ANALYSIS OF LAND AND VEGETATION COVER DYNAMICS
USING REMOTE SENSING & GIS TECHINIQUES,A CASE
STUDY OF NECHISAR NATIONAL PARK
Abstract
The research aims to analyze the trend of land and vegetation cover dynamics over the period from 1976, 1986 2000 and 2007 thus examine the conservation status of the area and generate
up-to-date land cover map. Information is extracted from various Satellite images of multidated Landsat, ASTER and MODIS images. The Landsat images are the basic remote sensing data to generate the thematic maps which are further analyzed to show the cover dynamics in the park for 24years. All datas from the satellite images are processesed and analyzed using digital image processing techniques. Besides, different vector data are extracted from the images as well as other thematic maps. MODIS-NDVI images are analyzed for the different land cover classes and each vegetation cover seasonal response is compared for the year 2000 and 2005.
The land cover classes identified in the study area from 1976, 1986, 2000 and 2007 are water body, riparian and ground water (GW) forest, wood land, dense bush land, bushy shrubbed grass land, open grass land, degraded grass land, cultivated land, swamp vegetation and bare
land. Rate of land cover change and fragmentation of habitat were discussed for the different
land cover classes. Rate of land cover change, fragmentation index and land cover conversion
matrix clearly shows the dynamics of the different cover classes has happened for the past decades and generally the park conservation status is found to be poor. Bush encroachment in the study area is a major challenge to the park particularly for the grass land and overgrazing
on the Nechisar plain has caused expansion of invasive plants erosion and land degradation.
The community livelihood dependency both in the rural and urban setting is concluded and discussed as a challenge to the park from biodiversity conservation point of view.
Key Words: Land cover dynamics, National park, Vegetation cover, Remote sensing and GIS,
Habitat fragmentation, degradation, biodiversity conservation.
2. 1.Introduction
• deforestation, grazing cattle, human habitation
and over fishing in the park have caused severe
stresses and degradation of park ecosystems,
leaving the sustainability of NSNP resources in
question.
(Alison M. Jones, 2005)
• The park had been excessively encroached by
invasive plant species and the small but rare
groundwater forest had been subjected to
unsustainable illegal harvesting
(Report of African Parks foundation, 2005)
1.1.Statement of the problem
3. Freeman(2006) Natural Resource Challenges of NSNP
NSNP
Forest
Springs
Wild Animals
(land-based)
Grass land
Wild animals
(lake-based)
Fish
Swamp,Sokke
Grazing land
Cut for firewood
Habitat being destroyed
over-grazing
Nowhere for towns
people to graze cattle
Being
over-cut
Depleted by over-fishing
Fish breeding
area destroyed
Risk to numbers caused by lack of fish
Potential
use by
investors
Unknown
impact on
forest
Unknown
impact on lake
and resources
Farming
land
Land degraded
4. 1.2.Objectives of the study
1.2.1 General objective :
to analyze the trend of land and
vegetation cover dynamics for the
period starting from 1976-2007 thus to
assess and examine the conservation
status of the area and generate up-to-
date land cover map
5. • To assess the trends of land and vegetation cover
dynamics in the study area during the period from
1976, 1986 and 2000,
• To generate up to date land cover map for the year
2007,
• Studying the conservation status of NSNP focusing
on invasive plants encroachment and land
degradation,
• To show the annual biomass growth response,
1.2.2.Specific objective
6. 2.Description of the study area
2.1. Location •Located in Africa’s
Great Rift Valley,
575km southwest of
Addis Ababa,
• Diverse Habitat
-Open grass land
-Ground water &
riparian vegetation
-Lake shores
-Dense Bush land
-Acacia woodland
10. 2.4 Topography and Hydrology
Nechisar plains are bounded
by:North lake Abaya ,South
and southwest lake Chamo
Abaya receives 86.484mm3
of discharge annually, while
lake Chamo receives only
29.250mm3
Kulfo –flows from North of
Arebamench & runoff joins
lake shores of Chamo-
Irrigation at chamoleto
Arebamench(Traditionaly 40
springs)-GWF
Seremele- perennial & runs
along the eastern most
cultivated fertile flood
plains,
11. 2.5 Socioeconomic
Rural:Weredas adjacent to the
national park livelihood are
farming, fishing, Tourism & petty
trade
Communities inside the park
Kore-Live in Amaro Mountains &
cultivate the Sermele floodplain
Guji- are nomadic pastoralists
grazing approximately 4000 – 5000
heads of stock in the park
Urban :Arebaminch 168,172ha
estimated area,Population 232,000
with 50.8% male and 49.2% female
proportion (SNV,2006).
