See discussions, stats, and author profiles for this publication at: [Link]

net/publication/275225369

Forest Cover Change Detection Using Remote Sensing and GIS in Banja
District, Amhara Region, Ethiopia

Article in International Journal of Environmental Monitoring and Analysis · December 2014

DOI: 10.11648/[Link].20140206.19

CITATIONS READS

38 7,566

4 authors, including:

Adane Birhanu Gedif Solomon Addisu

Bahir Dar University Bahir Dar University College of Agriculture and Environmental Sciences Natural R…
13 PUBLICATIONS 500 CITATIONS 94 PUBLICATIONS 1,207 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Adane Birhanu Gedif on 14 October 2024.

The user has requested enhancement of the downloaded file.

International Journal of Environmental Monitoring and Analysis
2014; 2(6): 354-360
Published online December 23, 2014 ([Link]
doi: 10.11648/[Link].20140206.19
ISSN: 2328-7659 (Print); ISSN: 2328-7667 (Online)

Forest cover change detection using remote sensing and

GIS in Banja district, Amhara region, Ethiopia
Abyot Yismaw1, Birhanu Gedif2, Solomon Addisu3, Ferede Zewudu4
1
Bureau of Mines and Energy Agency, Amhara National Regional State, Bahir Dar
2
Director for Geospatial Data and Technology Center, Lecturer at College of Agriculture and Environmental Sciences, Department of Natural
Resources Management, Bahir Dar University
3
PhD Research Scholar at Andhra University, College of Science |& Technology, Environmental Sciences Department
4
Assistant Professor, at College of Social Sciences, Department of Geography and Environmental Studies, Bahir Dar University

Email address:
abotyismaw@[Link] (A. Yismaw), birhanu1968@[Link] (B. Gedif), soladd2000@[Link] (S. Addisu),
feredez@[Link] (F. Zewudu)

To cite this article:

Abyot Yismaw, Birhanu Gedif, Solomon Addisu, Ferede Zewudu. Forest Cover Change Detection Using Remote Sensing and GIS in Banja
District, Amhara Region, Ethiopia. International Journal of Environmental Monitoring and Analysis. Vol. 2, No. 6, 2014, pp. 354-360.
doi: 10.11648/[Link].20140206.19

Abstract: Forest cover change analysis between 1973 and 2003 was conducted in Banja district, Awi zone, Amhara region
using remote sensing and geographic information system supported with field verifications. Information was extracted from
various Satellite images and different digital maps. The objectives of this study were to detect the extent and rate of forest cover
change over the last 30 years. Three dates of Landsat image data of the 1973, 1986 and 2003 were used to produce land cover
map in general and forest cover map in particular. Normalized difference vegetation index (NDVI), image differencing and
post-classification comparison change detection methods were employed. In addition to this, socioeconomic data were used in
explaining the drivers of forest cover changes in the study area. The results showed that during the last 30 years, forest cover
declined from 6044ha in 1973 to 2855.9ha in 1986 and 2446.9ha in the year 2003. The annual rate of forest cover change
between 1973 and 2003 was 120ha/year. The socioeconomic factors like population growth, the demand for the expansion of
agricultural land, fuel wood and construction materials were the major driving forces for the observed forest cover changes.
Therefore, in order to reduce the problem of forest cover change, remedial actions are recommended.
Keywords: Forest Covers Change, Remote Sensing, Geographic Information System, Satellite Image Analysis

