenarios
showed many different options for planting biofeul crops. Crop substitution modeling
regarding land suitability had area to substitute to grow cassava (278.68
model regarding profit had area that can be used to substitute to cassava (1196.76
Some implications on agricultural land use affected by land quality in Sakon
Basin, Northeast Thailand. (Mongkolsawat, 2011). Crop requirements are normally
confined to certain land qualities which in turn reflect to land use pattern in the areas.
Exploring land qualities under a given land use was conducted with objective of
identifying the land quality limitations and its consequences on land use pattern. The
study area, Amphoe Wanon Niwat, is located in Sakon Nakhon basin and
significantly differs in land use pattern when comparing to the extensive areas in the
Northeast. We used the 1995 Landsat TM and the 2002 orthophotography to identify
the change of land use pattern of the areas. Evaluation of land suitability for cassava
was conducted asked on the integrated land qualities concerned by using GIS. With
the established GIS database, the overall insight into each land quality affecting the
crops could be determined. The spatial land qualities and their associated attributes
were used to analyze the causes and their consequences on land use patterns. Our
work demonstrates that an analysis of satellite data and aerial orthophoto can provide
detailed, spatially explicit identification of land qualities causing the consequent
agricultural land use pattern in the Sakon Nakhon Basin. The shallow lateritic soils,
improper land use and mis-management of land have significantly caused the current
land use patterns with relatively low agricultural productivity.
A study on estimation of cassava area and production using Remote Sensing
and Geographic Information Systems in the northeast region of Thailand. (Rajendra,
1999). A study on cassava plantation area and production was conducted in the
northeast region of Thailand using an integrated Satellite Remote Sensing (SRS) and
Geographic Information Systems (GIS). Although, SRS and GIS are considered as the
efficient tools for resource inventorying and monitoring, little work has been done in
Thailand with regards to the large area crop monitoring and production estimation.
The objective of the study was to explore the use of NOAAAVHRR data for mapping
cassava plantation areas. GIS was employed to create geographical database, such as
soils, topography, and land use and also for improving the results of image
classification. The study conducted for the two crop seasons of 1995 and 1996
indicated that the NOAA-AVHRR data can be used to map the cassava plantation
areas at the regional scale in Thailand. The results of the study were compared with
existing cassava statistics produced from the Thai Tapioca Development Institute
(TTDI) and the Office of Agricultural Economics (OAE), Thailand. The estimated
cassava plantation areas from the study were underestimated by - 9.7 and - 16.4
percent to that of TTDI and OAE, respectively for 1995 and overestimated by 4.0
percent but underestimated by - 14.4 percent, respectively for the year 1996.
Land suitability for cassava and assessing cassava cropping area with
satellite data and Geographic Information Systems (olsawat, 2008). Cassava a major
annual crop in Northeast Thailand requires minimal cultural attention and cash input,
it is tolerant to drought, efficient extractor of nutrients in infertile soil.
Over 50% of the cassava production in Thailand has been originated from the
Northeast. Land suitable for cassava should be depicted to support the increase of its
yield with effective land utilization. The study made with the aims of analyzing the
integration of land qualities to evaluate land based on the FAO guideline, with respect
to cultivation practices and minimal soil loss as well as economic viability. The study
area, Northeast Thailand, covers an estimated 170,000 km2 with over 1 million ha of
cassava cropped area and is characterized by gently undulating topography. The
overall evaluation of land for cassava in the Northeast was based on the integrated
requirements of crop, management, conservation and economic viability.
Evaluation of cassava planting potential with Remote Sensing and
Geographic Information System (GIS) (Zhang Chao, 2010). Along with the
development of the starch and grain alcohol industries, biomass energy production
has recently become important. Cassava is an important biomass energy plant. In this
paper, Geographic Information System (GIS) and remote sensing (RS) are used, along
with knowledge of the growing environment needed for cassava and the farmland and
ecology protection policy of China, to evaluate the cassava growing potential of
Winning County. The processing of spatial data is done first. Then, the evaluation
principles are defined according to the spatial data and the required growth
conditions. The evaluation data are obtained by spatial data analysis according with
the evaluation principles. Lastly, the cassava planting potential results are verified by
referencing these to cassava planting statistical data for 2005 provided by Nanning
City Government of Guangxi province.
Conformity of agricultural land use and physical stability in Khon Kaen,
Northeast Thailand. This study is to establish the database for land suitability for
cassava. (S.Sukchan, 2003) The evaluation of land in terms of its suitability for
cassava is based on the procedure as described in FAO guideline for land evaluation
for agriculture. The study area, Khon Kaen Province, covers an area of about
1,088,599 ha. The cassava requirement includes a number of land qualities which
affect the plant growth and yield. The land qualities, on which the suitability are based
consist of rainfall, irrigated area, soil texture and drainage, soil depth, base saturation
cation exchange capacity, available phosphorus, landform, slope and salinity of soil.
Each of land qualities in terms of spatial data were digitally encoded in Geographic
Information System databases to create thematic layers. With the selection criteria,
the overlay of those land qualities was digitally performed to produce resultant,
polygonal layer, each of which is a land unit. The land suitability model applied to the
resultant layer provides the suitability class. The result indicated that the most
extensive areas are marginally suitable and cover areas of about 37% of the province