Livelihood: Farming crops:maize,
sorghum, Fruits:bananas , mangos
,apples and pears,FishingFishing
--TourismTourism service (hotels and tour
operators), Trade ,Employment
12. 3.Materials and Methods
• Software & program :ArcGIS 9.2, ERDAS
Imagine 9.1, Map source,
• GPS-receiver (Garmin),
• Digital Camera,
• Topo map of the study area (1:50,000 scale)
• Multi dated Satellite images (Landsat images
1976-MSS,1986-TM and 2000-ETM+, 2006-
ASTER)
• MODIS NDVI-image(2000&2005)
• SRTM (90m contour interval)
3.1.Materials
13. Landsat
MSS
(1976)
Landsat
ETM+
(2000)
Landsat
TM
(1986)
Image processing (Remote sensing techniques)
Image rectification
Image enhancement
Image classification
Unsupervised
Supervised
Image interpretation
Accuracy assessment
ASTER image
2006
-Topo map
(1:50,000)
-Secondary data
and previous land
use/cover maps,
-Google Earth
image, KML files
-SRTM
-Field ground
verification
-Senior park staff
and Scout
interview
-Community
discussion
-Workshop
proceedings
Out puts
Land cover change comparison thematic maps
Up-to-date land cover map of NSNP
Statistical results of cover change dynamics(charts,
graphs, index values)
Impacts of the cover dynamics and the conservation
status of the national park
GIS data integration and statistical analysis
-Cover change comparison
-Land Cover change rate & Conversion matrix
-NDVI Seasonal vegetation cover response
-Landscape fragmentation index
MODIS-NDVI
Images for 2000
and 2005
3.2.Methodology
14. 1. Identifying Land use/cover data sources
Bolton(1970) distribution of main land cover types
Topo map,1979-1985
FAO/UNDP(1992), RVLB LU/LC map, Assistance to
land use planning project
2.Sattelite image visual interpretation (TCC/FCC Band
combination, select training site pixels, Google
earth image KML files, field inventory)
3.Unsupervised & Supervised cover classification and
accuracy assessment
4. Land cover classification and result interpretation,
statistical analysis (cover change rate, conversion
matrix, fragmentation index)
3.2.1.Data analysis and Organization
15. 4.Results and Discussion
-Unsupervised classes
of Landsat 2000 gives
the classes which are
verified at field
-10 major land cover
classes are identified
and training classes are
set for each year
satellite image
-Accuracy assessment
for each year cover map
conducted
4.1.Unsupervised classification and select basic land cover units
16. Land and vegetation cover classes identified in NSNP for this study
REPARIAN & GW FOREST1 ACACIA WOOD LAND2
BUSHY SHRUBBED GRASSLAND3 OPEN GRASS LAND4
19. 4.2.Land Cover classification thematic maps
1%
4%
2% 25%
11%
11%
11%
6%
30%
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubbed grass land
Open grass land
Open bushed grass land
Cultivated land
Swamp vegetation
Water body
2%
26%
1%
10%
13%
31%
7%
6%
4%
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubbed grass land
Open grass land
Open bushed grass land
Cultivated land
Swamp vegetation
Water body
20. 1%
5%
12% 14%
12%
14%
6%
6%
30%
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubbed grass land
Open grass land
Open bushed grass land
Cultivated land
Swamp vegetation
Water body
2%
26%
1%
10%
13%
31%
7%
6%
4%
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubbed grass land
Open grass land
Open bushed grass land
Cultivated land
Swamp vegetation
Water body
21. Land cover classes for 2007
4798, 10%
1819, 4%
3231, 7%
10124, 21%
7610, 16%
3549, 7%
2532, 5%
14531, 30%
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubbed grass land
Open grass land
Degraded grass land
Cultivated land, Swamp and bare
Water body
Accuracy assessment
The final up to date cover map is
generated using the information of
past classification , field inventory
and accurate training selection
GCP were used to validate the
Landsat and ASTER image
classification accuracy assessment
and the error matrix gives average
results:
78% accurate for 1976 Landsat-MSS
image,
80% accurate for 1986 Landsat-TM
image,
72.22% accurate for 2000 Landsat-
ETM+ image and
95% accurate for 2007 Land cover
map using ASTER image
Up- to date Land cover map & Classification Accuracy assessment
22. 1.Cover change comparison for 1976,1986 ,2000
Land cover comparison
0
2000
4000
6000
8000
10000
12000
14000
16000
1 2 3
years(1-1976,2-1986,3-2000)
Area(ha)
Riparianand GW forest
Wooded grass land
Dense bushland
Bushyshrubed grass land
Opengrass land
Openbushed grass land
Cultivated land
Swamp vegetation
Waterbody
Uneven distribution of land cover class for the
spatial and temporal dynamics in the past 24
years
From 1976 to 1986,
-a total loss of
