Abstract:
Monitoring and study of lakes has been hampered by hydrological data gaps. For
instance, Lake Ziway is being highly affected by human interference that causes a
noticeable change on characteristics of the lake over the recent decades. However, the
magnitude of the change is not quantified well due to lack of regular and up-to-date lake
bathymetric data. Therefore, additional data sources other than traditional data sources
are needed to evaluate the change in lake characteristics. In this thesis, the combined use
of multisource satellite products was evaluated for monitoring storage change of the lake.
The approach started with extraction of multi-temporal lake surface areas after
processing 72 Landsat ETM+/OLI optical satellite imageries. Also, timeseries data of
211 water level measurements was obtained from two satellite altimetry products (GRLM
and DAHITI). Next, 15 coinciding pairs of remote sensing based lake water level and
surface area were related using regression. Lastly, equation of the developed water level
area curve was integrated and used to estimate lake water volume variation. Accuracy
evaluation was conducted on three water extraction indices to identify the best performer.
The lake surface area was extracted with an overall accuracy of 91.0% to 99.2% and
kappa coefficient (κ) of 0.82 to 0.98. The water level which was retrieved from mean of
the altimetry products was compared to in situ measurement and yielded R2=0.97,
MAE=6.59 cm and RMSE=8.52 cm. This study unveils that Lake Ziway has been
experiencing water level reduction, surface area shrinkage and storage loss from 2009 to
2018 at a rate of 4.56 cm/year, 0.147 km2/year and 33.3 Mm3/year respectively which
accounts 2.05% total storage of the lake annually. The decadal reduction of the lake
volume is largely affiliated to uncontrolled water abstraction and to some extent by the
erratic pattern of climatic variables. Finally, it can be concluded that the combined use of
satellite imagery and altimetry is viable approach for continuous, frequent and up-to-date
monitoring of lake water volume variation and for long-term change detection. Therefore,
it can be suggested that the data and methods used in this study could be adopted to
evaluate temporal storage change of lakes having shortage of hydrological data but
having appreciable coverage from satellites.