This chapter gives an overview of data collection, data requirements and the
method applied for processing as well as the modeling approach, geographical dataset
creation and analysis techniques adopted. The findings this research in procedure
adopted work form the basis for driving statics in classification of plantation cassava
in Banteay Meanchey Province and subsequently in the overall.
The methodology workflow to achieve the objectives as shown this below

Figure 3.1 Classification of workflow

This study chapter will discuss data acquisition and the processing of data
collected in preparation for its inclusion in analysis for finding classification of
cassava plantation. The discussion begins with brief overview of satellite image.

Figure 3.2 SMMS (HJ-1A)

The figure above was representing the Satellite image of Chines Huan Jing-
1A (HJ-1A) on the 02 March 2015 by code LT51280512011018BKT00 path 07 and
row 105. The reference datum of this image was WGS 1984 and pat the map
projection was UTM zone 47N.

Figure 3.3 Enhanced operational land imager (OLI) and the thermal infrared sensor
(TIRS) image of 2015

The satellite image Landsat (OLI) and (TIRS) shown in figure upper was
selected four part on the 03rd ,07th February 2015 and 02nd ,18th March 2015 by
Landsat 8 code LC81270502015054LGN00, LC81270512015038LGN00 path 127
and row 50,51 and LC81280502015077LGN00, LC81280512015061LGN00 path
128 and row 50,51using the OLI and TIRS. The reference datum of this image was
WGS 1984 and pat the map projection was UTM zone 47N.

Geocorrection of Landsat Imagery

Geometric correction of satellite images involves modeling the relationship
between the image and ground coordinate systems. There are both systematic and
non-systematic geometric errors present in satellite imagery (Jenson, 1996). The
systematic errors in Landsat imagery are well documented, and are primarily
functions of scan skew, mirrors can velocity, panoramic distortion, platform velocity,
perspective and earth rotation (Mather, 1999; Jenson, 1996). Data on sensor
characteristics and ephemeris information are modeled and applied to the raw imagery
as part of the systematic correction performed by the Landsat receiving stations
(Masek et al., 2001). Assuming an accurate ephemeris based correction software
model is implemented, systematic errors are corrected in commercially available
Landsat imagery. Non-systematic errors are mainly caused by variation through time
in the position and attitude angles of the satellite platform (Jenson, 1996).
The stage in Geo-correcting an image is to select an appropriate map
reference system for the area of interest. Correction of all data sources to a single map
projection will allow accurate and easy integration, so choosing the correct system is
critical to success. To Geo Correcting your scanned field slip you will input four
control points with known x and y coordinates into the corners of the images. The
coordinate values used will be junctions of North -South and East-West oriented
divisions of the British National grid, which can be assumed to have a precise position
and value.
Geocorrection rectification it assigning coordinates to known locations
Ground Control Point and GCPs. The Data pixels must be related to ground locations
such as in UTM coordinates two main methods: Image to Image correction involves
matching the coordinate systems of one digital image to another image acting as a
map reference

