Delineation of the ''Paleo-Lake Manyara'' in Northern Tanzania using Radar- and Multispectral Remote Sensing
Bachofer, Felix1; Quénéhervé, Geraldine2; Maerker, Michael2; Hochschild, Volker1
1University of Tuebingen, Geosciences, GERMANY; 2Heidelberg Academy of Sciences and Humanities, GERMANY

Motivation: The surroundings of Lake Manyara are focus of several paleo-archeological investigations, since the location is close to the famous Olduvai Gorge where paleo-anthropological findings can be traced back to the homo habilis. In the catchment of Lake Manyara, especially close to the village of Makuyuni, two hominin-bearing sites (0.63 and 0.78 Ma), lots of vertebrate fossils and handaxes from different periods were found (Frost et al. 2012). All findings contribute to the knowledge of human evolution and expansion. For a better integration of these discoveries, a better understanding of the paleo-environment in the region is necessary. The use of remote sensing in combination with topographic analysis for the delineation of paleolakes has been already successfully applied by Gaber et al. (2009) on the Sinai Peninsula. ElSheik et al. (2011) used Remote Sensing, GIS and geophysical methods to delineate a paleolake in northern Darfur.

Study area and paleolake evidence: Lake Manyara (954 m asl.) is located in an endorheic basin in the eastern arm of the East African Rift System in northern Tanzania (Fig. 1). Today it is a shallow soda lake that periodically dries out completely. The Manyara basin is an asymmetrically shaped half graben, with a 200 to 600 m high escarpment in the west and a west dipping monocline in the east (Schwartz et al. 2011). Lacustrine strata known as the Manyara Beds in the west of Lake Manyara and approximately 100 m above today's lake surface define a maximum extent proven by sediments so far, even if some authors assume an even higher lake level and a hydrological connection with Lake Natron and Magadi basins (Schwartz et al. 2011).

The Manyara Beds can be subdivided into a lacustrine grayish lower member (mudstones, siltstones, diatomites, marls and tuff) which was sedimented between 1.03 and 0.633 Ma and a fluvial and terrestrial up to 13 m thick reddish brown upper member (siltstones, mudstones, conglomerates and breccias) sedimented between 0.633 until 0.44 (0.27) Ma. The transition between both members is in most sections marked by a distinct tephra layer (Ring et al. 2005). The sections are best exposed close to the town of Makuyuni, where the sediments are partly overlain by a thin layer of Holocene soils and caliche and where big gully systems bite into the savanna landscape. The region is affected by minor en echelon step faults (NNE and E). Ring et al. (2005) assume that the eastern shoreline of Lake Manyara did not only migrate westward but also to the south due to the southward movement of faulting and volcanism in the Middle and Late Pleistocene.

Fig.1: The Lake Manyara Basin

Casanova and Hillaire-Marcel (1992) documented more recent evidence for paleo-shorelines. They found stromatolites in the northwest of today's lake extent useful to provide information about the paleoclimatic fluctuations in the East African rift system. They dated three generations of Late Pleistocene and Holocene stromatolites 20 m above today's lake level.

Methodology: For the delineation of the paleo-shorelines we applied radar and optical remote sensing, as well as terrain analysis methods to contribute to a better understanding of the lakes development and extent. SRTM-X DEM (30 m) data covering the nearly the whole Manyara Basin were combined with SRTM-3 DEM (90 m) to achieve a coverage of the whole basin. We applied a morphology preserving multidirectional Lee Filter for noise and artifact reduction. We processed ICESat altimetry data (15 overpasses over Lake Manyara in 2003 – 2009) to calculated a mean lake level of 954.25 m (EGM96, stdv. 35 cm), which we could use to correct the absolute height.

We used the Next ESA SAR Toolbox (NEST) for the processing of ALOS PALSAR (Level 1.1; Polarization: HH/HV; Date: 24.05.2008 and 15.07.2010) and TerraSAR-X Stripmap (Level 1B; Polarization: HH; Date: 02.09.2011 and 13.09.2011) scenes. Each scene was terrain corrected and backscatter intensity (sigma nought) was processed. While the paleo-shorelines are hardly noticeable in optical remote sensing images, they are highlighted by their intense backscatter (Fig. 2). To extract the linear paleo-shorelines and terraces a canny filter was implemented in a Python-script. The corresponding heights where extracted from the DEM so that it was possible to distinguish between different levels and to trace them over the image.

Fig. 2: Paleoshorelines from TerraSAR-X

For the delineation of the sediments of the pleistocene Manyara Beds a different approach was necessary. After the collection of field reference, multispectral ASTER scenes (23.08.2006; VNIR + SWIR) were used to calculate mineral indices using relative band depth techniques. Together with a height tracing algorithm applied on the DEM, the Upper and the Lower Manyara Beds could be distinguished from younger sediments and soil with similar spectral properties and absorption features.

Central Conclusions: With the combination of DEM-analysis and multispectral band rationing techniques it was possible to extract the maximum through outcropping visible extent of the Lower Manyara Beds and therewith the a preliminary maximum lake level which lies more than 27 km west of today’s shoreline. More detailed information can only be provided by drillings and not yet detected gully and river sections. Radar images proofed to be very useful in detecting paleo-forms which could not be delineated so detailed by optical remote sensing techniques and hardly by field work because of their spatial extent and inaccessible locations.

CASANOVA, J. & HILLAIRE-MARCEL, C. (1992): Chronology and paleohydrology of late Quaternary high lake levels in the Manyara basin (Tanzania) from isotopic data (18O, 13C, 14C, ThU) on fossil stromatolites. Quaternary Research, 38, S. 205-226.
ELSHEIKH, A. et al. (2011): Geology and geophysics of the West Nubian Paleolake and the Northern Darfur Megalake (WNPL–NDML): Implication for groundwater resources in Darfur, northwestern Sudan. Journal of African Earth Sciences, 61, S. 82-93.
FROST, S. R. et al. (2012): Refined age estimates and Paleoanthropological investigation of the Manyara Beds, Tanzania. Journal of Anthropological Sciences, 90, S. 1-12.
GABER, A. et al. (2009): Delineation of paleolakes in the Sinai Peninsula, Egypt, using remote sensing and GIS. Journal of Arid Environments, 73, S. 127-134.
RING, U. et al. (2005): Kinematic and sedimentological evolution of the Manyara Rift in northern Tanzania, East Africa. Geological Magazine, 142, S. 355-368.
SCHWARTZ, H. et al. (2012): Geochronology of the Manyara Beds, northern Tanzania: New tephrostratigraphy, magnetostratigraphy and 40Ar/39Ar ages. Quaternary Geochronology, 7, S. 48-66.