grassland 562.6ha/year
-Bushy shrubbed
grass land has
increased at a rate of
714.1ha/year,
From 1986-2000
-a drastic loss of the
swamp vegetation at a
rate of 92ha/year
-Riparian and GW
forest have shrunk at a
rate of 3.34ha/year
-cultivated land
expand during the two
periods at a rate of
1.12ha/year and
12ha/year
23. Cover change comparison for 1976,1986 ,2000 & 2007
Land cover class comparison
0
2000
4000
6000
8000
10000
12000
14000
16000
RiparianandGW
forest
Woodland
Densebushland
Bushyshrubbed
grassland
Opengrassland
Barebushed
grassland
Cultivated,Swamp
andbareland
Waterbody
Land cover class
Area(Ha)
1976
1986
2000
2007
-Progress in the Riparian & GW forest, Open grass land,Bushhy shrubbed grass
land
- Cover decline in cultivated land ,Bare bush grass land, Wood land, Dense bush
land
-Water body is constant and spatial changes are insignificant
24. 2.Rate of Land cover change
+6.56
-46.95
+82.5
+197.25
-265.7
+122.05
+25.35
-147.95
-7.025
Net change rate
121.12Water body
-92-1.9Swamp vegetation
12144Cultivated land
54.78394.5Bare bushed grass land
31.2-562.6Open grass land
-470714.1Bushy shrubbed grass land
275-224.3Dense bush land
129-424.9Wood land
-60.8546.8Riparian and GW forest
1986-20001976 - 1986Land cover class
Rate of land cover change (ha/year)
0
2000
4000
6000
8000
10000
12000
14000
16000
0
2000
4000
6000
8000
10000
12000
14000
16000
Series2 1986 3559.2 1104.58 2873.42 12367.7 6328.4 4881.3 721.197 1892.22 14553.1
Series3 2000 2707.96 2910.78 6726.57 5787.85 6765.12 5648.31 2417.42 595.298 14723.7
Series1 1976 3091.1 5354.35 5116.53 5226.02 11954.4 936.037 280.389 1911.06 14541.9
Cover
classificati
on
Riparian
and GW
forest
Wooded
grass land
Dense
bush land
Bushy
shrubed
grass land
Open
grass land
Open
bushed
grass land
Cultivated
land
Swamp
vegetation
Water
body
Area(ha)
25. 3.Land cover change Conversion matrix
Conversion Matrix for 1976 and 1986
0
1000
2000
3000
4000
5000
6000
R
iparian
and
G
W
forest
W
ood
land
D
ense
bush
land
Bushy
shrubed
grass
land
O
pen
grass
land
Bare
bushed
grass
land
C
ultivated
land
Swam
p
vegetation
W
aterbody
Land cover types
Area(ha)
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubed GL
Open grass land
Bare bushed GL
Cultivated land
Swamp vegetation
Water body
Conversion matrix for 1986 and 2000
0
1000
2000
3000
4000
5000
6000
R
iparian
and
G
W
forest
w
ood
land
D
ense
bush
land
Bushy
shrubed
grass
land
O
pen
grass
land
Bare
bushed
grass
land
C
ultivated
land
Swam
p
vegetation
W
aterbody
Land cover class
Area(ha)
Riparian and GW forest
Wood land
Dense bush land
Bushy shrubed GL
Open grass land
Bare bushed GL
Cultivated land
Swamp vegetation
Water body
1976 to 1986
Most covers are
converted dominantly:
to dense bush land, bare
bushy grass land
cultivated land and Bushy
shrubbed grass land
1986 to 2000
Open grass land ,swamp
vegetation and Riparian
vegetation taken by mainly
Bushy shrubbed grass
land and Bare bushed
grass land
26. 4.Landscape fragmentation
368036373635Water body9
182423Swampy vegetation8
134846Cultivated and bare land7
226440Bare bushed grass land6
60253341Open grass land5
2111969Bushy shrubbed grass land4
333434Dense Bush land3
223555Wood land2
79147134Riparian and GW forest1
Pa value for
2000
Pa value for 1986Pa value for
1976
Land cover classesNo
The park habitat is highly treated and most cover classes have shown a
fragmentation index value that describes the NSNP is highly disturbed
Fragmentation of ecosystems into small patches can reduce habitat for wildlife
species that require larger, connected patches and introduce predators, parasites,
and competitors (Core national indicator, 2000)
27. TRRM rain fall data (2000-2006)
GW & Riparian
Grass land
Bare land
Water body
Wood land
Swamp veget
Bush land
High Biomas
growth
-April,May
,October &
November
Less biomas
growth
-January to
march,June to
september and
december
4.3.MODIS NDVI Image analysis-to see the temporal biomass growth
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
200
250
300
RainfallinNechSarNP(mm)
2000
2001
2002
2003
2004
2005
2006
Average
2005 NDVI profiles
2000 NDVI profiles
28. 4.4.NSNP natural resource challenges and conservation status
Though natural processes contribute to change in land cover,
the major driving force is human induced land uses (Allen and
Barnes, 1985)
Major ecological challenge faced in the NSNP due to anthropogenic and
natural factors
the surrounding population is currently growing rapidly, consequently
inducing high pressure on the resource base of NSNP
Different studies show Cattle grazing in the Nechisar plain, poaching of wildlife,
