Remote sensing and gis application for forest change detection in kim tien commune kim boi district hoa binh province

In this study, Landsat 5 TM satellite images in 1995, 2002, 2011 and Landsat 8 OLI/TIRS in 2018 of Kim Tien, Kim Boi were used to classify and identify areas of forest change, degraded f

MINISTRY OF AGRICULTURE AND RURAL DEVELOPMENT

VIETNAM NATIONAL UNIVERSITY OF FORESTRY

Faculty: Forest Resources and Environmental Management

Student: Nguyen Thi Hue Student ID: 1453091261

Class: K59B Natural Resources Management Course: 2014 – 2018

Advanced Education Program Developed in collaboration with Colorado State University, USA

Supervisor: Assoc Prof Tran Quang Bao

Hanoi, 2018

TABLE OF CONTENTS

ACKNOWLEDGEMENT 3

ABBREVIATIONS 4

LIST OF TABLES 5

LIST OF FIGURES 6

ABSTRACT 1

Chapter I 1

INTRODUCTION 1

Chapter II 3

LITERATURE REVIEW 3

2.1 General information 3

2.2 Development of remote sensing and GIS 4

2.3 Features of Landsat images 5

2.4 Remote sensing and GIS application 8

2.4.1 Agriculture 8

2.4.2 Forestry 9

2.4.3 Urban planning 10

2.4.4 Land cover mapping 10

Chapter III 11

OBJECTIVES AND METHODOLOGY 11

3.1 Objectives 11

3.2 Methodology 11

3.2.1 Study site 11

3.2 Methodology 12

3.2.1 Data sources 12

3.2.2 Interviewed data 13

3.2.3 Data processing 14

3.2.4 Image classification using NDVI 16

3.2.5 Field survey 17

Chapter IV 20

NATURAL AND SOCIAL-ECONOMIC CONDITION 20

4.1 Natural condition 20

4.2 Social-economic condition 23

5.1.2 NDVI of study area 30

5.1.3 Forest Cover in Period 1995 - 2018 31

5.1.4 The accuracy of classified forest map 33

5.2 Forest cover change detection in study site from 1995 to 2018 38

5.2.1 Forest change detection from 1995 to 2002 38

5.2.2 Forest change detection from 2002 to 2011 40

5.2.3 Forest change detection from 2011 to 2018 42

5.2.4 Forest change detection from 1995 to 2018 44

5.3 Driving forces of forest cover change 46

5.3.1 Forest increase 46

5.3.2 Forest decrease 47

5.4 Solutions for forest protection and management 49

5.4.1 Strengthen forest management and protection 49

5.4.2 Support local people on forest plantation and improving their livelihoods 49

5.4.3 Raise awareness of local people in forest protection 50

5.4.4 Increase investment on forest protection and afforestation projects 50

5.4.5 Enhance technological application in forest protection and management 51

Chapter VI 52

CONCLUSION, LIMITATION, FURTHER STUDY 52

6.1 Conclusion 52

6.2 Limitation and further study 53

APPENDIX 1 1

APPENDIX 2 4

REFERENCES 1

ACKNOWLEDGEMENT

This thesis would not have been possible without the support and help from several

people I would like to express our special appreciation of following people who supported

me with my sincere gratitude:

I would like to express my sincere thanks and appreciation to my supervisor Dr Tran

Quang Bao for his encouragement untiring and excellent guidance, valuable suggestions in

my thesis His comments and advices have helped me to finish my thesis

I would like to thank the teachers in Vietnam National University of Forestry who

have imparted us the knowledge to perform this topic

I am also thankful to Prof Lee MacDonald for his enthusiasm in guiding me to

construct thesis proposal

I would like to express my gratitude to the Head of commune for support and giving

me chance to study in commune Besides, I also thank the local authorities, people for

providing me valuable time and information for my study

Last but not least, I want to give my gratitude to my parents who always encourage

