- Email: [email protected]
Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ)
Quaternary International xxx (2015) 1e6
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ) Frank D. Eckardt a, *, Fenton P.D. Cotterill b, Tyrel J. Flügel c, Beth Kahle d, Marty McFarlane e, Christie Rowe f a
Environmental and Geographical Science, University of Cape Town, Rondebosch, 7701, South Africa Geoecodynamics Research Hub, Department of Botany and Zoology, University of Stellenbosch, Matieland, 7602, South Africa Department of Military Geography, Stellenbosch University, Private Bag X2, 7395, Saldanha, South Africa d Department of Geological Sciences, University of Cape Town, Rondebosch, 7701, South Africa e Marty McFarlane, Bosele Investments (Pty) Limited, Box 66, Maun, Botswana f Earth & Planetary Sciences Department, McGill University, Montr eal, QC, H3A 0E8, Canada b c
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
The Makgadikgadi basin and wider Middle Kalahari region of Botswana and beyond host landforms which have been attributed to Quaternary environmental change, including palaeolake level fluctuations and aeolian activity. Tectonic processes and landforms on the other hand, have mostly been linked to the Okavango graben and associated rift zone (ORZ) to the west of the Makgadikgadi. In this paper we establish the extent of tectonic surface expression associated with the Makgadikgadi Rift Zone (MRZ). We identify a series of parallel, NNE-SSW, normal faults and scarps, expressing horst and graben structures linked to seven major blocks in the northern Makgadikgadi basin, using both Shuttle Radar Topographic Mission (SRTM) and geomagnetic data. Subtle expression of rifting has controlled endorheic drainage topology, replicated regional dune-field patterns and displaced the 945 m palaoelake contour since lake desiccation. These observations underscore the role of neotectonic “piano key” block movement in shaping surface landforms across a large expanse of the Kalahari region. This paper provides the first detailed map and introduction to the Makgadikgadi Rift Zone (MRZ) and its geomorphology. © 2015 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Kalahari Geomorphology Makgadikgadi Rifting SRTM ICESat
1. Introduction This study is centered on subtle neotectonic landforms in the Middle Kalahari, considered terminal branches of the East African Rift System (Scholz et al., 1976), with the Okavango Rift Zone (ORZ), an alluvial fan, being the most documented example (McCarthy and Ellery, 1998; Modisi, 2000; Gumbricht et al., 2001; Kinabo et al., 2007, 2008; Shemang and Molwalefhe, 2011; Bufford et al., 2012; Podgorksi et al., 2013). In this paper we will particularly focus on the Makgadikgadi Rift Zone (MRZ) to the east of the ORZ (Fig. 1). Research emphasis in the MRZ in contrast, has been on Late Quaternary lacustrine sandy ridges, flanking the perimeter of the current dry lake basin (Grove, 1969; Heine, 1978; Cooke, 1980; Mallick et al., 1981; Helgren, 1984; Shaw and Cooke, 1986; Shaw et al., 1997; Ringrose et al., 1999, 2005; Burrough et al., 2009; Ringrose et al.,
* Corresponding author. E-mail address: [email protected] (F.D. Eckardt).
2009; Moore et al., 2012; Riedel et al., 2014) as well as fossil dune forms (Stokes et al., 1997). Although the ORZ and MRZ both include recent and ongoing horst and graben development (Baillieul, 1979a and b; Reeves, 1972), and hosted the same regional palaeolake (Cooke, 1980), it is fair to say that the MRZ has remained comparatively understudied. With this in mind, we set out to identify, map and describe faults and related geomorphic landforms, covering a 40 000 km square area of north-eastern Botswana and western Zimbabwe. The results include the identification of seven major fault blocks, control of drainage channels flowing into the basin, the distribution of palaeo dune topography and the vertical displacement of the sandy ridges associated with the paleo 945 m lake level. 2. Methods Surface traces of faults were mapped using SRTM3 (Shuttle Radar Topography Mission, 3-arc second resolution) version 4 data obtained from the CGIAR-CSI site (Consultative Group on International Agricultural Research e Consortium for Spatial Information,
http://dx.doi.org/10.1016/j.quaint.2015.09.002 1040-6182/© 2015 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
2
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Fig. 1. Introduction of the MRZ. The Middle Kalahari of northern Botswana, home to the Makgadikgadi Rift Zone (MRZ) to the east and the Okavango Rift Zone (ORZ), (Kinabo et al., 2007) to the west, as well as the former 945 m palaeolake outline (light grey shading), extracted from SRTM data. Gidikwe Ridge and associated ridge lines (dotted lines) were obtained from Mallick et al., 1981. Preliminary MRZ faults (thick solid lines) depicted here are taken from Baillieul (1979a and b). Dunes ridges (thin solid EeW lines) were mapped from vegetation stripes in Landsat 7 (Image Source: Global Land Cover Facility).
