The Earth Exposed - Salem State University
Download
Report
Transcript The Earth Exposed - Salem State University
Vegetation Change in North Africa
through the analysis of satellite data,
1982 - 2006
Professor Stephen Young
Department of Geography, Salem State College
IGC - 2008, Tunis, Tunisia – August 13, 2008
Vegetation Change in North Africa
OUTLINE
Why this research?
Overview of Data Set & Methodology used.
Overview of where land cover change is
occurring.
Overview of persistent vegetation change.
Overview of the Driving Forces behind the
changes.
Vegetation Change in North Africa
Why
this research?
Global Vegetation Change
From 1982/83 to 1998/99
Why this research?
Have already mapped vegetation change at the
global scale and now focusing on the regional
scale.
Have researched Australia, East Asia, South Asia,
and Southwest Asia – now looking at North Africa.
Looking for strong temporal signals of
PERSISTENT vegetation change.
Trying to understand those signals in context with
global environmental change.
Why this research?
In particular
Colleagues at this conference may be
able to help me understand the
vegetation changes which are occurring
in North Africa.
Data Used
GIMMS AVHRR land cover data
NOAA / NASA AVHRR data – 8 km resolution
Coarse resolution picks up broad change
Normalized Difference Vegetation Index
(NDVI)
(NIR - red) / (NIR + red)
High correlation with photosynthesis
Data Pre-Processing
GIMMS AVHRR data
Base data: NDVI 15-day MVC
Re-composited: Monthly Max Composites,
January 1982 – December 2003
Re-composited:
Annual Average Composites
Re-composited endpoints
(1982 + 1983) ÷ 2 = 1982/83
(2002 + 2003) ÷ 2 = 2002/03
Global Scale NDVI GIMMS data
Annual Average 1987
Desert Pixels & All Global Land Pixels
6
5
Percent
4
3
2
1
0
-1 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 0 1 2 3
-2
Years
Global Land Cover
Sahara 2
Linear (Global Land Cover)
Linear (Sahara 2)
Data Used
SPOT VEGETATION data (1km)
SPOT IMAGE – 1 km resolution
Finer resolution than GIMMS
Normalized Difference Vegetation Index (NDVI)
(NIR - red) / (NIR + red)
High correlation with photosynthesis
Data Pre-Processing
SPOT VEGETATION data
Base data: NDVI 10-day MVC
Re-composited: Monthly Max Composites,
Re-composited: Annual Average Composites
April 1998 – Dec 2006
1998 used Jan, Feb, March of 1999
Re-composited endpoints
(1998 + 1999) ÷ 2 = 1998/99
(2005 + 2006) ÷ 2 = 2005/06
SPOT VEGETATION NDVI
Annual Average 2001
SPOT VEGETATION
200
180
160
Scaled NDVI
140
120
100
80
60
40
20
0
1998
1999
2000
2001
2002
2003
Years
All Africa
Sahara
Vegetation
2004
2005
2006
Methodology
Pairwise Comparison (Univariate Differencing)
Percent Change
GIMMS: 2002/03 minus 1982/83 = change
SPOT: 2005/06 minus 1998/99 = change
(2002/03 minus 1982/83) / 1982/83 = Percent Change
(2005/06 minus 1998/99) / 1998/99 = Percent Change
Profiling
Graphing change through the period
Results
All Africa Change
GIMMS: 1982 to 2003
Table – All Africa Change
Classification (% Change)
Major Decrease (decrease > 20%)
Decrease (decrease 10 to 20%)
Minor Decrease (decrease 5 to 10%)
Little Change (between -5 and +5%)
Minor Increase (increase 5 to 10%)
Increase (increase 10 to 20%)
Major Increase (increase > 15%)
All Land Pixels
# of Pixels
3895
22426
60269
343420
72695
70061
33456
606222
% of Land Pixels
0.1%
3.8%
9.9%
56.9%
12.1%
11.6%
5.6%
100%
North Africa Change
GIMMS, 1982 to 2003
Table – North Africa Change
Classification (% Change)