Data exploration and preprocessing

Data exploration
The classification analysis is based on the assumption that the band data and
the training sample data follow normal distribution. To check the distribution of the
data in a band, use the interactive Histogram tool on the Spatial Analyst toolbar. To
check the distribution of individual training samples, use the Histograms tool on the
Training Sample Manager (ArcGIS Resource, 2013).
Stretching of band data
The classification process is sensitive to the range of values in each band. To
have the attributes of each band considered equally, the value range for each band
should be similar. If the value range of one band is too small (or too large) relative to
the other bands, it can use the mathematical tools in the Spatial Analyst toolbox to
stretch it. For example, we can use the Times math tool to multiply the band with a
constant value to stretch its value range.
Creating a multiband image
The Image Classification toolbar works with a multiband image layer. To
load individual bands to a new multiband image, use the Composite Bands tool.

Collecting training samples

In supervised classification, training samples are used to identify classes and
calculate their signatures. Training samples can be created interactively using the
training sample drawing tools on the Image Classification toolbar. Creating a training
sample is similar to drawing a graphic in Arc Map except training sample shapes are
managed with Training Sample Manager instead of in an Arc Map graphic layer.
To create a training sample, select one of the training sample drawing tools (for
example, the polygon tool) on the Image Classification toolbar and draw on the input
image layer. The number of pixels in each training sample should not be too small or
too large. If the training sample is too small, it may not provide enough information to
adequately create the class signature. If the training sample is too large, you might
include pixels that are not part of that class. If the number of bands in the image is n,
the optimal number of pixels for each training sample would be between 10n and

Figure 3.4 Collecting training sample


Evaluating training samples

When training samples are drawn in the display, new classes are
automatically created in the Training Sample Manager. The manager provides it with
three tools to evaluate the training samples the Histograms tool, the Scatterplots tool,
and the Statistics tool. You can use these tools to explore the spectral characteristics
of different areas. You can also use these tools to evaluate training samples to see if
there is enough separation between the classes

Figure 3.5 Evaluation training sample editing classes

Depending on the outcome of the training sample evaluation, it may need to

merge the classes that are overlapping each other into one class. This can be done
using the Merge tool in the manager window. In addition, it can rename or renumber a
class, change the display color, split a class, delete classes, save and load training
samples, and so forth. The following image shows how to merge two classes

Figure 3.6 Editing class


Creating the signature file

Once determine the training samples are representative of the desired classes
and are distinguishable from one another, a signature file can be created using the
Create Signature File tool in the manager window.

Figure 3.7 Creating the signature file

Examining the signature file

The dendrogram tool allows examining the attribute distances between
sequentially merged classes in a signature file. The output is an ASCII file with a tree
diagram showing the separation of the classes. From the dendrogram, it can determine
whether two or more classes or clusters are distinguishable enough; if not, it might
decide to merge them in the next step

Editing the signature file

The signature file should not be directly edited in a text editor. Instead, it
should use the Edit Signatures tool in the Multivariate toolset. This tool allows you to
merge, renumber, and delete class signatures.

Applying classification
To classify the image, the Maximum Likelihood Classification tool should
be used. This tool is based on the maximum likelihood probability theory. It assigns
each pixel to one of the different classes based on the means and variances of the class
signatures (stored in a signature file). The tool is also accessible from the Image
Classification toolbar.
The Interactive Supervised Classification tool is another way to classify
image. This tool accelerates the maximum likelihood classification process. It allows
to quickly preview the classification result without running the Maximum Likelihood
Classification tool.

Post-classification processing
The classified image created by the Maximum Likelihood Classification tool
may misclassify certain cells and create small invalid regions. To improve
classification, it may want to reclassify these misclassified cells to a class or cluster
that is immediately surrounding them. The most commonly used techniques to clean
up the classified image include filtering, smoothing class boundaries, and removing
small isolated regions. A more visually appealing map results from the data cleanup
Filtering the classified output
This process will remove single isolated pixels from the classified image. It
can be accomplished by either the Majority Filter tool or the Focal Statistics tool with
Majority as the statistics type. The difference of the two tools is that the Majority
Filter tool assumes a 3 x 3 square neighborhood during the processing, while the
Focal Statistics tool supports more neighborhood types
Smoothing class boundaries
The Boundary Clean tool clumps the classes and smooth the ragged edges of
the classes. The tool works by expanding and then shrinking the classes. It will
increase the spatial coherency of the classified image. Adjacent regions may become
Generalizing output by removing small isolated regions
After the filtering and smoothing process, the classified image should be
much cleaner than before. However, there may still be some isolated small regions on
the classified image. The generalizing process further cleans up the image by
removing such small regions from the image. This is a multi-step process which
involves several Spatial Analyst tools.
1. Run the Region Group tool with the classified image to assign unique
values to run each region on the image.
2. Open the attribute table of the new raster layer created by the Region
Group tool. Use the pixel counts to identify the threshold of small regions that you
want to remove.
3. Create a mask raster for the regions you want to remove. This can be done
by running the Set Null tool to set the regions with small numbers of pixels to a null
4. Run the Nibble tool on the classified image. Use the mask raster created
from the Set Null tool from the previous step as the Input mask raster. This will
dissolve the small regions on the output image.