Jarvis et al., 2008), The unenhanced 90 m DSM (Digital Surface Model) data was side shaded and illuminated with threefold vertical exaggeration, using variable azimuths including 0, 45, 90, 125, 180, 225, 270 and 315 and applying sun elevation of 45 . Faults appeared as continuous or discontinuous straight features, indicative of tectonic fault block movement (Figs. 2 and 3). Fault traces were also delineated using national aeromagnetic data, supplied by the Department of Geological Surveys, Botswana (Fig. 4). The survey was flown at a height of 80 m with a line spacing of 200 m and the data converted to a 50 m grid. A range of derivative and trigonometric filters were applied in order to enhance the shallow and subtle features in the magnetic dataset. Mapping of faults from this datasets relies on a magnetic contrast between units on opposite sides of the fault, or between the fault itself and the rocks that it displaces. A fault may appear as a magnetic anomaly in its own right, presumably as a result of fluid flow and precipitation or stripping of magnetic minerals. While it may be difficult to separate dykes and faults with a high degree of confidence, the mapped faults do not coincide with the prominent Karoo-age swarm dykes (NWeSE) in the extreme SW of the dataset. Approximate Kalahari thickness data was also depicted in Fig. 4. This contour data by Haddon and McCarthy (2005) is suitable for mapping regional trends of Kalahari depth but lacks the resolution to identify relative uplift and subsidence of underlying horst and graben structures. Additionally SRTM was used to generate the 945 m contour, point heights for sandy ridges, resolve stream topology, identified fossil dune ridges and provided the topographic cross-sections (Fig. 3) and terrain view (Fig. 5). It is important to note that while SRTM generally overestimates absolute elevation in southern Africa by approximately 5 m, its relative accuracy however is better than 5 m (Rodriguez et al., 2005). For additional validation purposes we also compared SRTM elevation data against the ICESat (Ice, Cloud, and land Elevation Satellite) laser altimeter points, which has a much greater absolute vertical accuracies (Global elevation data product (GLAH06), Level-1B, laser campaigns: L1A, L2A, L2A, L2C, L3A, L3B, L3C, L3D, L3E, L3F, L3G. L3H, L3I and L3J, release 33)
(Schutz et al., 2005) for the Makgadikgadi basin specifically (spatial domain: Latitude 20.17 to 21.46, Longitude þ24.72 to þ26.61). While ICESat laser coverage is not dense enough to produce digital elevation models, it is well suited for validation purposes, since the 65 m diameter laser altimeter footprints, separated by 172 m at ground level, have returned a vertical accuracy of <2 cm (Fricker et al., 2005). For comparative purposes we ensured that all elevation data used here conformed to the WGS 84 ellipsoid. Fossil dune distribution patterns for the middle Kalahari have traditionally been inferred from the vegetation patterns detectable in Landsat satellite imagery (Mallick et al., 1981). However the vegetation continues to reveal the past dune forms, even after the dune ridges have been degraded or even substantially flattened. While mapping fossil dunes near the Gumare Fault (Fig. 1, McFarlane and Eckardt, 2007) and comparing the vegetation stripes in Landsat imagery against SRTM data, actual dune topography appeared most pronounced at the margins of, or within downfaulted graben and along incised streams, essentially in areas of lowered local base level. It was proposed that the action of water is the key to the development of linear forms visible in SRTM data, a process that was termed “dune replication”. The striking contrast between actual dune topography (“replication”) detectable in SRTM data and vegetation patterns (“ghosting”) visible in Landsat 7 (Red, Green, Blue/Bands 7, 4 and 2) data, is also observed in the MRZ. 3. Results and discussion We identified a series of normal faults to the east of the ORZ. A rose diagram indicates that the majority strike from the NNE to SSW in both the SRTM (approx. 30 , Fig. 2) and geomagnetic (approx. 35 , Fig. 4) datasets. Secondary, west to east faultsets are identifiable in the geomagnetics (90 ) that have lesser manifestion in SRTM data, where ENE to WSW faults are better expressed (70 ). Detailed microseismic studies by Scholz et al. (1976) demonstrated the current activity of the north-northeast striking faults. Focal mechanism solutions indicate that earthquake slip is mostly