and support me

Hanoi, 5 October 2018

Hue Nguyen Thi Hue

ERTS Earth Resources Technology Satellite

ESRI Environmental Systems Research Institute

GIMMS Geographic Information Mapping Manipulation System

NASA National Aeronautics and Space Administration

NDVI Normalized Difference Vegetation Index

SPOT Satellite Pour l’Observation de la Terre

LIST OF TABLES

Table 5.1: The categories classification framework 35

Table 5.2: The accuracy of classification method in 1995 35

Table 5.3: The accuracy of classification method in 2002 36

Table 5.4: The accuracy of classification method in 2011 36

Table 5.5: The accuracy of classification method in 2018 37

Table 5.6: Forest cover change in the period 1995-2002 38

Table 5.7: Forest cover change in the period 2002-2011 40

Table 5.8: Area of forest cover change in the period 2011-2018 42

Table 5.9: Forest cover change in Kim Tien in the period 1995-2018 44

LIST OF FIGURES

Figure 3.1: Map of study site 11

Figure 3.2: Landsat image of study site in 1995, 2002, 2011, 2018 13

Figure 3.3: Landsat processing and classification 15

Figure 5.1: NDVI classification of Kim Tien commune 31

Figure 5.2: Forest distribution in Kim Tien in 1995, 2002, 2011, 2018 32

Figure 5.3: Chart of area of land use types in Kim Tien over period 1995-2018 33

Figure 5.4: Map of ground true points distribution in 2018 34

Figure 5.5: Map of forest cover change in Kim Tien in the period 1995-2002 39

Figure 5.6: Chart of forest cover change in Kim Tien in the period 1995-2002 39

Figure 5.7: Map of forest cover change in Kim Tien in the period 2002-2011 41

Figure 5.8: Chart of forest cover change in the period 2002-2011 41

Figure 5.9: Forest cover change in Kim Tien in the period 2011-2018 43

Figure 5.10: Chart of forest cover change in Kim Tien in the period 2011-2018 43

Figure 5.11: Map of forest cover change in Kim Tien in the period 1995-2018 45

Figure 5.12: Chart of forest cover change in the period 1995-2018 45

ABSTRACT

Remote sensing technology and GIS are considered as an effective and objective tool

in monitoring and evaluating environmental resources, especially in the determination of forest area fluctuations In this study, Landsat 5 TM satellite images in 1995, 2002, 2011 and Landsat 8 OLI/TIRS in 2018 of Kim Tien, Kim Boi were used to classify and identify areas

of forest change, degraded forest and rehabilitated forest NDVI (Normalized difference vegetation index) method was chosen to classify the information from satellite image The accuracy of image classification method given 84.35% in 2018, 81.75% in 2011, 80% in

2002 and 75.65% in 1995 This result shows that, image classification using NDVI method combined with field survey had provided high accuracy to construct forest change in case of lacking of data to examine historical images The study has constructed a database on forest land and forest status maps for 1995, 2002, 2011, 2018; maps of forest change during periods

1995 – 2002, 2002 – 2011, 2011 – 2018 in Kim Tien commune The result show that, from

1995 to 2018, an area of forest land had gone down about 208.17ha

The study also shows that the difficulty of forest protection management in the commune by poor people’s lives, people are not fully aware of the responsibility for forest management and forest protection, the weakness in forest management network of the commune On this result, the study has recommended some solutions for better forest management, include: Strengthening forest management and protection, raising awareness of local people in forest protection and enhancing the applications of technology

Chapter I

INTRODUCTION

Forest plays a significant role in environmental protection Forests are important renewable natural resources and have an important role in preserving an environment suitable for human life It participates in air regulation, absorbs carbon dioxide and releases oxygen, provides fresh air Forest restricts flood, drought, prevent erosion and landslide It also provides shelter, food source of many wildlife animals In Vietnam, forest represents the characteristics of tropical rainforest In the past, forest area was declined seriously According

to the forest inventory and planning institute, from 1979 to 1990 natural forest declined 2.7 million hectares account for 1.7%/year In period 1999- 2005, the area of rich natural forest decreased by 10.2% and the medium forest decreased by 13.4% (Hanoi Association for forest plantation and ecological protection, 2010) The main reasons of that status are, illegal logging, conversion of forest to agricultural land In recent years because of forest plantation and protection policies of the government, forest cover has increased significantly Forest cover in 2016 is 41.19% (MARD, 2016)

Nowadays, the development of technology of earth observation satellite, remote sensing imagery and geographic information systems have proposed applications in many field of science and management In natural resources and environmental management, they support human in construction, management and storage database Natural resource management involved in forestry sector Currently many state, and private forestry agencies, governments are implemented Geographic Information Systems (GIS) and remote sensing (RS) for various applications Geographic data with spatial dimensions is useful for understanding ground conditions, also useful for strategic decisions Remote sensing techniques used to investigate changes in land use land cover in the field of forestry Remote sensing technology is very useful for analyzing change detection and mapping of the land

cover of the forest, culture data extraction Satellite data has multi-time, multi-resolution, covers huge area quality to help us collect information quickly without directly contact The using of high resolution remote sensing images in resource management has been a new direction for planning natural resource Landsat program consists of a series of remote sensing satellites developed by NASA which provide since 1972 for land monitoring Since