Change detection is the process of identifying differences in

1. Introduction the state of an object or phenomenon by observing it at
Forest cover changes is a dynamic, widespread and different times. Remote sensing based change detection
accelerating process, mainly driven by natural phenomena applies comparison of a set of temporal images covering time
and anthropogenic activities, which in turn drives changes that period of interest using specific change detection algorithms.
would impact natural ecosystem. Understanding forest To address this issue, technology has developed and the
patterns, changes and interactions between human activities possibilities are virtually unlimited in different areas of
and natural phenomenon are essential for proper forest applications, which can be addressed through earth
management and decision improvement. Today, data from observation satellite data and decisions support tools such as
satellites are very applicable and useful for forest cover Geographic Information System [1].
change detection studies. Detecting forest conditions as well The problematic nature of forest degradation in connection
as monitoring the changes of various forest structural and with spatio-temporal change of forest cover change needs a
biophysical variables can enable accurate understanding of profound and ongoing research to act accordingly. In this
forest ecosystem services. Monitoring of forest cover change respect, the depletion of forest cover has become a major
is one of the main applications of remote sensing based change ecological problem Ethiopia. Regarding this, EFAP [2] stated
detection. that by the early 1950s forest cover of the country was reduced
to 16% of the total area, early 1980s by 3.6% and about 2.7%
 International Journal of Environmental Monitoring and Analysis 2014; 2(6): 354-360 355

in 1989. The rate of deforestation is estimated at 150,000ha Both primary and secondary data have been used in this
per year. This indicated that rapid deforestation and research. The use of each data has been found to be relevant
accelerated land degradation have led to deterioration of the and important in the research process. Primary data used in the
country’s environment. study were: Landsat MSS data image acquisition date of
In relation to the available information of forest cover in 02/01/1973 with four bands, Landsat TM data image
the country, EFAP [2] documented that there is no adequate acquisition date of 01/03/1986 with seven bands, Landsat
information on the location, extent of the remaining forest ETM+ data image acquisition date of 01/26/2003 having eight
cover of the country and the rate at which this resource is bands and SRTM (Shuttle Radar Topographic Mission) DEM
depleted. However, there were some works done to offer data 90m resolution. In order to assess the causes that brought
information on forest cover of the country. Due to massive about forest cover change; questionnaires have been designed
exploitation, the forest resource of the country has confined and distributed. The survey instruments used for collecting
itself to small remnants on the highlands particularly, almost data were structured and semi-structured questionnaires. The
all located at unreachable areas [3]. In the study area the research findings from the primary data have been
Kathasa Priority State Forest is one of the few remnant supplemented by secondary data sources such as published
forests which are found in Banja district. Furthermore, like in and unpublished materials. Garmin GPS was used for the
many other parts of the country, the problem of forest cover collection of ground control points for image classification.
change is a very serious environmental problem in the study Field observation sheet was prepared for storing all land use
area and extensive areas of forest cover including shrub lands /land cover information types. Totally 121 ground truth points
have been deforested. However, the rate and areal extent of were collected. Images (MSS 1973, TM 1986 and ETM+
the forest cover change is not well studied till date. Thus, for 2003) were projected to similar projection and datum, UTM
a sustainable forest resource management, it is necessary to projection and Adindan datum.
estimate forest cover change on large spatial and temporal Radiometric correction, Histogram equalization haze and
scales. The General Objectives of this research was to detect noise reduction were performed. The three images have
the magnitude of forest cover change. The Specific objectives different spatial resolution. Spatial resampling was performed
include identifying forest cover and its spatial distribution, to using ENVI 4.2 software. By doing this Accordingly, the
analyze the spatio-temporal change, to create forest cover spatial resolution of all images become 28.5m by 28.5m.
map of the study area for the years 1973, 1986 and 2003 and Furthermore, resolution merges was also conducted to
to assess the cause of forest cover change. increase the spatial resolution of multi-spectral images. Both
unsupervised and supervised image classification techniques
2. Methods were applied. Unsupervised classification was done before
field work. For the supervised image classification training
The study was conducted in Banja District, Awi areas were established based on the ground truth taken during
Administrative zone, Amhara National Regional State, field work. Among different algorithms in the supervised
Ethiopia. Geographically, the study area lies within 10052’to classification maximum likelihood was utilized.
110 3’N latitude and 36038’to 3708’E longitude at a distance of Having applied the techniques of image classification
440 km Northwest of Addis Ababa and 120km South of Bahir methods, land use / land cover types were identified in order to
Dar, the capital of Ethiopia and Amhara regional state detect forest cover change. With the help of visual
respectively. The altitude of the study area ranges from 1800 interpretation and the different reflectance characteristics of
to 2953m.a.s.l. Most of the eastern parts of the district are the features in the satellite images of 1973, 1986 and 2003
2425m.a.s.l (Fig.1). eight Land use/Land cover classes, namely; Forest, Shrub land,
Grazing land, Agricultural land, Settlement area, Bare land,
Water body and Marsh have been identified with the support
of ERDAS Imagine Software. Each land use/land cover
description was stated as follows:-
Forest: It represents both natural and fragmented
plantation forest areas that are stoked with trees capable
of producing timber or other wood products.
Shrub land: Areas covered with small trees, bushes and
shrubs, mainly with less crown cover.
Settlement: are those areas composed of intensive use
with much of the land by rural villages, towns and roads.
Agricultural land: are lands covered with agricultural
activities.
Grazing land: All areas covered with mainly natural
pasture, but also other small sized plant species.
Bare land: Areas under degraded grasslands and with
Figure 1. Location of the study area
some areas that are bare ground (rocks).
356 Abyot Yismaw et al.: Forest Cover Change Detection Using Remote Sensing and GIS in Banja District,
Amhara Region, Ethiopia