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
Fig. 2. Surface Morphology of the MRZ. Major faults (black lines), scarps (multi-line) and seven fault blocks (indicated by numbers) mapped from side shaded SRTM data using a variety of illumination angles (note: only one view depicted here). The modern 945 m contour (dashed line), and shoreline ridge point elevations were obtained using SRTM. The river knickpoints (solid triangles), by Nash and Eckardt (2015) were extracted from river long profiles using SRTM also. SRTM only depicts dune ridge topography, the flat dune vegetation stripes depicted as dashed EeW lines are seen in Landsat 7 imagery only. The cross-sections (Fig. 3) AeB and CeD are indicated by the heavy black lines. The MRZ was home to a magnitude 4.2 earthquake on May 26th in 1984. Its epicenter is marked by a star (Source: United States Geological Survey database). Previously unidentified shore features to the west, are demarcated by “þ” around the modern 960e980 m contour.
3
4
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Fig. 3. Topographic cross sections. Transects AeB and CeD (both 150 km in length) as indicated in Fig. 2, with emphasis on “piano key” block depiction and major associated landforms. AeB less pronounced fault blocks hosting both Nxai and Ntwetwe Pan. CeD transect, capturing most notable blocks, scarps and replicate dune patterns evident alongside major horsts and river incisions also depicted in Fig. 5.
normal, with a minor component of dextral oblique motion. Fault dip indicated by the best located earthquakes is ~60 NW, and earthquake locations correlate well to the surface traces of the most topographically pronounced faults, suggesting these are active faults with more conspicuous surface expression. We interpret these faults as low-displacement horst and graben structures, expressed in local topographic relief (Figs. 2, 3 and 5), which control the flow of the Nunga, Lememba and Nata Rivers into the northern Makgadikgadi basin. Topologies of the endorheic MRZ drainage channels are aligned parallel to these faults, in contrast to the Okavango River, which drains perpendicularly across the ORZ faults (Fig. 1). Recent work by Nash and Eckardt (2015) has shown that some of the knickpoints associated with the MRZ drainage appear
to be controlled by the faults identified here. The mapped faults and associated forms clearly depict neotectonic processes in the MRZ. This appears to be the case even in an area where the Kalahari cover, according to Haddon and McCarthy (2005) may exceed 200 m in thickness (Fig. 4). Underlying calcrete, silcrete and hybrids thereof, as well as relatively cohesive saprolite, could have enabled fault expression through to the surface sands, as also occurs along the Gumare fault (McFarlane et al., 2005, 2010). According to SRTM data, crest heights along the contiguous Gidikwe ridge (Figs. 1 and 3), along the western rim of the basin, range in elevation from 935 to 950 m, centered along the 945 m contour. To the northeast, on the other hand, sandy ridges appear both stepped and discontinues, range in height from 930 to 965 m
Fig. 4. Geomagnetics of the MRZ. Shallow fault traces delineated in geomagnetic data provided by Department of Geological Survey, Botswana. Contour depth of Kalahari cover was obtained from Haddon and McCarthy (2005) and Council for Geoscience (2000). Rose diagram depicts fault orientations for both SRTM (Fig. 2) and magnetic data. Magnetic data for Zimbabwe not available and not included in the rose diagram. The location of knickpoints from Nash and Eckardt (2015) are indicated by solid triangles. Present day pans are shown by the grey shading.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
5
Fig. 5. Terrain view of MRZ. Extreme vertically exaggerated terrain view using SRTM and Landsat 7 drape, covering the area shown in Fig. 2 and including most of the transect CeD. Area is approximately 130 by 70 km in size, view looking north. Individual fault blocks are evident as is Lememba River valley and the result of dune replication in river valley and on the dune horst between major blocks 4 and 5. The extent of vegetation dune stripes with little topographic expression is evident in Landsat 7 imagery. Note “piano key” block offsets between 945 m lake level and sandy shore ridge also depicted in Fig. 2.