1972, there has been eight Landsats were launched: Landsat-1, Landsat-2, Landsat-3, Landsat-4, Landsat-5, Landsat-6, Landsat-7 and Landsat-8 Landsat acquisition over million images of all place in the world, and millions of images were downloaded Given the huge availability of Landsat data, land cover monitoring is affordable and reliable

GIS and remote sensing provide scientific and reliable data that traditional mapping method can not do With the requirement of updating quickly, high accurate information about change of forest cover, the application of GIS and remote sensing is necessary for today and the future Besides that Kim Tien is a part of buffer zone of Thuong Tien Natural Reserve that used to be a problem area about uncontrolled harvest forest, and now there is no

research about forest land in the area Because of this reasons I have this study “Remote sensing and GIS application for forest change detection in Kim Tien commune, Kim Boi district, Hoa Binh province from 1995 to 2018”

Chapter II LITERATURE REVIEW 2.1 General information

Remote sensing is the science and art of obtaining information about an object, area

or phenomena through the analysis of data acquired by a device that is not in directly contact with the object, area or phenomena This process consists of making observations using sensors mounted on platform (satellites, airborne,…) The nature of remote sensing is technology that identifies and recognizes objects or environmental conditions through their own reflective or electromagnetic characteristics However, magnetic field and gravity field are also used Remote sensing uses aerial images or satellite images to collect information There are three common remote sensing methods are by airplane, satellite and drone These methods can provide detail information about Earth’s surface, natural resources as well as environmental condition

Geographic information system (GIS) is a computer application designed to perform a wide range of operations on geographic information Geographic information is defined as information about locations on or near the surface of the Earth A geographic information system (GIS) is a framework for gathering, managing, and analyzing data GIS integrates many types of data It analyzes spatial location and organizes layers of spatial information into visualizations using maps and 3D scenes, it describes the spatial relationship between them GIS discovers deeper insights into data, such as patterns, relationships, and situations, helping analyze and display spatial information from real world to solve the problem of collecting information and support users in making decisions to better manage natural resources

Remote sensing data is the source of database for GIS on the basis of various information classes Thus, the combination of remote sensing and GIS has become an

effective integrated technology for collecting, updating and analyzing spatial data that serve various fields

2.2 Development of remote sensing and GIS

Remote sensing began in the 1840s when balloons were used to take pictures of the ground In 1909, the first aerial photograph was taken from an aero plane Aerial photography became a reconnaissance tool in the First World War for two purposes: spying and mapping

In the Second World War aerial photos were used for military purposes: mapping of strategic location, military targets, assessing damage In 1957 Sputnik was invented, the putting cameras on orbiting spacecraft was reliable Two American satellites: Explorer I and II were launched in 1958 and 1959 Remote sensing developed from 1970 when the first satellite dedicated to monitoring land and ocean to map natural resources In 1972 ERTS-I (Earth Resources Technology Satellite) was launched in 1972 Then it was followed by ERTS-2 in

1975 Then their name were changed to LANDSAT-1, 2 The European Radar satellite 1) was launched in 1991 USA, France and India have planned a series of satellites with improvement in capability, so the users can get better resolution of aerial images

(ERS-GIS has been developed from mid 20th century Roger Tomlinson-the father of (ERS-GIS,

he worked to initiate the development of the Canadian Geographic System (CGIS) as the root

of Geographic Information System This system was invented to collect, store, and analyze data about land usage in Canada In1964, SYMAP-the first compute mapping software was created by Howard Fisher Then in 1965 he constructed the Harvard Laboratory for Computer Graphics, developed ODYSSEY GIS, GIMMS, GRID Many of the early concepts for GIS and its applications were created at the Lab In the late 1980s, ESRI was found as one of GIS software vendors, now it is the largest GIS software in the world In the 1981, ESRI released

helps people to create their own digital map to handle problems It also becomes a based tool for collecting, storing and manipulating map-based land data

computer-2.3 Features of Landsat images

Landsat is the general name for satellite system used for purpose of archive Earth images It is a satellite system that have become an international nature The first Landsat satellite was launched in 1972 by NASA and continuing with Landsat 7 Landsat program was designed to collect data monitoring multi-spectral information from Earth’s surface helps scientist to assess the change in Earth’s surface

There are four sensors used in the Landsat program The Multi Spectral Scanner (MSS) was carried in Landsat 1-3 Landsat 2 was operated in 1975, Landsat 3 in 1978 There three satellites image had four spectral bands covering green, red and near infrared band with