Equipment of Analysis
In this section, there is some key equipment in participatory research and
analysis. Computer tool to assist in the research study, collection data writing thesis
and analysis with other programs and overall useful such as:

Central Processing Unit (CPU) cor i5 2.6GHZ
Read and Memory (RAM) 4GHZ
Printer: The use for print thesis
Camera: IPad Air
Global Position System (GPS)
Software Application
The ArcGIS was used to present the resulting map and image from the
analyses carried out on the research and used Microsoft office for facilitates sit thesis.
The table below was presented outline these software
Table 3.1 the using software

Software Relevance
ArcGIS was used for displaying and subsequent
1 ArcGIS 10 processing and enhancement of the Image as well as the
resulting maps
Microsoft Office: Excel was used in producing the bar chart, while
2 MS word & Excel Microsoft word was used generally for the presentation
of the research in text, chart and map formats

The object of this study form the basis of all the analysis carried out in this
chapter. The results are presented inform of maps, chart and statistical table and
present study application geospatial technology of compare with image satellite
Landsat 8 (OLI) and SMMS (Hj-A1) since 2015 supervised image classification of
land use was classified in nine classes of Banteay Meanchey province. Seven
categories of land us were identified these are: cassava, field crop, forest, water,
perennial tree/fruit tree, rice and urban using maximum likelihood.

Land use from SMMS (Hj-A1)

The Land use in Banteay Meanchey province of the year 2015 by the land
use classification of Landsat satellite images covering a total of about 6185.64 square
kilometers or about 618,564.44 hectares by the results of the study are as follows:
Cassava with an area of approximately 972.15 square kilometers or
97,215.33 hectares an about 15.72 percent of the total area.
Field crop with an area of approximately 742.74 square kilometers or
74,273.76 hectares or 12.01 percent of the total area.
Forest with an area of approximately 221.46 square kilometers or
approximately 22,146.02 hectares or 3.58 percent of the total area.
Water with an area or approximately 283.40 square kilometers or 28,340.40
hectares or 4.58 percent of the total area.
Perennial tree/Fruit tree with an area of approximately 1,075.64 square
kilometers or 107,564.40 hectares or 17.39 percent of the total area.
Rice with an area of approximately 2,531.59 square kilometers or
253,158.56 hectares or 40.93 percent of the total area.
Urban with an area of approximately 358.66 square kilometers or 35,865.98
hectares or 5.80 percent of the total area.
Table 4.1 Land use of Banteay Meanchey province from SMMS (Hj-A1) 2015

Land use Hectare Percentage

Cassava 97,215.33 972.15 15.72

Filed crop 74,273.76 742.74 12.01

Forest 22,146.02 221.46 3.58

Water 28,340.40 283.40 4.58

Perennial trees/
Fruit trees 107,564.40 1,075.64 17.39

Rice 253,158.56 2,531.59 40.93

Urban 35,865.98 358.66 5.80

Total 618,564.44 6,185.64 100.00

Filed crop
Rice 12.01%

Perennial Forest
Trees/Fruit Trees 3.58%

Figure 4.1 Percentage of land use Banteay Meanchey SMMS (HJ-A1)


Filed crop

1,500.00 Water

Perennial Trees/Fruit Trees


Figure 4.2 Area of land use Banteay Meanchey SMMS (HJ-A1)

Figure 4.3 Map SMMS (HJ-A1) of Banteay Meanchey showing location of cassava
plantation fields of studie