Marshy: Lands that are seasonally under water. matrix which is a report of the overall proportion of correctly
Water body: lands completely covered with water. classified pixels. Finally, Kappa Statistics was calculated for
All the existing images were classified in to eight Land the different areas that were classified.
use/Land cover types. To detect the changes of forest cover The Kappa coefficient result values are between 0 and 1,
over time, NDVI and Post-classification Change Detection where the latter shows complete agreement, and is often
comparison methods were applied. multiplied by 100 to give a percentage measure of
NDVI is calculated using the following equation: classification accuracy. Kappa values are also grouped into
three groups: a value of kappa coefficient greater than
NDVI = (NIR - RED) / (NIR + RED) 0.80(80%) represents strong agreement, a value of kappa
Where, NIR and RED is the reflectance in the near infra-red coefficient between 0.40 and 0.80 (40 to 80%) represents
and red bands, respectively. NDVI is a nonlinear function that moderate agreement, and a value of kappa coefficient below
ranges between -1 and +1. Healthy vegetation has high 0.40(40%) represents poor agreement [4].
positive NDVI values because of their relatively high
reflectance in NIR and low in visible wavelength. Having 3. Results and Discussions
conducting NDVI analysis of the year 1986 and 2003, the
mean and standard deviations values are summarized to 3.1. Land Use and Land Cover Results
evaluate the trends of vegetation cover change. The land use/land cover units of the study area were
Change detection involves the use of multi-temporal data classified in to classes of forest, Shrub land, Grazing land,
sets to differentiate areas of land cover change between dates Agricultural land, Settlement, Bare land, Marsh and Water
of imaging. This kind of change detection method identifies body. The statistics of land use /land cover change in general
where and how much change has occurred. In the meantime, and forest cover change in particular were computed and
four conditions of forest cover change detection summarized to detect the nature of the changes based on the
characteristics such as, detecting the changes that have years 1973, 1986 and 2003.
occurred, identifying the nature of the change, measuring the The dominant land use, grazing land and agricultural, takes
areal extent of the change, and assessing the spatial pattern of 33,848.2ha of the total area, forest covers 6044ha, shrub land,
the change are explored. Besides, change detection matrix has which mainly contains small trees, bushes and shrubs, covers
been produced to explore the trends and patterns of land 4910.9ha, settlement covers 751.4, bare land covers 305.3ha,
use/land cover change in general and forest cover change in marshy and water body covers 97.2 and 29.3, respectively.
particular. For the current study, the rate of forest cover change Agricultural land take the highest share in the 1986 land
was also calculated using the formula below:- use/land cover class covers 21,631.6ha, followed by grazing
land which accounts 12,308.1. Shrub land which mainly
r = a − b ........................Equation 2
t contains small trees covers 7,989.5 ha, forest 2,855.9 ha,
settlement 891.9ha, bare land 214ha and marshy 53.9ha and
Where, r = Rate of forest covers change water body 40.9ha takes the minimum area coverage
a = Recent year forest covers in ha (Fig.2).The land cover classes for 2003 takes the highest share
b = Initial year forest covers in ha for agricultural and grazing land covering 28,545.6 and
t = Number of years between a and b 10,157.8 respectively followed by forest 2446.9ha, shrub land
To validate and crosscheck the result of the Landsat image 1937.9ha,settlement 1552.5ha, bare land 951.9ha,
classification with known ground truth data, accuracy marshy346.1 and water body 47ha of land cover from the total
assessment was done for the year 2003 using ENVI4.2 area of the Banja district.
software. Accuracy assessment evaluation includes an error