60 meters spatial resolution Landsat 4 was launched in 1982 and Landsat 5 in 1984, both satellites carried a new sensors was known as Thematic Mapper (TM) Thematic Mapper is

an advanced, designed to achieve higher image resolution This sensor observe the Earth under six spectral bands and one thermal band (band 6) with 30 meters spatial resolution Landsat 6 was launched in 1993 but it was lost just after not reaching the velocity to obtain orbit After that Landsat 7 was launched in 1999 and is equipped with sensor Enhanced Thematic Mapper Plus (ETM+) The ETM+ sensor has the same spectral bands as the TM sensor, with an addition of a 15 meters panchromatic band Landsat 8 was launched to orbit

in 2013 and has two main sensors: the Operational Land Imager (OLI) and the Thermal Infrared Sensor (TIRS) Images of Landsat 8 include nine spectral bands with 30 meters spatial resolution for bands 1-7 and 9 Band 8 is panchromatic with 15 meters resolution Bands 10 and 11 are collected at 100 meters

Table 2.1: Sensor feature of Landsat 4-5

Source: Paul R.Baulmann (2010)

Table 2.2: Sensor features of Landsat 7

Source: Paul R.Baulmann (2010)

Table 2.3: Sensor features of Landsat 8

Band 7-Short wavelength infrared 2 2.107-2.294 30

Source: Department of the Interior U.S Geological Survey (2016)

Landsat image is applied in many fields from studying the status to monitoring fluctuation of land use/land cover This following table presents the major applications of Landsat images:

Table 2.4: The major application of Landsat image

Landsat

band

Blue 0.45pm – 0.52pm Used to investigate shoreline, distinguish

vegetation and soil, identify other objects

Green 0.52pm – 0.60pm Used to measure the maximum reflectance

spectrum of plant, identify vegetation state and other objects

Red 0.63pm – 0.69pm Used to determine chlorophyll absorption zone to

help classify vegetation, identify other objects Near Infrared 0.76pm – 0.9-pm Used to determine plant types, status and biomass,

soil moisture

Short-wave

Infrared

1.55pm – 1.75pm 2.08pm – 2.35pm

Used to determine the moisture of vegetation and soil, research mineral rock, separate snow and cloud

Thermal

Infrared

10.4pm – 12.5pm Used to determine time that vegetation is shocked,

soil moisture and temperature mapping

Panchromatic 0.53pm – 0.9pm Low resolution and continuous spectrum, this band

is used to overlay with other bands, from that measure objects exactly

Source: Climategis.com

2.4 Remote sensing and GIS application

With the development of computer and modern technology, remote sensing is becoming popular for resource management The use of remote sensing and GIS has gained mass momentum in recent years It is used almost every field in social and natural sciences, giving accurate, efficient methods for collecting, viewing and analyzing spatial data In the last few decades majority of work in remote sensing was mainly focused on environmental studies The implication of remote sensing and GIS to forest cover change and urban planning is now getting attention and interest among GIS and Remote sensing professionals GIS provides foresters with powerful tools for recording, keeping, updating information, supporting the decisions of manager and policy makers

The applications of remote sensing and GIS have been widely applied in many field from agriculture, forestry, land use land cover, natural disasters, geology and so on

2.4.1 Agriculture

Agriculture plays an important role in economy of country It provides food to everyone and getting highest food productivities is the desire of the farmers and agricultural agencies In agricultural field, GIS can be used to manage and monitor farming practices at different levels Satellite data can help determine the location and extent of crop stress Remote sensing can be used to prepare maps of crop types Satellites can provide information about the health of vegetation This information serves to predict grain crop yield, collecting crop production, mapping soil productivity, identification factors that influences crop stress (P K Kingra, 2016) and monitoring farming activity Remote sensing and GIS is essential in mapping areas that are vulnerable to natural disasters such as flood, drought

Remote sensing imagery gives the general spatial information of land It can identify

predicting soil degradation Soil surface after erosion have different colors, tone and structure than non-eroded soil, so that eroded parts can be identify by spatial images Remote sensing inputs are evolved for analyzing crop area and is a very useful tool in crop yield forecasting (Menon Arr, 2012)

2.4.2 Forestry

Remote sensing forms a valuable tool in mapping and monitoring of biodiversity and provides valuable information to quantify spatial patterns, biophysical patterns, ecological patterns, determine species richness and factors affecting species richness and for predicating response of species to global changes (Menon Arr, 2012).One of the basic applications of GIS is forest cover typing and species identification Forest cover typing consist of investigation mapping over a large area, species inventories are measurement of stand structure, characteristics GIS gives geographic and spatial data that allows manager to effectively manage planning process Remote sensing and GIS can identify various forest types that could be difficult to classify by traditional survey

