1. Detection of trends and teleconnections of
stream flows with large scale climate signals
in the Omo-Ghibe River Basin, Ethiopia
Mekonnen Adnew & Woldeamlak Bewket
Addis Ababa University
Department of Geography & Environmental Studies

2. Contents
Introduction and rationales
Objective
Data and Method
Findings
Conclusion

3. Introduction and rationales
 There is a general scientific consensus on the intensification of the earth’s
hydrological cycle due to global warming
 The change is very likely on extreme hydrological variables
 Climate change could alter the magnitude, frequency, timing, duration and the
natural variability of extreme hydrological events (Burn and Hag Elnur,
2002)
 The natural river-flow regimes can be also altered due to anthropogenic
activities
 The long-term changes could have significant effects in the designs of water
engineering structures and watershed development programs

4. Cont…
 Observational evidence is also very important to address the uncertainty in
climate model scenarios at the local scale (Hannaford and Marsh, 2006)
 On the other hand, the lower part of the study area (Omo-Ghibe River
Baisn) one of the drought and flood prone areas in the country
 There are also huge hydroelectric (5 hydro power plants) and sugarcane
irrigation projects (about 600, 000 ha) which are very sensitive to extreme
hydrological events
 Very few studies covered this part of Ethiopia (Demissie et al., 2013 and
Worku et al., 2014)

5. Objectives
The objectives of this study are:
 to identify possible changes in the hydrological regimes with emphasis to
extreme events
 to explore the streamflow associations with global SSTs

6. Fig. 1 The study area
Methodology
Study area

7. Data and Method
 Daily streamflow data for 15 gauging station for two time periods 1972-
2006 and1982-2007 was employed
 Indices of important SST modes that are known to affect the climate of
Africa (see Rowell, 2013) and Ethiopia in particular (Gissila et al., 2004;
Korecha and Barnston, 2007; Segele et al., 2009a, b; Diro et al., 2011a)
Equatorial east Pacific (Nino/ENSO)
Central Indian Ocean (CindO)
Indian Ocean Dipole (IOD)
Equatorial east Atlantic (EqEAtl)
Tropical Atlantic Dipole (TAD)
 All the daily flow data were cross-checked with the data from the nearby
stations for any possible errors

8. Table 1. Hydrological indices used to represent extreme flow regimes
Index Description Unit
AMS Annual maximum daily flow series m3/s
MAX7day Kiremt maximum 7-day flow volume m3/s
MAX30day Kiremt maximum 30-day flow volume m3/s
POT1Mag Peak-over-threshold magnitude on average one event/year m3/s
POT3Mag Peak-over-threshold magnitude on average three event/year m3/s
POT1Fre Peak-over-threshold frequency on average one event/year day
POT3Fre Peak-over-threshold frequency on average three event/year day
MIN1day Annual minimum daily flow m3/s
MIN7day Annual minimum 7-day flow volume m3/s
MIN30day Annual minimum 30-day flow volume m3/s
MIN10P Number of days exceeded the 90th percentile of the record period days
ANNmean
Wet and dry
Annual mean flow volume
For wet and dry seasons
m3/s
m3/s

9. Cont…
 Long-term changes were examined using the non-parametric Mann-
Kendall’s test and Sen’s slope estimator using MAKESENS
 Pre-whiten method You et al. (2002b) and Petrow and Merz (2009) was
used to remove the serial correlation from a data set prior to apply MK
trend test

10. Results
Trend results for mean annual and seasonal flow
events
 There are no clear and systematic trends both in annual and
seasonal flow volumes both for 1972-2006 and 1982-2006

12. Trends for high-flows/flood events
There is no clear emergent trends for indices of flood events

13. Trend results for low-flow events
There is a tendency towards increasing trends

14. Streamflow-to-SSTs relations
 Indices of SSTs from 5 Oceanic regions that are known to affect the climate
of the Ethiopian summer season were correlated against streamflow indices
for the period 1972-2006

15. Streamflow
index
Sing of
correlations
CindO OID ENSO EqEAtl TAD
July Positive 0 3 0 2 0
Negative 5 2 5 3 5
Agust Positive 2 0 0 2 0
Negative 3 5(1) 5(4) 3 5
September Positive 1 0 0 2 1
Negative 4(1) 5(2) 5(4) 3 4
JAS Positive 0 0 0 1 0
Negative 5 5(1) 5(4) 4 5
MAX7day Positive 2 3 2 3 1
Negative 3 2(1) 3(2) 2 4
MAX30day Positive 0 3 1 2 1
Negative 5 2(1) 4(3) 3(1) 4
POT3fre Positive 1 1 0 2 1
Negative 4(1) 4(1) 5(3) 3 4

16.  There are reasonable associations between the streamflow-SSTs
 Changes in the SSTs and associated changes in local and zonal atmospheric
circulations were reported (Segele et al., 2009; Jury and Funk, 2012)
 There is a change zonal walker circulation towards low level northeast (cool
moist air) and upper level westerly (weakening the TEJ –with upper level
divergence and increase subsidence) over Ethiopia
 An increase in warming and ascending air trends over Indian Ocean associated
with westward Rossbay wave and increase Subsidence
 All these lead towards decreasing rainfall and streamflow
 There is a tremendous change in land use and land cover in the study area
(Worku et al., 2014)
 Overall forestland, woodland, woody grassland & shrub land decrease by 53%
 Cropland and Grassland increase by 83%

17.  LULC changes from FL, WL, WG & SL into CL &GL would be expected
contribute increase runoff due to reduced interception, less infiltration and
lower actual evapotranspiration

18. Conclusion
 There is no clear and systematic trends for most of the streamflow indices
 Trends in streamflow indices did not reflect the effect exerted from changes
in atmospheric circulation and LULC
 However, the findings of this study generally provide information for the
local scale climate change adaptation and watershed management activities

19. Thank you!

20. Thank you!