# of Pixel
Major Decrease (decrease > 20%)
502
Decrease (decrease 10 to 20%)
2081
Minor Decrease (decrease 5 to 10%)
9173
Little Change (between -5 and +5%)
125815
Minor Increase (increase 5 to 15%)
4645
Increase (increase 10 to 20%)
3067
Major Increase (increase > 15%)
1710
All Land Pixels
146993
% of Land Pixels
0.1%
1.4%
6.2%
85.6%
3.2%
2.1%
1.2%
100%
Profiles of Change from 1982 to 2003
Profiles of Change from 1982 to 2003
Vegetation Change in North Africa
GIMMS Data - 1982 - 2003
Change is highly variable year to
year
Few Persistent Changes
All Africa Change
SPOT: 1998/99 to 2005/06
All Africa Change
SPOT: 1998/99 to 2005/06
North Africa – Areas of Decline
North Africa – Areas of Decline
160
150
140
Scaled NDVI
130
120
110
100
90
80
70
60
1998
1999
2000
2001
2002
2003
Years
Series1
Series2
Series3
2004
2005
2006
North Africa – Areas of Decline
160
150
140
Scaled NDVI
130
120
110
100
90
80
70
60
1998
1999
2000
2001
2002
2003
Years
Series4
Series5
2004
2005
2006
North Africa – Areas of Decline
North Africa – Areas of Decline
160
150
140
Scaled NDVI
130
120
110
100
90
80
70
60
1998
1999
2000
2001
2002
2003
2004
Years
Series1
Series2
Series3
Series4
Series5
2005
2006
North Africa – Same Areas of Decline
GIMMS Data, 1982/3 – 2002/3
Mesopotamian Wetlands, Iraq
Areas of Decline
GIMMS Data
Wetlands
0.3
0.28
0.26
NDVI
0.24
0.22
0.2
0.18
0.16
0.14
0.12
0.1
82
83
84
85
86
87
88
89
90
91
92 93
94
Years
Wetlands
95
96
97
98
99
0
1
2
3
India
South Interior Karnataka
Annual Integrated NDVI
Trend line:
y = 0.0222x - 39.313
NDVI
5.5
5
5 yr moving average
4.5
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
YEARS
India
South Interior Karnataka
Annual Integrated NDVI
Trend line:
y = 0.0222x - 39.313
NDVI
Annual Rainfall
Rainfall in
mm
Trend line :
y = -33.752x + 8499.8
12000
5.5
11000
5 yr moving
average
10000
9000
8000
5
7000
5 yr moving average
6000
5000
YEARS
Years
02
20
00
98
20
19
96
94
19
19
92
90
19
19
86
88
19
19
84
19
19
4.5
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
82
4000
North Africa – Areas of Increase
1 – West Nile Delta, 2 – North Central Algeria, 3 – Atlas Mountains, 4 – South Morocco
North Africa – Areas of Increase
1 – West Nile Delta, 2 – North Central Algeria, 3 – Atlas Mountains, 4 – South Morocco
GIMMS Data – Western Nile Delta
Arc of Irrigation Temporal Profiles
210
Scaled NDVI
190
170
150
130
110
90
70
50
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
0
Years
Region - 1
Region - 2
Region - 3
Region - 4
Region - 5
1
2
3
3-year Running Average
190
Scaled NDVI
170
150
130
110
90
70
50
82
83 84
85
86
87
88 89
90
91
92 93
94
95
96 97
98
99
0
Years
Region - 1
Region - 2
Region - 3
Region - 4
Region -5
1
2
3
Conclusions
Unlike other parts of the world, patterns of
declining photosynthesis (NDVI) in North
Africa are highly variable – with few areas
of persistent change.
Climate must be the major driver of
change for most of the region.
Conclusions
Areas in Morocco indicated in the
literature as suffering from desertification –
also have patterns of great variability.
The influence of human activity is hidden
by strong climate variability.
Conclusions
There are some regions in North Africa which do
show signs of persistent increase.
.
Like other regions of the world,
the main driver
of persistent increases in photosynthesis (NDVI)
is agriculture and in particular related water
resources.
Future research will focus on analysing
precipitation data for North Africa with NDVI
data.
Thank You