Figure 4.4 Map SMMS (HJ-A1) classification of Banteay Meanchey showing location
fields of studies

Land use from image satellite Landsat 8 (OLI)

The tables below in Land use image satellite Landsat 8 (OLI) in the 2015 by
the land use classification of Landsat satellite images covering a total of about
6,185.64 square kilometers or about 618,564.44 hectares by the results of the study
are as follows:
Cassava with an area of approximately 837.57 square kilometers or
83,757.37 hectares or about 13.54 percent of the total area.
Field crop with an area of approximately 653.87 square kilometers or
65,386.54 hectares or 10.57 percent of the total area.
Forest with an area of approximately 268.25 square kilometers or 26,825.14
hectares or 4.34 percent of the total area.
Water with an area of approximately 421.43 square kilometers or 42,143.43
hectares or 6.81 percent of the total area.
Perennial tree/Fruit tree with an area of approximately 1,170.80 square
kilometers or 117,079.71 hectares or 18.93 percent of the total area.
Rice with an area of approximately 2,508.45 square kilometers or
250,845.26 hectares or 40.55 percent of the total area.
Urban with an area of approximately 325.27 square kilometers or 32,527.00
hectares or 5.26 percent of the total area.

Table 4.2 Land use of Banteay Meanchey province from classified Landsat 8 (OLI)

Land use Hectare kilometers Percentage

Cassava 83,757.37 837.57 13.54

Filed crop 65,386.54 653.87 10.57

Forest 26,825.14 268.25 4.34

Water 42,143.43 421.43 6.81

Perennial trees/
Fruit trees 117,079.71 1,170.80 18.93

Rice 250,845.26 2,508.45 40.55

Urban 32,527.00 325.27 5.26

Total 618,564.44 6,185.64 100.00


Filed crop

40.55% Forest

Trees/Fruit Trees
18.93% Water

Figure 4.5 Percentage of land use Banteay Meanchey image satellite Landsat 8 (OLI)

Filed crop

2,000.00 Forest

1,500.00 Water

1,000.00 Perennial Trees/Fruit Trees


Figure 4.6 Area of land use Banteay Meanchey image satellite Landsat 8 (OLI)

Figure 4.7 The map of satellite Landsat 8 (OLI) at Banteay Meanchey showing
location of cassava plantation fields of studies

Figure 4.8 The map of image satellite Landsat 8 (OLI) of Banteay Meanchey province
showing location fields of studies

Land use comparing of SMMS (Hj-A1) and Landsat 8 (OLI)

Compare in land use Small Multi-Mission Satellite identify and analyze that
the Cassava area has increased 13,457.96 hectares or 2.18 percent, Field crop area has
increased 8,887.22 hectares or 1.44 percent, Forest area has decreased markedly
4,679.12 hectares or 0.76 percent, Water area has decreased markedly 13,803.03 or
2.23 percent, Perennial Trees/Fruit tree has decreased markedly 9,515.31 hectares or
1.54 percent, Rice area has increased 2,313.30 hectares or 0.37 percent, Urban area
has increased markedly 3,338.98 hectares or 0.54 percent.

Table 4.3 Comparison of land use Banteay Meanchey province image satellite
Landsat 8 (OLI) and SMMS (Hj-A1)

SMMS (Hj-A1) LANDSAT 8 (OLI) Difference

Square Square Square
Land use Hectare Hectare Hectare
kilometers kilometers kilometers
Cassava 97,215.33 15.72 83,757.37 13.54 13,457.96
74,273.76 12.01 65,386.54 10.57 8,887.22 1.44
Filed crop
Forest 22,146.02 3.58 26,825.14 4.34 -4,679.12 -0.76
Water 28,340.40 4.58 42,143.43 6.81 -13,803.03 -2.23
trees/Fruit 107,564.40 17.39 117,079.71 18.93 -9,515.31 -1.54
253,158.56 40.93 250,845.26 40.55 2,313.30 0.37
35,865.98 5.80 32,527.00 5.26 3,338.98 0.54

Total 618,564.44 100.00 618,564.44 100.00










Figure 4.9 Comparison of land use Banteay Meanchey province image satellite
Landsat 8 (OLI) and SMMS (Hj-A1)

Values accuracy of model

By adopting the use of the land use map from model year 2015 the models
check the accuracy of land-use maps for 2015 obtained from the classification of
satellite data. By randomly sample points the results of the review accuracy of models
land use in 2015 found that the model was the overall accuracy of 75.56 percent from
the land use classify SMMS (Hj-A1) satellite and the model was the overall accuracy
of 81.48 percent land use from satellite Landsat 8 (OLI).