and adhere much less to the 945 m contour. We attribute these anomalies to the neotectonic “piano key” movement of seven major fault blocks separated by approximately 10e20 km (Figs. 2 and 3). These blocks form the core of the MRZ. However, faults and scarps appear less pronounced in the west (Block 1, 2 and 3, Figs. 2 and 3, transect AeB), than the east (block 4, 5, 6 and 7, Figs. 2 and 3, transect CeD). Examining the blocks in turn, block 1 to the west is among the lowest and also hosts Nxai Pan at its southern end. The Nunga River is in part controlled by its eastern faults. Block 2 appears to have dropped less than block 1, based on the stepped shape of the 945 m contour. For block 1 and 2 the 945 m contour is devoid of major sandy ridges which we attribute to block movement since lake emplacement. Lake shore ridges for block 1 and 2 lie south of the modern 945 m contour at an elevation of approximately 930e940 m. Block 3, a horst, appears among the most elevated relative to its surrounds, with its sandy mound clearly bound by the 945 m contour. Block 4 is lower than 3 but above 5, 6 and 7. To the north, block 4 is distinctly incised by the Lememba River, whose course is controlled by both replicated dune straats and faults. To the south, block 4, accommodates the Ntwetwe pan handle (Fig. 2). Block 5 is slightly lower than 4. Its southern end is bound by a sandy ridge and the 945 m contour has “spilled” onto the block surface (Figs. 2 and 5). Block 6 is higher than both 5 and 7, bounded by distinct faults and scarps (~40 m). Finally, we propose that block 7 is a discrete graben, which accommodates the smooth and broad Nata River valley. The sandy ridges formerly associated with the 945 m lake contour appear vertically displaced by the horst and graben development, suggesting block movement to have incurred in the MRZ since paleolake desiccation. Sandy ridges in blocks 1, 2 and 5 appear below the 945 m contours, in blocks 3, 4 and 6 the ridge crests appear above the contour. At this point it would be appropriate to consider the accuracy of SRTM data. Compared to 45 291 ICESat elevation points, SRTM overestimates absolute height in the MRZ, on average by 4.6 m, supporting the earlier estimates for southern African by Rodriguez et al. (2005). We therefore acknowledge that the orthometric elevation of features will need to be subjected to more accurate height estimation in future. However, considering the standard deviation of 1.8 m, we trust that the
data has sufficient fidelity to support our mapping effort of both faults and blocks. Furthermore, we noted variation in the surface morphology of the individual fault blocks as depicted in shaded SRTM data (Fig. 2). In particular the east-west running replicated dune ridges are predominantly observed in the incised Lememba River valley of block 4 and on the elevated horst blocks such as 6 and to some extend 5. This is distinctly different from the lower lying blocks 1, 2 and 7, which feature no replicated dune ridges or incised rivers. Both of these also appear to be absent on the former lake floor below the 945 m contour. The replicated ridges are particularly well exemplified on the narrow horst (approximately 30 4 km in size), extending northward from the Ntwetwe Pan between blocks 4 and 5 (Fig. 5). On the other hand blocks 2 and 5, and to some extend 4 and 6, feature dunes that are mostly identifiable as vegetation stripes in Landsat imagery only. The contrast between replicated dunes detectable in SRTM data and ghost dune vegetation patterns visible in Landsat (Fig. 5), is analogous to the fault and drainage controlled ridge topography first identified in the ORZ (McFarlane and Eckardt, 2007). Here again, we propose that the degraded dunes are replicated from semi-consolidated sands responding to changes in the local base level. Blocks that have risen relative to the surrounding terrain in particular, appear to have the necessary piezometric gradient to potentially achieve suffosion acting on interdune straats, which in part could also account for the pitted and karst like surface textures of blocks 5 and 6. While the proposed process of suffosion needs verification, it appears that vertical transport processes, by faulting and possibly leaching, play a role in shaping the topographic ridges in the Middle Kalahari. 4. Conclusion This is the first study to identify extensive surface expression of the MRZ in the context of Kalahari surface geomorphology, and link these subtle landscape patterns to underlying fault systems. The geomorphic landforms of the Makgadikgadi basin, including its shorelines (Ringrose et al., 2005; Burrough et al., 2009; Ringrose et al., 2009) and fossil dunes to the north east (Stokes et al., 1997) have to date solely been interpreted as palaeo