Figure 2. Land use/Land Cover map for 1973(left), 1986(right) and 2003 (Below)
 International Journal of Environmental Monitoring and Analysis 2014; 2(6): 354-360 357

3.2. Change Detection forest and shrub land became reduced in areal coverage as
compared to the 1973 and 1986 land cover / land use classes.
Change detection method was employed to categorize But the agricultural land has been on the pace of increment,
forest cover changes between 1973, 1986 and 2003. In accounting for 65% of the total area.
addition, the forest cover change in the form of maps and To clearly understand the major land cover change source
statistics has been assembled to examine the specific nature and destination of cover classes change conversion matrix for
and extent of the forest cover changes between the stated dates each period is analyzed. Change matrix has been produced on
of imageries in the study area. The rate of land use/ land is the basis of 1973, 1986 and 2003 satellite images
presented in Table 1. classification and presented in Tables 2, Table 3 and Table
Based on the 1973 land use / land cover classes, about [Link] three Confusion Matrix table show that the areal
48% was accounted for grazing land where as agricultural distribution of each land cover/ land use classes that have
land shared 25% and forest shared 13% and shrub land takes undergone transformation from one type to another or being
11% of the total area of the Banja district. The smallest lost their areal extents or remained intact.
share went to settlement, bare land, marshy and water body. The amount of land in hectare changed into other land use
In1986, agricultural land showed an increase in areal type obtained from other land use type. For instance, 11,176ha
coverage due to the need to secure more agricultural lands. of grazing land, 2553.3ha of shrub land and 1503ha of forest
However, grazing land which was likely to be changed in to of 1973 were converted to agricultural land. Whereas the
other land use / land cover unit showed change in decline. bolded diagonal values stand for the unchanged land use / land
Apparently, shrub land and settlement coverage have got cover that maintained its original land cover / land use unit
increment in areal coverage. Regarding 2003 land use / land (Table 2).
cover unit, the size of three classes, namely; grazing land,
Table 1. Rate of land covers change

Land use Land cover 1973-1986 1986-2003 1973-2003

categories Area (ha) Rate of change(ha/yr) Area (ha) Rate of change(ha/y) Area (ha) Rate of Change (ha/yr)
Agricultural land 10349.3 796.1 6914 406.7 17263.3 575.4
Grazing land -10257.8 -789.1 -2150.3 -126.5 -12408.1 -413.6
Forest -3188.1 -245.2 -409 -24 -3597.1 -120
Shrub land 3078.6 236.8 -6051.6 -356 -2973 -99.1
Settlement 140.5 10.8 660.6 38.9 801.1 26.7
Bare land -91.3 -7 737.9 43.4 646.6 21.6
Marshy -43.3 -3.3 292.2 17.2 248.9 8.3
Water body 24.4 1.9 6.1 0.4 17.7 0.6

