application of remote sensing for the assessment of drought in somalia – case study in puntland...
TRANSCRIPT
APPLICATION OF REMOTE SENSING FOR THE ASSESSMENT OF
DROUGHT IN SOMALIA –Case Study in Puntland
Ambrose Oroda
Ronald Vargas, Simon Oduori and Christian Omuto
12th June, 2007. Nairobi, Kenya
BACKGROUND
Lack of proper natural resources management can result in:
• Significant reduction in the production potential of natural resources.
• Environmental degradation, particularly of rangelands and range resources.
• However, anthropogenic impacts are often worsened by natural phenomena such as droughts.
DROUGHT
• Drought refers to a period of months or years when an area or a region experiences a deficiency in its water supply due to consistent below average precipitation:
– Meteorological drought – Consistently below average precipitation.
– Agricultural droughts – insufficient water supply for crop production or ecological production of the range.
– Hydrological drought – below average water reserves in the sources such as the aquifer system.
• Impacts of droughts can, however, be worsened by anthropogenic mismanagement of natural resources.
• There is, therefore, a need to develop a methodology that can provide reliable and rapid information decision support in environmental management.
DROUGHTS CONT.
Consequences of drought in the context of the study area
• Decreased water supply• Increase in livestock diseases• Livestock deaths• Crippling of the economic sector resulting in loss of
income, increased poverty, etc.• Environmental degradation leading to mass migration –
“environmental refugees”• Famines and subsequently malnutrition• Social instability:
Impacts of drought can also be looked at sector-wise
Significance of this study
• Droughts like many other natural disasters have been on the increase.
• The economic, social and environmental costs of droughts have increased dramatically (Donald Wilhite, 2000).
• There has been prolonged periods of dry spells in the recent years over sub-Saharan Africa (Hare and Ogallo, 1992).
• Methodologies and seasonal forecasts have always been lacking – and this is made worse where there are no central management systems (Thomas Downing and Richard Washington, 2002).
• Remote sensing has been found to be a useful tool in accurate and timely assessment and monitoring of environmental conditions including droughts.
Objective
• This study was to assess and test applicability of remote sensing in measuring the phenological dynamics and physiognomic variability of vegetation in two study areas of Somalia as an indicator to assessing drought events.
METHODOLOGYSTUDY SITES
METHODOLOGYSTUDY SITES
METHODOLOGY CONT.
• Remote sensing– Acquisition of Multi-temporal, multi-spectral and multi-resolution
remote sensing products (between 1973 and 2005)
– Using Idrisi software performed a temporal-spatial analysis of vegetation to show impacts of settlements on physiognomic conditions
– WINDISP 5 software was used to generate NDVI data from NOAA-GAC data (1982 and 2002) to assess drought events
– VEDAS software was used to generate NDVI data from SPOT and MODIS images.
METHODOLOGY CONT.
• Fieldwork– Physiognomic assessment of the study area
to verify different land cover classes - Ground-truthing.
– Questionnaire administration to gather:• Indigenous knowledge• Historical memory of environmental conditions
including years of drought.
RESULTS_1 – Temporal-spatial assessment of drought conditions
Very Severe drought
Less severe drought
Limited or no drought
VI = (NDVIi – NDVImin)
(NDVImax – NDVImin)
Vegetation Condition Index as a measure of drought severity
-40
-20
0
20
40
60
80
100
1 6 11 16 21 26 31 36 41 46
VI (
%)
VI
RESULTS_1 CONT.
RESULTS_2Assessment of phenological dynamics and physiognomic
variability, and other environmental conditions
NDVI AND RAINFALL ESTIMATES FOR THE STUDY AREAS FROM NOAA
0
2
4
6
8
10
12
14
16
18
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Dekads
RF
E (
mm
)
0
0.05
0.1
0.15
0.2
0.25
ND
VIg
Northern Study Area
Southern Study Area
Northern Study Area
Southern Study Area
Results_3
Monthly Potential Evapotranspiration derived from NOAA
0.00
50.00
100.00
150.00
200.00
250.00
Jan Feb Mar Apri May Jun Jul Aug Sept Oct Nov Dec
Tim e (m onths)
Pote
ntia
l Eva
potra
nspi
ratio
n (m
m)
Northern Study Area
Southern Study Area
Measured PET for 4 Stations neighbouring the study areas
0
50
100
150
200
250
Jan Feb Mar April May June July Aug Sept Oct Nov Dec
Time (months)
PET (
mm
)
EIL
GALCAYO
LAS-ANOD
QARDO
The results show that:
•The southern study site receives more rainfall than the northern study area.