Figure 4.10 The map shows the sample location to determine accuracy

Table 4.4 Dislocation evaluate the accuracy of the classification land use in SMMS
(Hj-A1) satellite

Reference data
Classifieds Perennial
Data Filed
Cassava Forest Water trees/ Rice Urban Total
Fruit trees
26 3 0 0 2 3 1 35
Filed crop 4 13 1 1 1 2 0 22
Forest 0 0 5 1 0 0 0 6
Water 0 0 0 6 0 1 1 8
Perennial 1 1 2 1 52 3 2 62
Rice 4 8 4 6 2 90 4 118
Urban 2 1 1 2 0 1 12 19
Total 37 26 13 17 57 100 20 270
Overall Accuracy for SMMS (Hj-A1) 204/270*100=75.56 %
Table 4.5 Dislocation evaluate the accuracy of the classification land use in image
satellite Landsat 8 (OLI)

Reference data
Filed Perennial
Data Cassava Forest Water trees/ Rice Urban Total
crop Fruit tree
Cassava 28 3 0 0 0 3 1 35
Filed crop 4 16 0 0 0 2 0 22
Forest 0 0 5 0 1 0 0 6
water 0 0 0 6 0 1 1 8
trees/Fruit trees 1 0 2 1 57 0 1 62
Rice 3 8 4 6 2 93 2 118
Urban 2 1 0 0 0 1 15 19
Total 38 28 11 13 60 100 20 270
Overall Accuracy for Landsat 8 (OLI) 220/270*100=81.48 %

The result from land use to study the comparing of land use in Banteay
Meanchey province, using satellite imagery Landsat 8 (OLI) and SMMS (Hj-A1)
satellite classification for land use. Land use divided into seven categories of cassava
area, field crop area, forest area, water area, perennial tree/fruit tree area, rice area and
urban. The classification of satellite SMMS (Hj-A1) in 2015 with an area of seven
types of 97,215.33, 74,273.76, 22,146.02, 28,340.40, 107,564.40, 253,158.56 and
35,865.98 hectare respectively.
The Rice area has much space is available plantation that accounting for
40.93 percent of the total area, followed by cassava area, field crop area, forest area,
water area, perennial tree/fruit tree area, rice area and urban area and aquaculture
accounting for 40.93 percent, 15.72, 12.01, 3.58, 4.58, 17.39, 40.93 and 5.80.
The result of the classification of satelliteimagery Landsat 8 year 2015, with
an area of seven types of 83,757.37, 65,386.54, 26,825.14, 42,143.43, 117,079.71,
250,845.26 and 32,527.00 hectare of respectively. The rice area also has the most
accounting for 40.55 percent of the total area, but compared to the MSSM (Hj-A1)
satellite found space down 2,313.30 hectare representing 0.37 percent of the total area
followed by field crop area, urban area and cassava area and aquaculture accounting
for 1.44, 0.54 and 2.18 percent. And another area has increase such as forest area,
water area and perennial tree/ fruit tree area and aquaculture accounting for 4,679.12,
13,803.03 and 9,515.31 hectare respectively, by 0.76, 2.23 and 1.54 percent
When compared SMMS (Hj-A1) with image satellite Landsat 8 found that
cassava area, field crop area, rice area include urban area has increase and area
decreasing such as forest area, water area and perennial tree/fruit tree area.
Assessment of the accuracy of models of land use is was the overall accuracy of 75.56
percent from the land use classifies MSSM (Hj-A1) and the model was the overall
accuracy of 81.48 percent land use from satellite Landsat 8 (OLI) which shows that
the accuracy is the medium and high between the data obtained from monitoring

changes and data for location details of the trial and evaluated the accuracy of each
point in the survey concluded.