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
6
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
environmental proxies. Although tectonic activity in the MRZ was identified over thirty years ago (Baillieul, 1979a and b; Reeves, 1972) the various fault blocks and their effects on landforms have not been previously recognized. This study demonstrates the strength of combining SRTM and geomagnetic data analyses to identify subtle horst and graben topography, particularly to resolve sub-40 m fault scarps in the central and eastern MRZ. The results point to post-lake, surface modifications, as supported by the patches of replicated dune patterns preserved on elevated horsts. We also identify tectonic movement, revealed by drainage incision and the lateral separation of the modern 945 m contour, from the palaeolake shore form. It is also evident that the fault structures are more complex and extensive than we have depicted, because they not only continue south across the lake basin but also north through the Zambezi valley, and they possibly form a continuum with the ORZ to the west. The Makgadikgadi has long been associated with a palaeolake environment and to a lesser degree dunefield processes, but this study also points to the landscape's significant tectonic history, which has entailed both exogenic and endogenic controls acting on surface processes. The formeprocesses relationships that have shaped the replicated dune patterns require further examination as does post sedimentary modification of landforms associated with the palaeolake. These findings have implications for the palaeoenvironmental interpretation of landforms in the Middle Kalahari. Acknowledgments We would like to thank Richard Kahle for the preparation and analyses of the geomagnetic data and Andy Moore for his reviewer comments. References Baillieul, T.A., 1979a. Makgadikgadi pans complex of central Botswana: summary. Geological Society of America Bulletin 90, 133e136. Baillieul, T.A., 1979b. Makgadikgadi Pans complex of central Botswana. Geological Society of America Bulletin 90, 289e312. Bufford, K.M., Atekwana, E.A., Abdelsalam, M.G., Shemang, E., Atekwana, E.A., Mickus, K., Moidaki, M., Modisi, M.P., Molwalefhe, L.M., 2012. Geometry and faults tectonic activity of the Okavango Rift Zone, Botswana: evidence from magnetotelluric and electrical resistivity tomography imaging. Journal of African Earth Sciences 65, 61e71. Burrough, S.L., Thomas, D.S., Bailey, R.M., 2009. Mega-Lake in the Kalahari: a Late Pleistocene record of the Palaeolake Makgadikgadi system. Quaternary Science Reviews 28, 1392e1411. Cooke, H.J., 1980. Landform evolution in the context of climatic change and neotectonism in the Middle Kalahari of north-central Botswana. Transactions of the Institute of British Geographers 80e99. Council for Geoscience (South Africa), Erasmus, D.M., Roos, H.M., 2000. Isopach Map of the Kalahari Group. SADC by the Council for Geoscience. Fricker, H.A., Borsa, A., Minster, B., Carabajal, C., Quinn, K., Bills, B., 2005. Assessment of ICESat performance at the Salar de Uyuni, Bolivia. Geophysical Research Letters 32 (21). Grove, A.T., 1969. Landforms and climatic change in the Kalahari and Ngamiland. Geographical Journal 191e212. Gumbricht, T., McCarthy, T.S., Merry, C.L., 2001. The topography of the Okavango Delta, Botswana, and its tectonic and sedimentological implications. South African Journal of Geology 104, 243e264. Haddon, I.G., McCarthy, T.S., 2005. The MesozoiceCenozoic interior sag basins of Central Africa: the Late-CretaceouseCenozoic Kalahari and Okavango basins. Journal of African Earth Sciences 43, 316e333. Heine, K., 1978. Radiocarbon chronology of Late Quaternary lakes in the Kalahari, southern Africa. Catena 5, 145e149.