Table 2. Matrix of Land Cover/Land Use Changes Between1973 and1986

Land Cover/Land Use Units of 1986

Grazing Bare Water
Classes Agricultural land Forest Shrub land Settlement Marshy
Land land body
Land Agricultural land 5841 2865.4 396.8 1916.5 53.8 187.8 14.6 5.5
Cover/L Grazing land 11176 7499.6 409.5 2785 119 563.1 12.1 1.5
and Use Forest 1503 680 1760 1983.1 6.8 59.7 9.7 7.2
Units of Shrub land 2553.1 915.7 502.6 1085 22.2 64 14.4 3.9
1973 Bare land 137.8 83.3 13 63.8 2.8 3.3 1.2 0
Settlement 349.9 225.7 22.2 132.9 9.3 11.1 0.2 0
Marshy 34.6 37.7 1.7 15.8 0.2 2.8 1.3 3.2
Water body 2.1 0 0 7.5 0 0 0.3 19.4

Table 3. Matrix of Land Cover/Land Use Changes Between 1986 and 2003

Land Cover/Land Use Units of 1986

Grazing Shrub Bare Water
Classes Agricultural land Forest Settlement Marshy
Land land land body
Land Agricultural land 15252.8 4584.9 227.8 583.8 446.7 353.5 182.2 0.1
Cover/L Grazing land 6349.4 4359.6 37.7 96.3 349 1029.4 86.7 0
and Use Forest 887.8 82.2 1478.4 384.4 3.7 15.2 4.1 0
Units of Shrub land 5535 704 700.6 846.4 74.5 66.2 60.2 2.5
1973 Bare land 123.2 34 0.4 2.2 36.5 17 0.6 0
Settlement 378.8 389 1.5 7.7 41.6 71 1.8 0
Marshy 18 3.9 0.3 16 0 0.2 10.5 4.6
Water body 0 0 0 1.6 0 0 0 39.8
358 Abyot Yismaw et al.: Forest Cover Change Detection Using Remote Sensing and GIS in Banja District,
Amhara Region, Ethiopia

The type and amount of conversion in the land use land agricultural land (5535ha) (Table 3).
cover from 1986 to 2003 within 17 years. Land with areas of From 1973 to 2003 within 30 years. 7633.3, 5939.71726,
15252.8ha, 4359.6ha, 1478.4ha, 846.4ha, 36.5ha, 71, 10.5ha 266.6, 13.1, 34.5, 0.2 and 20.1 that classified as agricultural
and 39.8ha classified as agricultural land, grazing land, forest, land, grazing land, forest, shrub land, bare land, settlement,
shrub land, bare land, settlement, marshy and water body marshy and water body respectively remained unchanged. On
respectively remain unchanged. On the other hand, the the other hand, the conversion of land takes place from one
conversion of land takes place from one types of land use in to types of land use in to other land use types. For instance,
other land use types. agricultural land to grazing land was (2396.9), grazing to
For instance, agricultural land to grazing land (4584.9ha), agricultural land (14363.7) and shrub land to agricultural land
grazing to agricultural land (6349.4ha) and shrub land to (3275.6) (Table 4).
Table 4. Matrix of Land Cover/Land Use Changes Between1973 and 2003

Land Cover/Land Use Units of 2003

Bare Water
Classes Agricultural land Grazing Land Forest Shrub land Settlement Marshy
land body
Land
Agricultural land 7633.3 2396.9 254.2 301.4 233 351.5 104.2 6.5
Cover/
Grazing land 14363.7 5939.7 233.2 433.8 558.6 907.7 125 4.1
Land
Forest 2501.5 712.7 1726 887.3 33 126.2 50 7.6
Use
Shrub land 3275.6 884.8 217.3 266.6 83.7 116.2 61.6 5.1
Units
Bare land 207.5 53.8 4.4 9.3 13.1 15.1 2.2 0
of1973
Settlement 508.5 143 9.3 23.3 29.9 34.5 2.9 0.1
Marshy 52.4 27 2.4 9.9 0.6 1.2 0.2 3.5
Water body 2.8 0 0.2 6.3 0 0 0 20.1