GIS provides useful tools in fire management Forest fires significantly influence on vegetation cover, animals, soil, air quality Forest managers have used GIS to predict the fires Forest fire depends three elements: heat, fuel, and oxygen GIS can be used for fuel mapping, weather condition mapping and fire danger mapping GIS provide information about factors: slope, elevation, wind speed, relative humidity, cloud cover, temperature, dead fuel moisture (Sonti SH, 2015) Forest fire managers can base on these data and develop models to predict forest intensity

Remote sensing and GIS help to make decision in harvest planning Maps constitute a planning tool for identification of felling direction, exploitation route, and depots Users use remote sensing to collect data for example, timber harvesting, silviculture, predicting fuel wood, recreation opportunities and minimizing impacts of harvesting

2.4.3 Urban planning

Urban area is complex combination of buildings, roads, garden, cemetery, and sidewalk Urban provides several of problems: resource allocation, employment level, traffic congestion, infrastructure, GIS is a tool to manage the changing of urban area Remote sensing and GIS are useful tools to investigate and analyze city’s natural conditions, resource distribution, geographical layout of towns, road network, urban expansion and the change of urban area (Yinghui Xiao and Q.Zhan, 2009) Managers will use spatial data and maps to decide how to manage the landscape Remote sensing can identify traffic congestion, design the right place to build a road to ensure maximum relief of the roads already full capacity, the direction of the road, entry and exit junctions The same data also is used to ensure the best routes for public transport network

2.4.4 Land cover mapping

Land cover mapping is a typical application of remote sensing Land cover consists of forest, grassland, concrete pavement, water Land cover reflects human activities such as the use of land for resident, agriculture, industry GIS collects and updates data of land cover maps The change is detected by comparison between two multi-date images Remote sensing has an important contribution to make in documenting the change in land use/land cover on regional and global scales from the mid-1970s (Lambin et al., 2003) Map of land cover change is used to assess the external effects for example, human activities, natural disasters on land cover types and give appropriate policies for sustainable management

- Objective 1: To investigate the status of forest cover in Kim Tien

- Objective 2: To construct maps of forest cover change detection in Kim Tien

- Objective 3: To identify key drivers of forest change detection in the period

- Objective 4: To propose solutions for effective forest management

3.2 Methodology

3.2.1 Study site

Figure 3.1: Map of study site

Kim Tien is a mountainous commune located in South-West of Kim Boi District (Hoa Binh province), it is 4 kilometers far from Kim Tien to center of Kim Boi district, its

+ In the West, it is bordered by Hop Dong and Thuong Tien commune

Natural area of Kim Tien is about 2,178.79 ha

3.2 Methodology

3.2.1 Data sources

To track the change in the study site over period of time, the study tried to collect Landsat images with different data acquired of period from 1995 to 2018:

Table 3.1: Landsat images used in the study

instrument

Date acquired

LC81270462018158LGN00 Landsat 8 2018/06/07 30x30 127/46 LT51270462011187BKT00 Landsat 5 2011/07/06 30x30 127/46 LT51270462002290BJC00 Landsat 5 2002/10/17 30x30 127/46 LT51270461995175BKT00 Landsat 5 1995/06/24 30x30 127/46

Source: glovis.usgs.gov

Figure 3.2: Landsat image of study site in 1995, 2002, 2011, 2018

Besides that, the study collected documents related to forest land including: paper map of topography, current land use The study also inherited materials, reports about natural, socio-economic condition in Kim Tien, document of forest protection, support program for residents in study site

3.2.2 Interviewed data

Interview is a good way to get information directly from people that live in the study site Because they understand clearly about the place where they live so collecting information from them is very useful for research From interview, we can know the current status of land cover, forest cover change, management planning, as well as driving force of land cover change

Interview staff of management board to get information about the current forest cover types of the area, as well as polices and regimes management which are implemented in Kim Tien commune Therefore, the research could get an overview about the processing of formation and change in the area

Local people play an important role in management of forest Kim Tien commune, because they depend on and affect directly on the forest and land use They understand about the area better than anyone Therefore, interview local people is the best way to have real and accurate information about the area The information of interview provides a basis for assessing and monitoring local people’s impact on forest resources The interview local communities also can have an overview about effectiveness of management plan of the commune In this study, the interview local people to determine key driver of forest change detection during research period The interview is focus on old and middle-age people with experience So that, they could provide much more information about the area in many years ago

30 people belong to object of interview: staffs, authorities and residents of Kim Tien commune For the local people, the interview is focused on their purposed of using land, planting forest tree and harvesting forest activities Content of interviewing staff is focused

on the status of forest land, types of forest in the commune, plantation project was carried out, and local forest management