•Similarly the southern study has more vegetation cover than the northern one.
•There is very close correlation between the measured ET and ET derived using RS.
RESULTS_2 & 3
RESULTS_4: Annual NDVI vis-à-vis LTM - NOAA
Annual NDVI Values compared with the Long Term Mean - derived from NOAA-GAC
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Years
ND
VI V
alu
es
Annual_NDVI
LTM
Annual NDVI Values compared with the Long Term Mean
0.1
0.12
0.14
0.16
0.18
0.2
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Years
ND
VI V
alu
es
Annual_NDVI
LTM
Northern study area Southern study area
RESULTS_5 – Assessment of drought vis-à-vis years of reported droughts – NORTHERN STUDY SITE
Annual NDVI Values compared with the Long Term Mean - derived from NOAA-GAC
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Years
ND
VI
Va
lue
s
Annual_NDVI
LTM
Years of reported droughts from the field
0
2
4
6
8
10
12
14
16
18
1948 1950 1953 1961 1965 1970 1973 1975 1980 1982 1984 1990 1992 1994 1997 1999 2001 2003 2005 2007
Years
Fre
qu
en
cy
of r
ep
ortin
g
RESULTS_6 – Assessment of drought using RS vis-à-vis years of reported droughts – Southern Study site
Annual NDVI Values compared with the Long Term Mean
0.1
0.12
0.14
0.16
0.18
0.2
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Years
ND
VI V
alu
es
Annual_NDVI
LTM
Years of drought as reported from the field - south
0
5
10
15
20
25
30
35
1964 1973 1974 1975 1984 2000 2003 2004 2005 2006
Year of drought
Fre
qu
en
cy
RESULTS_7 – Annual SPOT-NDVI vis-à-vis Years of reported drought
Annual NDVI Values Compared with the 8-Year Mean
0.075
0.095
0.115
0.135
0.155
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Years
ND
VI
Va
lue
s
Annual NDVI
LTM
Annual NDVI Values compared with the 8-Year Mean - Sample Point 93 - southern site
0.11
0.13
0.15
0.17
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Years
ND
VI
Val
ues
Annual Mean
LTM
Years of reported droughts from the field
0
2
4
6
8
10
12
14
16
18
1948 1950 1953 1961 1965 1970 1973 1975 1980 1982 1984 1990 1992 1994 1997 1999 2001 2003 2005 2007
Years
Fre
qu
ency
of
rep
ort
ing
Years of drought as reported from the field - south
0
5
10
15
20
25
30
35
1964 1973 1974 1975 1984 2000 2003 2004 2005 2006
Year of drought
Fre
qu
en
cy
RESULTS_5, 6 & 7: Show that:
• There was great correlation between the field measured information and remote sensing information.
• Field results (local knowledge) show droughts in 1974-1976, 1984, 1990 – 1992 and 2001 – 2004.
• Drought periods were significantly identified with RS in the years 1988 – 1994 and 2003 – 2004 using SPOT NDVI.
RESULTS_6-Comparative view of SPOT NDVI images to assess droughts
NDVI for May 2000 –Non drought year
Severe drought
No drought
Drought
NDVI for May 2004 –Drought year
Severe drought
No drought
Drought
RESULTS_7: Comparative view of SPOT NDVI Images
Severe drought
No drought
Drought
NDVI for June 2000NDVI for June 2004
RESULTS_8
CONCLUSIONS AND RECOMMENDATIONS
• There was great correlation between the field measured information and remote sensing information.
• Remote sensing has good potential in the assessment of drought and can be employed in the accurate and near-real time assessment of environmental conditions.