Table 5. Trends and Rates of Forest cover change

Forest cover in hectare for Rate of Change

three years Area change (ha) (ha/yr) Change (ha) (ha/yr) Area change (ha) ha/yr
1973 1986 2003 1973-1986 1973-1986 1986-2003 1986-2003 1973-2003 1973-2003
6044.4 2855.9 2446.9 -3188.1 -245.2 -409 -24 -3597.1 -120

3.3. Areal Extent and Rate of Forest Cover Change Table 6. Total Forest Cover of 1973, 1986 and 2003

Three Land sat satellite images of 1973, 1986 and 2003 Forest Cover unit from the total area Forest Cover
Year
(in ha) In %
were used to monitor the areal extent and rate of forest cover
1973 6044.4 13
change within time series. Throughout the analysis stage, 1986 2855.9 6
digital image interpretation of forest cover area for each year 2003 2446.9 5
was performed and total area of the forest cover in hectare and
its percentage from each date of satellite interpretations were 3.4. Patterns of Forest Cover Change
calculated and summarized. Forest cover map and total forest
The result shows the areal share of forest cover lands and
cover of 1973, 1986 and 2003 is presented in Table 5.
also gives information about the amount of forest cover land
From this result, about 6044 ha of the area were covered
that was converted into other land cover and land use units of
with forest resource in the year 1973. Meanwhile, the forest
the three periods. The pattern of forest cover change into
cover land of the area accounted for 2855.9ha and 2446.9ha in
other land use/ land cover units between in the year 1973 and
the year 1986 and 2003 respectively.
1986, 1986 and 2003 and 1973 and 2003 is presented in Table
In the year 1973, 13% of the area was covered with forest
7. 4249.5ha of forest cover land are changed into other land
resources while from the total area of the area about 6 % was
cover and land use units between1973 and 1986. Specifically,
covered with forest resources in 1986. Meanwhile, forest
48% of the forest cover is changed into Shrub land followed
resources of the area were turned down in to 5% in the year
by forest cover transformed in to agricultural land (35%). The
2003 from the total area.
remaining 16%and1% of the forest cover land is converted
The rate of forest cover change from year 1973 to 1986 is
into grazing land and settlement respectively. From the 1986
-245.2 ha per year (6044.4ha –2855.9ha/13 years) and from
and 2003, 1,377.4ha of forest cover land are changed in to
year 1986 to 2003, it was -24 ha annually (2855.9-2446.9ha/
other land cover units. The conversion of forest land to
17years).
agricultural land takes the lion share, about 65%. The
Besides, considering the annual rate of forest cover change
remaining 28%, 6% and 1% of the forest is transformed into
between1973 and 2003, the computed result is -120ha per year
other land cover/use units of shrub land, grazing land and bare
(6044.4-2446.9/30). The rate of forest cover change from 1973
land respectively.
to 1986, 1986 to 2003 and 1973 to 2003 is presented in Table 6.
 International Journal of Environmental Monitoring and Analysis 2014; 2(6): 354-360 359

Table 7. Patterns of Forest Cover Change in to Other Land Cover/ Use Units

B/N 1973 & 1986 B/N 1986 & 2003 B/N 1973 & 2003
Forest Cover Change
Area(in ha) Percentage Area(in ha) Percentage Area(in ha) Percentage
Forest to Agricultural land 1503 35 887.8 65 2501.5 57
Forest to Grazing land 680 16 82.2 6 712.7 17
Forest to Shrub land 1983.1 48 384.4 28 887.3 21
Forest to Bare land 6.8 0 3.7 0 33 1
Forest to Settlement 59.7 1 15.2 1 126.2 3
Forest to Marshy 9.7 0 4.1 0 50 1
Forest to Water body 7.2 0 0 0 7.6 0
Total change 4249.5 100 1377.4 100 4318.3 100