Classification plantations of cassava using satellite imagery Landsat 8 (OLI)
Case Study Banteay Meanchey Province, Cambodia. Using the classification of
controlled (Supervised Classification) with the highest probability (Maximum
Likelihood Classification) the Banteay Meanchey Province here are plantations of
cassava approximately 837.57 square kilometers or 83,757 hectare accounted for
15.72 percent of the total study area. Compared with the results from the classification
using satellite imagery SMMS (HJ-A1) a different percentage of 2.18 was found the
area under cultivation of cassava from satellite SMMS (Hj-A1) has approximately
975.15 square kilometers or 97,215.33 hectares representing 15.72 percent. The result
of the inspection accuracy overall, the classification of satellite images Landsat 8
(OLI) and SMMS (HJ-A1) with an accuracy of approximately 75.56 and 81.48
percent respectively, with accuracy levels moderate and high. The accuracy of which
are similar to the findings of Yamamoto and Suckchan studying the appropriate
classification of land (Land suitability) for rice, sugarcane and cassava by using
factors include area classification. Results from monitoring the overall accuracy of the
classification (Overall accuracy assessment) found that the overall accuracy of the
classification with 81.48% accuracy by reason of satellite Landsat 8 (OLI) is due to be
over SMMS (Hj-A1) satellite Landsat 8 (OLI) has wavelength available upon request.
And the numerical value, which amounted to 16 bits or 65,536 more detailed satellite
images SMMS (Hj-A1) is the numerical value of 8-bit or 256.
This study revealed the capacity of cassava plantation area it will be useful
information to forecast the trend cassava over production. The results obtained from
this study serve as a good example of the application of remote sensing data and
geoinformation system for agriculture. This application can be seen as a regional pilot
project to be adopted by many other provinces in Cambodia for various kinds of
agricultural cultivation such as rice, sugarcane, maze, etc.


1. This study used satellite data Landsat 8 and SMMS satellite recorded
during different satellite the results of the data analysis area deviation from reality.
Therefore, it should be taken at the same time satellite images were analyzed to
provide information within creased accuracy.
2. The classification data with satellite images of cassava plantation wanted
ability and experience of a translator. It’s don’t have affect the accuracy of the maps
of land uses should check the accuracy of land use to obtain current information and
increased accuracy.
3. Distribution of numerical value Digital Number: DN of cassava, rice and
coin are similar for accuracy more than must be use high resolution satellite Landsat
analysis exploration another factor that relative with crop planted such as Drainage
of the soil, Depth of soil character of land index value NDVI
4. Continuing the trend of land use in the future is used as a guide in
developing spatial databases and the information used to plan development of the
area, including the management of natural resources, area agriculture and
environment preliminary.
5. The technology of remote sensing and GIS should be employed in major
studies, concerning nation issue such as agriculture and environment etc.
Cassava plantation at Banteay Meanchey Province

Picture 1 Cassava plantation at Tmor Puok district (Day One)

Picture 2 Cassava plantation at Tmor Puok district (Day Tree)

Picture 3 Cassava plantation at Tmor Puok district (Day Five)

Picture 4 Cassava plantation at Tmor Puok district (Weeks Two)

Picture 5 Cassava plantation at Pnom Srok district (Weeks Two)

Picture 6 Cassava plantation at Svay Chekk district (Weeks Two)

Picture 7 Cassava plantation at Malai district (Weeks Tree)

Picture 8 Cassava plantation at Tmor Pouk district (Weeks Tree)

Picture 9 Cassava plantation at Svay Chekk district (Months Two)

Picture 10 Cassava plantation at Malai district (Months Tree)

Picture 11 Cassava plantation at Svay Chekk district (Months Two)

Picture 12 Cassava plantation at Tmor Pouk district (Months Three)

Picture 13 Cassava plantation at Pnom Srok district (Month three)

Picture 14 Cassava plantation at Serey Sorpoun city (Months Six)


Name Mr. Sopheak Pen

Date of Birth May 30. 1982

Place of Birth Commune #4 , Khan Toul Kork

Phnome Penh Municipality, Cambodia
Present Address Borey Chamkar Dong , # 38D, Chamkar Dong
Phnom Penh Municipality, Cambodia
Position hell
2010-present Faculty of Science, Mean Chey
University of Computer Science, National Road
No 5, Banoy Village, Sangkat Teuk Thlar,
Serey Sophorn City, Banteay Meanchey
Province, Cambodia
Education Bachelor Science of Computer, Royal
2001-2005 University of Phnom Penh