illegal fishing and felling of timber from the groundwater forest, farming at Kulfo
and Seremele valley encapsulate the main problems
Socio-economic setting are influenced by the existing environmental
conditions and Political unrest during the early 1990s led to intense
damages in a relatively short period of time
Natural challenges includes Bush encroachment, invasive plant expansion
29. Riparian and GW forest situation
Progress in forest cover
Total cover in the year 2007
4798ha which is highest
compared to previous classes
Supportive measures
-The forest Guarding strategy
implemented by APF-Ethiopia
-Restoration of cultivated areas
at Chamo letto & Seremele
riparian vegetation
-Future afforestation programs
which will be implemented for
Arebaminch town wood supply
- Alternative energy option
project
-Community participation for
park resource management
Control forest
degradation
Sermele forsest restoration
Chamoletto forest degradation
30. Open Grass land situation
Challenges
The GL is degraded heavily where
livestock density is high,Land
degradation & erosion are seen at
overgraze plain
-Malvaceae family Invasive
plant(Abutilon bidentatum ) and
Bush encroachment (mainly
Dichrostachys cinerea, Acacia
mellifera and Acacia oerfeta)
Supportive measures
-17 hectares of GL were cleared of
invasive plant species (APF,2005)
-In 2007 Open GL covers 16% of
the total area, higher than the
1986 and 2000 GL spatial cover
32. 5.Conclusion and Recommendation
In the past three decades as the population around Nechisar
National Park and nearby Arba Minch increases, so do the
Pressure to the park natural resource such as grazing cattle
at Nechisar plain, farming expansion,etc
The results of the thematic maps for land cover classes of
1976, 1986, 2000 and 2007 shows uneven dynamism for the
different land cover classes except the changes for the water
body is relatively insignificant or uniform
There is a tremendous land cover change in the open grass
land cover as compared to other cover classes and it’s
calculated a total loss of grassland, 562.6ha/year over the
10years of 1976-1986
Bushy shrubbed grass land has increased highly for the first
10 years period from 1976-1986 at a rate of 714.1ha/year, and
second period from 1986-2000 by 470ha/year.
33. Habitat fragmentation is also a serious
challenge in the park for the different land and
vegetation cover classes as the broken
patches of land cover results a fragmented
landscape which is unhealthy for the
ecosystem functioning.
Remote sensing and GIS to study Land cover
dynamics could facilitate for a proactive
planning process that enhances the
implementation of sustainable natural
resource management for decision makers,
park managers and other stakeholders.
34. Recommendation
• The Nechisar plain is severely affected by environmental and
anthropogenic challenges thus needs priorities to completely
stop Farming, overgrazing and plan appropriate bush
encroachment controlling strategy.
• The dependency of the surrounding community to the natural
resources of the park has to be a priority agenda and for
successful and win-win solution there should be the integration
of different actors to start participatory park management and
ecotourism projects
• The ground water forest which is unique habitat for the country
has to be protected from any exploitation and investment
projects such as water bottling industries which are already
proposed by investors and this should be discouraged unless
detailed impact assessment is conducted.
35. The situation in the park has shown improvement once APC
Ethiopia is taking the responsibility to manage the national park,
to realize the vision of APC & to set the Nechisar national park a
model to other treated parks in the country all actors
participation is important
Due to current situation there should be a quick response for a
wise decision before the park resource is aggressively
devastated while conflicts among the surrounding communities
arise during transitional periods between APC–Ethiopia,
Regional Governments and the Federal Government.
Finally I recommend for further researchers to study the impact
of vegetation cover dynamics on the situation of endangered
wild life such as Swayenes heartbeets and to conduct a scenario
analysis on the relationships of natural resource degradation
and wildlife extinction.
37. Acknowledgement
• My supervisor Prof. Zerihun Woldu
• HoARENC
• AAU Earth science department
• AAU Biology department,Herbarium center
• African Parks Foundation (APC p.l.c –NSNP)
• GO’s and NGO’s
• Friends
• Family