3.5. The Socio-Economic Results others) and baking injera among the total sampled house hold
98% of the sample house hold responds use of open hearth the
Forest cover change is the direct reflection of the dynamics remaining answered “mitine stove” 1% and Charcoal Stove
of socio-economic development. Concerning this, Badege 1%. The growing demand for forest resources by rural and
(2001) states that forest cover change has occurred starting urban sectors is responsible for the massive scale of
from an early time a small level of expansion of agriculture by deforestation. Deforestation in the study area is happening as a
deforestation of the land that covered by the natural forest. result of expansion of agricultural activities, wood
Similarly, numerous factors inspired by the action of human consumption for fuel, construction and other uses.
being are accountable for the huge changes of the forest cover This section helps to understand respondents’ observed
land into other land cover and land use units in the study area. about the impacts of forest destruction on their locality. All the
In the case of forest cover change in study area is a respondents were asked the impact of the destruction of forest
miniature case of the general deforestation that occurs resource on their locality. Accordingly, the majority of the
throughout Ethiopia. Out of the total sampled 99% of the respondents 62% observed more erosion. It is only 13% of the
respondents reported that there was a decline of forest cover respondents have observed more flooding. The rest 25 %
compared the current forest status with forest status before encountered loss and decline of fertility of land (Table 8).
1973. The remaining was preferred the increment in the status Regarding the solutions for deforestation, about 40% of the
of the forest. respondents suggested strong law as a solution to control
Table 3 indicated that there is an increasing of agricultural forest depletion. Providing environmental education as a
land from 11,282.3ha in 1973 to 21631.6ha in the year 1986 solution was also indicated by 23% of the respondent. Those
and 28,545.6ha in the year 2003. The indication of agricultural who said looking for another source of income was only 19%.
land expansion in terms of spatial coverage means agricultural The rest 18 percent pointed out planting trees pointed out as a
land expanded in the expense of other land cover/land use solution.
units. For instance, between the years 1973 and 1986 about
1503ha, between the year 1986 and 2003 about 887.8ha and Table 8. Principal energy sources in the study area
between the years 1973 and 2003, about 2501.5ha forest cover Percentage of household
land is drastically changed into agricultural land. This also Energy source Number of households
using
supported by the survey 76% of respondent’s views confirmed Fuel wood 137 96
that, the expansion of various types of agricultural activities is Cow dung 2 1
the major sources of forest cover change in the study area. Crop residues 4 3
6.1% thought that human consumption for fuel, 2.6% charcoal
making, and 15.4% cutting of trees for house and fens 4. Conclusion
construction as the major causes of deforestation.
Rural households are totally dependent on biomass for Forest cover change in the form of deforestation is a major
[Link] this, the Ethiopian Forestry Action Program environmental problem manifested at Banja district .In the
(1994) estimated that in 1990/91 Ethiopia was used about 15 district, forest resources decreased by 3188.1ha, 409 and
million tons of energy, of which 95% was wood, dung, crop 3597.1ha between 1973and 1986, 1986 and 2003, and 1973
residue and charcoal. For instance, as indicated in the current and 2003 respectively. An increasing demand for agricultural
survey 96% of the sample households in the study area derived land was the cause for the change of forest resources in the
their energy source from fuel wood the rest from cow dung district. Totally, 2501.5ha of the forest resources were
and crop residues. The major energy sources of the area converted into agricultural land between 1973 and 2003.
presented in Table 14. Woody biomass accounts the total fuel From the examined results, the extent of land use and land
needed for domestic cooking and heating. Fuel wood is the cover in general and forest cover change in particular was
major source of energy since there is no shortage of biomass radically changed between 1973 and 2003.. Particularly,
fuels for their energy source. According to the survey, type of expansion of agricultural land and decline of both forest cover
domestic appliance that the family use for cooking (wet and as well as shrub land were observed. In addition, the areal
360 Abyot Yismaw et al.: Forest Cover Change Detection Using Remote Sensing and GIS in Banja District,
Amhara Region, Ethiopia