3.2.3 Data processing

The research used ArcGIS software to build map of the forest over the periods Image clustering channels were been collected including individual spectral channels due to needing combination and composition to easy conduct steps later After composition step, the study site boundary was created and used to cut study site from Landsat image

The process of interpretation and classification of images Landsat are presented in the figure 3.3:

Figure 3.3: Flowchart of Landsat images classification and change mapping

Clip study area

Composite

bands

Band combination

Raster calculation

3.2.4 Image classification using NDVI

Vegetation index is used as an indicator to quantify the greenness of plants within satellite image The vegetation index is widely used to determine the density of the vegetation, to assess the growth and development status of the plant, as the basis for data to predict pests, droughts, area productivity and crop yields The most used index is Normalized Difference Vegetation Index (NDVI) NDVI is developed for estimating vegetation cover from the reflective bands of satellite data The method employs the multispectral remote sensing data technique to find spectral signature of different objects such as vegetation index, land cover classification, concrete structure, road, urban areas, rocky areas and remaining areas NDVI is useful to determine vegetation change detection NDVI can be expressed as:

NDVI = (NIR – RED)/(NIR + RED) Where: NIR is near infrared channel RED is red channel

The NDVI is motivated by the observation vegetation, which is the difference between the NIR and red band, it should be larger for greater chlorophyll density In contrast, bare land has higher reflectance at red wavelengths and lower reflectance at near infrared wavelengths Value of NDVI ranges from -1 to 1 Value of NDVI low represents where vegetation cover is low Value of NDVI high where vegetation cover is rich Negative value indicates non-vegetated surface features such as water, barren land, ice, snow, or clouds If there is a land cover change somewhere between two dates, the NDVI differentiated image should have pixel value greater than or smaller than 0

NDVI very low value of NDVI about 0.1 and below represent a area of snow, rock, sand Water like ocean, lakes, and rivers have a low reflectance in spectral channels that have

a negative NDVI value Value from 0 to 0.5 is the area of spare vegetation, 0.5 to 1 shows

to 0.3 represent shrub and grassland High values represent the temperate and tropical rainforests from 0.6 to 0.8 While bare soil is represented with NDVI value close to 0

3.2.5 Field survey

Field survey is conducted to collect ground control point with the help of Global Positioning System (GPS) The survey arm is determining the land cover types, providing data for classification image, and for accuracy assessment The research selects 230 point in the field, including forest and other land use types The total number of selected points is distributed evenly across the entire study site

Table 3.2: The object-based classification framework

3.2.6 Accuracy assessment

Accuracy assessment is used to evaluate the accuracy of image classification methods

or compare the reliability of the results of different remote sensing image classification methods The study could know how accurate the classification is by using selected points from field survey and Google Earth In this study, a total of 230 reference points include: forest, agriculture and other were selected to serve as samples for the classification

This accuracy assessment represents the relationship between the classes in the digital map and field survey/Google Earth This study assesses the accuracy of classification method

by using Kappa coefficient:

Where:

N: Total of sample point

r: The number of classes are classified

xii: Number of correct point in ith class

xi+: Total number points of ith class

x+i: Total number points of ith class after classified

Kappa is always less than or equal to 1, sometime it can be negative Kappa

coefficient can be interpreted as the table below:

Table 3.3: Interpretation Cohen's Kappa value

Source: Landis, J R and Koch, G G.1977 Biometrics 33: 159-174

3.2.6 Forest change detection

This step is used to detect forest change in study site during the period of time Detection of forest change is achieved by overlay and post-classification comparison of the forest cover map of different time This process was conduct by ArcGIS software The information of overlay map is a coincidence of unchanged objects and the difference of

Topography of commune is divided into two types:

- The mountainous terrain of the commune is surrounded by high mountains with steep slopes, alternating between small scattered valleys

- The lowland area is relatively flat valley, which are the production and living areas of the local people Because of mountainous areas and surrounded by high mountains, the main source of water for production activities is from streams

4.1.2 Climate

Kim Tien commune lies in the area that affected by tropical monsoon climate The weather is divided into two distinct seasons: rainy season lasts from February to September The climate is hot and temperature quite high Dry season lasts from the end of September to January of the next year The weather is dry, cold and often has foggy

Generally, the climate and temperature of Kim Tien commune is relatively favorable for the development of agriculture and forestry Due to two season and other weather conditions such as storm, thunderstorm and Northeasterly monsoon require measures to prevent flood and drought

4.1.3 Soil

- Hilly land group: these include red soils on neutral macaque, basalt, limestone and barren soil on metamorphic rocks, of which yellowish feralite is the most common These soil types are suitable for forestry and fruit trees