Helgren, D.M., 1984. Historical geomorphology and geoarchaeology in the southwestern Makgadikgadi Basin, Botswana. Annals of the Association of American Geographers 74, 298e307. Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole-filled SRTM for the Globe Version 4. CGIAR-CSI SRTM 90m Database. http://srtm.csi.cgiar.org. Kinabo, B.D., Atekwana, E.A., Hogan, J.P., Modisi, M.P., Wheaton, D.D., Kampunzu, A.B., 2007. Early structural development of the Okavango rift zone, NW Botswana. Journal of African Earth Sciences 48, 125e136. Kinabo, B.D., Hogan, J.P., Atekwana, E.A., Abdelsalam, M.G., Modisi, M.P., 2008. Fault growth and propagation during incipient continental rifting: insights from a combined aeromagnetic and shuttle radar topography mission digital elevation model investigation of the Okavango Rift Zone, northwest Botswana. Tectonics 27 (3), TC3013. Mallick, D.I.J., Habgood, F., Skinner, A.C., 1981. Geological Interpretation of Landsat Imagery and Air Photography of Botswana. Overseas Geology and Mineral Resources No. 56. Institute of Geological Sciences, London. McCarthy, T.S., Ellery, W.N., 1998. The Okavango delta. Transactions of the Royal Society of South Africa 53, 157e182. McFarlane, M.J., Eckardt, F.D., Ringrose, S., Coetzee, S.H., Kuhn, J.R., 2005. Degradation of linear dunes in Northwest Ngamiland, Botswana and the implications for luminescence dating of periods of aridity. Quaternary International 135, 83e90. McFarlane, M.J., Eckardt, F.D., 2007. Palaeodune morphology associated with the Gumare fault of the Okavango graben in the Botswana/Namibia borderland, a new model of tectonic influence. South African Journal of Geology 110, 535e542. McFarlane, M.J., Eckardt, F.D., Coetzee, S.H., Ringrose, S., 2010. An African surface weathering profile in the Kalahari of North West Ngamiland, Botswana: processes and products. Zeitschrift für Geomorphologie 54, 273e303. Modisi, M.P., 2000. Fault system at the southeastern boundary of the Okavango Rift, Botswana. Journal of African Earth Sciences 30, 569e578. Moore, A.E., Cotterill, F.P.D., Eckardt, F.D., 2012. The evolution and ages of Makgadikgadi palaeo-lakes: consilient evidence from Kalahari drainage evolution. South African Journal of Geology 115, 385e413. Nash, D., Eckardt, F.D., 2015. Drainage development, neotectonics and base-level change in the Kalahari Desert, southern Africa. South African Geographical Journal 1e13. Podgorski, J.E., Green, A.G., Kgotlhang, L., Kinzelbach, W.K., Kalscheuer, T., Auken, E., Ngwisanyi, T., 2013. Paleo-megalake and paleo-megafan in southern Africa. Geology 41, 1155e1158. Reeves, C.V., 1972. Rifting in the Kalahari ? Nature 237, 95e96. Riedel, F., Henderson, A.C., Heußner, K.U., Kaufmann, G., Kossler, A., Leipe, C., Shemang, E., Taft, L., 2014. Dynamics of a Kalahari long-lived mega-lake system: hydromorphological and limnological changes in the Makgadikgadi Basin (Botswana) during the terminal 50 ka. Hydrobiologia 739, 25e53. Ringrose, S., Downey, B., Genecke, D., Sefe, F., Vink, B., 1999. Nature of sedimentary deposits in the western Makgadikgadi littoral zone. Journal of Arid Environments 32, 