coverage of forest is reduced from time to time. From the total GIS Geographic Information Systems
area of the district about 6044 ha of land was covered with GLCF Global Land Cover Facilities
forest in 1973. But, this figure is declined to 2446.9 ha in the GPS Global Positioning System
year 2003The socio-economic data analysis; anthropogenic MSS Multi Spectral Scanning
factors were identified as major causes for forest cover change NDVI Normalized Difference Vegetation Index
in Banja district. To the alarming rate of population growth RGB Red, Green and Blue
resulted in expansion of agricultural land, demand for fire SPOT Satellite Probatoire d’Observation de la Terre
wood, Charcoal production and for hose construction. This SPSS Software package for social scientists
conditions leads to further depletion of forest resources in the UTM Universal Transverse Mercater
area. Consequently the forest cover reduced which highly
contributed to land degradation in the district. Author’s Contributions
Hence, to preserve the forest resources from further
destruction and to use the forest resources in a sustainable AY, and FZ conceptualized the research problem, designed
manner, farmers should be encouraged to plant fast growing the study, conducted field work, collected data, data entry and
trees on their farm boundaries, homesteads or on degraded analysis.
lands instead of cutting trees from the existing forest, BG and SA were involved in revision of the overall research
introduction of fuel saving stoves instead of using fuel wood activates, data analysis, writing and drafting of the manuscript
in traditional three stone stoves, creating awareness among the for publication.
society regarding to optimum utilization of the forest
recourses and conservation systems by concerned bodies
could play significant role in rehabilitation and minimizing of References
environmental degradation.
[1] ESCAP (1996).Manual on geographic information System, for
Planners and Decision makers. United Nation.
Acknowledgements
[2] EFAP (1993).The challenge for Development. Draft Final
This study would never be completed without the Report Volume II. Addis Ababa, Ethiopia. 78pp.
contribution of many people to whom we would like to [3] Azene Bekele, A., Birnie, A., Tengnas (1993). Useful Trees and
express our gratitude. The administrative kebele's, Shrubs for Ethiopia: Identification, Propagation and
development agents, district agricultural officials, local Management for Agricultural and Pastoral Communities.
guiders, committee leaders and respondent households in each Published By Regional SoilConservation Unit, Swedish
sample kebeles were indispensable for the successful International Development Authority, Nairobi, Kenya.
completion of the fieldwork. We would like also to [4] Rahman, M.M, Csaolovics,E., Koch, B., & Kohl, M.(2006) .
acknowledge people who contribute their knowledge and time Interpretation of TropicalVegetationUsingLand sat ETM+
in data collection and entry processes. Imagery.
[5] Badeg Bishaw (2001).Deforestation and land Degradation in
Nomenclature the Ethiopian Highlands: A strategy for Physical Recovery.
North East African Studies 8.1, 7-25 New York.
Authors declared that they have no competing interest
[6] FAO (2005).GlobalForest Resources Assessment progress
Acronyms towards sustainable Forest Management. P.11-12
AVHRR Advanced Very High Resolution Radiometer
BoA Bureau of Agriculture [7] Singh A. (1989). Digital change detection techniques using
CSA Central Statistics Agency remotely sensed data. International Journal of Remote Sensing,
Vol. 10, 989–1003,.
CSE Conservation Strategy of Ethiopia
ETM Enhanced Thematic Mapper [8] WHO, IUCN, and WWF (1993).Guidelines on the
FCC False color composite Conservation of Medicinal [Link], Gland, Switzerland.
EFAP Ethiopian Forestry Action Program

View publication stats