- Field land: The main is feralite soils that are modified by rice cultivation, which is quite humus-rich, with less acidic, light mechanical components and low nutrient content Rice, crops and annual industrial trees are suitable to grow on these soil types

Table 4.1: Land status of Kim Tien commune

1 Hilly land group 1,472.58 Surrounding the 8 hamlets

4.1.4 Natural resources

4.1.4.1 Land resource

Total natural area of Kim Tien commune is 2,178.79 ha, that consist of:

- Agricultural area is 1,640.27 ha, accounting for 75.28 percent of total area of the commune Including:

 Land for agricultural production is 165.43 ha accounting for 7.59% of total area

 Land for forestry is 1,472.58 accounting for 67.59% total area of the commune

 Aquaculture land is 2.26 ha accounting for 0.1% total natural area

- Non-agricultural land is 197.37 ha accounting for 9.06% total area, that includes:

 Rural land is 118.65 ha accounting for 5.45% of total area

 Specialized land is 58.63 ha accounting for 2.69% of total area

 Cemetery land is 10 ha accounting for 0.46%

 Streams, surface water area is 10.09 ha accounting for 10.09 ha accounting for 0.46%

The vegetation in the commune is abundance with many kinds of plants Vegetation is mostly grow for economics purposes Apart from the role of economic production, it also has the effect of protecting and creating landscapes and regulating the climate of the region Current forest area is only poor forest and forests are restored after shifting cultivation, due to illegal logging, exploitation does not take into account long-term benefits Percentage of forest area is low mainly is plantation forest Besides forest, natural vegetation in commune is shrub, weed

4.1.4.3 Water resource

Water surface includes Chao river, that has total length of 5.5 kilometers and canal system inside field, ponds, lakes, dams that are distributed scattered in the commune Water surface system is exploited to provided water for agricultural production, and daily live of local people

Groundwater source has not been specifically measured but the practical use of people

in the commune shows that: dug wells have a depth from 4 meters to 10 meters, small family wells often have a depth from 15 meters to 30 meters

4.1.4.4 Minerals

According to the survey, Kim Tien is a poor commune for mineral resources, mainly

is sand and gravel Area of mountain and hill is quite large, slope is average This is a good condition for growing forestry trees and husbandry

4.2 Social-economic condition

Kim Tien commune includes 8 hamlets: Chao I, Chao II, Doi I, Doi II, Go Cha, Go

Mu, Vo Khang, Go Khanh

In 2018 Kim Tien has 1025 household and population is 4860 98% of population of Kim Tien is Muong ethnic, the rest is Tay, Kinh ethnic

4.2.1 Road system

The whole commune has 6 main roads leading through 6 hamlets with a total length

of 5.82 kilometers, of which 3.8 kilometers is concreted, accounting for 65.8% and the remaining 2.02 km is the land road, occupying 34.82% There are 60 hamlet-route with total length is about 23.93 kilometers, of which total length of concreted road is about 8.52 kilometers accounting for 34.61%, but the road is still narrow, the road width is 3.5 meters The rest is dirt road with total length is 15.14 kilometers, accounting for 64.39%, road is small hand narrow It does not meet the demand of daily life and production of people in the commune, which needs to be expanded and enhanced

Currently, there are 9 field-routes in Kim Tien commune, with total length is 2.15 kilometers All field-routes are dirt roads and have not been concreted, surface road is narrow, does not meet the requirements of mechanization in production and travel

4.2.2 Irrigation system

The whole commune has one pumping station in Chao I hamlet, provides water for Chao I, Chao II, Go Cha, one water reservoir in Vo Khang that provides water for irrigation

of Vo Khang, Chao I, Chao I, Go Khanh However, pumping station and reservoir have been built for a long time so they was degraded and have not met the demand for irrigation

Bai Cai pumping station (Chao I hamlet) has total capacity of 80 m3/h provides water for 6ha, pumping station has small capacity and has been built for a long time that has met the demand for irrigation in production

The commune has been had 14.74 kilometers irrigation canals All of which, there are 4.67 kilometers are concreted, the remain 10.07 kilometers is canal ditch The commune plans to open new canals and concretize main and important canal routes to ensure the irrigation demand for production

4.2.3 Electricity

The current electricity supply to the commune is from the 35KV transmission line in

Ha Bi The power supply system for the commune includes 3 transformer stations with the total capacity of 335KVA

- Vo Khang transformer station: capacity is 75KVA, provides electricity for 260 households

- Chao I transformer station: capacity is 160KVA, provides electricity for 421 households

- Doi I transformer station: capacity is 100KVA, provides electricity for 185 households

Low voltage line is 10 kilometers and includes three lines:

- Line 1: Vo Khang, Go Mu hamlet, the length is 3.3 kilometers

- Line 2: Go Khanh, Chao I, Chao II, Go Cha, the length is 4.5 kilometers

- Line 3: Chao II, the length is 1.8 kilometers

The entire power system (includes transformer stations and line) is managed by Kim Boi Power Company and sells electricity to household Low voltage line system installed to meet the standards of the power industry, the quality of the infrastructures ensure good, meet the new rural criteria 80.6% of total households is used electricity often and safety

4.2.4 Education

4.2.4.1 Status of education level

- The commune has two hamlet get the standard of cultural village, accounting for 25%

- 100% residents in commune reach secondary-education level

- Total students graduated from secondary school are 56, and 37 students attend high school, it accounts for 66.07%

4.2.4.2 School

4.2.4.2.1 Kindergarten

Kim Tien commune has one main kindergarten and 5 small kindergarten located in hamlets Total area of kindergarten is 6000 square meters The area of the school meets the criterion of area according to criteria of the new rural

School has just been built but has not reached the level 1 of National standard Learning equipment has been degraded and has not met the long-term demand

Kindergarten has 13 classes, 13 classrooms, 3000 square meters playground, 33 teachers of which there are 3 teachers graduated from university, 5 teachers graduated from college

4.2.4.2.2 Primary school

Primary of Kim Tien has not considered as standard school according to the National standard The whole commune has one primary school is located in Chao II, near the main

road is convenient for students to go to school The school was built in 2003, has two floors The area of school is 6,784 square meters The area of playground is 3,600 square meters

School has 13 classes and 15 classrooms There are 29 teachers in schools, of which 6 teachers graduated from university, 13 teachers graduated from college, 10 teachers reach intermediated level

At present, the campus has reached the target of the new rural criteria However, furniture and equipment for students have deteriorated significantly, failing to meet the requirements for long-term learning Kim Tien commune aims to plant more trees, invest in equipment and facilities according to education branch standards, aiming to achieve the level

At present, the campus has reached the target of the new rural criteria However, furniture and equipment for students have deteriorated significantly, failing to meet the requirements for long-term learning Kim Tien commune aims to plant invest in equipment and facilities according to education branch standards, aiming to achieve the level 1 of National standard

School has 8 classes, 8 classrooms, no class has reached the National standard School

School has 21 teachers, 11 teachers meet university level, 8 teachers reach college level, and one teacher meets intermediated level

4.2.5 Cultural facilities

The commune has not had culture house and playground, each hamlet has its own culture house and playground However the area of playgrounds has met the standard of new rural criteria

Status of culture houses of each hamlet:

- Area of Doi I’s culture house is 1,300 cubic meters

- Area of Doi II’s culture house is 4,500 cubic meters

- Area of Chao I’s culture house is 800 cubic meters

- Area of Chao II’s culture house is 700 cubic meters

- Area of Go Khanh’s culture house is 400 cubic meters

- Area of Go Mu’s Culture house is 200 cubic meters

- Area of Go Cha’s culture house is 500 cubic meters

- Area of Vo Khang’s culture house is 2,600 cubic meters

The area of culture houses of hamlets reaches the criteria of new rural criteria about area, but the infrastructure of culture houses is not enough, needs to be improved and upgraded

4.2.6 Residential houses

Total residential houses in 2011 is 947 Of which:

- The number of houses of poor households need to be repaired and improved is 36, accounting for 3.8% Houses of poor households are built from rocks, cement, wood

or bamboo, rooftop is roof tiles and it can not prevent the heat effect from sun Houses and doors are small, low so that air circulation is poor, house is easy to get wet and moldy, affects to human health

- Cottages are 365 accounting for 38.5%

- Semi-permanent houses are 475, accounting for 50.2%

- Solid houses are 71, accounting for 7.5%

Semi-permanent houses and solid are built suitable with the climate condition of commune, they are airy, cool and convenient for daily life of local people

The economy of the commune is moving towards reducing the proportion of agriculture, increasing the proportion of industry

At present, the commune has one company is Mat Troi waterfall Trading & Tourism Joint Stock Company with investment of over 5 billions VND, 10 households engaged in transportation, 65 households do cottage industry – Trade in Services – Building Trade Agricultural production: Although the area of cultivation is reduced because of changing to serve other purpose, total production of grain in 2011 is 1,884 tons Average per capita food

is 459.9 kilograms/person/year, ensure the food security

In recent years, livestock sector of the commune has grown quite rapidly However, because of extreme weather, cold weather, in 2008, 95 cows died and during the extreme cold from January to March in 2011, 6 buffaloes and cows died So that the growth rate of the