2009-2010 Pedagogy of National Institute Education,

Phnom Penh Municipality, Cambodia

2013-2015 Master of Geoinformatic Faculty of

Geoinformatics, Burapha University

Chonburi, Thailand

Email address


HING Vutha, THUN Vathana. (2009). The Case of Cassava and Rubber in Cambodia.
Phnom Penh: CDRI.
Kevin Pollpeter, Eric Anderson Jordan, Wilson, Fan Yang. (2014). A report prepared for
the U.S.-China Economice and Security Review Cmention. IGCC.
Leisz, S. J. (2005). Developing a methodology for identifying, mapping and potentially
monitoring the sitribution of general farming system types in vietnam's Northen
Mountain Region. Agricultural Systems(340-363), 85(3) Spec.
Lillesand, T.M., Kiefer, R.W. and Chipman, J.W.,. (2008). Remote sensing and image
and image interpretation . John Wiley & Sons, Inc., New York.: Sixth edition.
mapsofindia. (2012, Sectember 28). Retrieved from
Meanchey, D. o. (2014).
(n.d.). Nations, Food and Agriculture Organization of the United.
Rahman, H. (2012, September 20). Strengthening of Satellite Based Crop Monitoring &
Estimation System for Food Security Application in Bangladesh. Retrieved from
Rubatsky, V.E and M. Yamaguchi. (1997). World Vegetables: Principles, Production,
and Nutritive Values, 2nd ed. New York, NY: Chapman & Hall.
Saha, M.V.K. Sivakumar P.S. Roy K. Harmsen S.K. (2003). In Satellite Remote Sensing
and GIS Applications in Agricultural Meteorology. Dehra Dun, India: World
Meteorological Organisation 7bis, Avenue de la Paix 1211 Geneva 2,
The Cambodian Ministry of Agriculture, F. a. (n.d.).
Tourism, C. (2015). Retrieved from
Tourism, C. o. (2015). webmaster at tourismcambodia. Retrieved from
Understanding GIS: The ARC/INFO Method. (1990). Redlands, CA: Environmental
System Research Institute.

USGS. (2009a, November 12). Retrieved from Landsat Thematic Mapper Data (TM):
USGS. (2010a, December Collected 12). Thematic Mapper. Retrieved from
USGS. (2010a, november 12). Thematic Mapper. Retrieved from
USGS. (2010b, November 12). Using Landsat data. Retrieved from from:
USGS. (2010b, December 6). Using Landsat data. Retrieved from
USGS. (2010b, December 6). Using Landsat data‘. Retrieved from
USGS. (201b, December 6). Using Landsat data. Retrieved from
USGS, C. U. (n.d.).
Yem, D. (2010). Key factors affecting the expanding irrigation in Cambodia.
Zulu. (2011, July 21). Retrieved from

Dewan, A. M., & Y. Yamaguchi (2009). Land use and land cover change in greater
Dhaka, Bangladesh: Using remote sensing to promote sustainable urbanization.
Applied Geography, 29(3):390-401
Berberoglu, S., & A. Akin (2009). Assessing different remote sensing techniques to
detect land use/cover changes in the Eastern Mediterranean. International
Journal of Applied Earth Observation and Geoinformation, 11(1), 46-53.
Goetzke, R., M. Braun, H. Thamm and G. Menz (2008). Monitoring and modeling urban

land-use change with multitemporal satellite data.. (2008) International Geoscience and
Remote Sensing Symposium (IGARSS), 4 (1), 4779770
Leisz, S. J., N.T.T. Ha, N.T.B. Yen, N.T. Lamand T.D. Vien (2005). Developing a
methodology for identifying, mapping and potentially monitoring the
distribution of general farming system types in Vietnam's Northern Mountain
Region. Agricultural Systems, 85(3) Spec. Issue):340-363.

Dong, L., W. Wang, M. Ma, J. Kong & F. Veroustraete (2009).The change of land cover
and land use and its impact factors in upriver key regions of the Yellow River.
International Journal of Remote Sensing, 30(5):1251-1265