281e299. Ringrose, S., Huntsman-Mapila, P., Kampunzu, A.B., Downey, W., Coetzee, S., Vink, B., Matheson, W., Vanderpost, C., 2005. Sedimentological and geochemical evidence for palaeo-environmental change in the Makgadikgadi subbasin, in relation to the MOZ rift depression, Botswana. Palaeogeography, Palaeoclimatology, Palaeoecology 217, 265e287. Ringrose, S., Harris, C., Huntsman-Mapila, P., Vink, B.W., Diskins, S., Vanderpost, C., Matheson, W., 2009. Origins of strandline duricrusts around the Makgadikgadi Pans (Botswana Kalahari) as deduced from their chemical and isotope composition. Sedimentary Geology 219, 262e279. Rodriguez, E., Morris, C.S., Belz, J.E., Chapin, E.C., Martin, J.M., Daffer, W., Hensley, S., 2005. An Assessment of the SRTM Topographic Products. Technical Report JPL D-31639. Jet Propulsion Laboratory, Pasadena, California, p. 143. Shaw, P.A., Cooke, H.J., 1986. Geomorphic evidence for the Late Quaternary palaeoclimates of the middle Kalahari of northern Botswana. Catena 13, 349e359. Shaw, P.A., Stokes, S., Thomas, D.S.G., Davies, F.B.M., Holmgren, K., 1997. Palaeoecology and age of a Quaternary high lake level in the Makgadikgadi Basin of the Middle Kalahari, Botswana. South African Journal of Science 93, 273e276. Shemang, E., Molwalefhe, L., 2011. Geomorphic landforms and tectonism along the eastern margin of the Okavango Rift Zone, Northwestern Botswana, as deduced from geophysical data. In: Sharkov, E. (Ed.), New Frontiers in Tectonics Research, General Problems, Sedimentary Basins and Island Arcs. InTech, pp. 169e182. Scholz, C.H., Koczynski, T.A., Hutchins, D.G., 1976. Evidence for incipient rifting in southern Africa. Geophysical Journal International 44, 135e144. Schutz, B.E., Zwally, H.J., Shuman, C.A., Hancock, D., DiMarzio, J.P., 2005. Overview of the ICESat mission. Geophysical Research Letters 32 (21). Stokes, S., Thomas, D.S.G., Shaw, P.A., 1997. New chronological evidence for the nature and timing of linear dune development in the southwest Kalahari Desert. Geomorphology 20, 81e93.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ) Frank D. Eckardt a, *, Fenton P.D. Cotterill b, Tyrel J. Flügel c, Beth Kahle d, Marty McFarlane e, Christie Rowe f a
Environmental and Geographical Science, University of Cape Town, Rondebosch, 7701, South Africa Geoecodynamics Research Hub, Department of Botany and Zoology, University of Stellenbosch, Matieland, 7602, South Africa Department of Military Geography, Stellenbosch University, Private Bag X2, 7395, Saldanha, South Africa d Department of Geological Sciences, University of Cape Town, Rondebosch, 7701, South Africa e Marty McFarlane, Bosele Investments (Pty) Limited, Box 66, Maun, Botswana f Earth & Planetary Sciences Department, McGill University, Montr eal, QC, H3A 0E8, Canada b c
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
The Makgadikgadi basin and wider Middle Kalahari region of Botswana and beyond host landforms which have been attributed to Quaternary environmental change, including palaeolake level fluctuations and aeolian activity. Tectonic processes and landforms on the other hand, have mostly been linked to the Okavango graben and associated rift zone (ORZ) to the west of the Makgadikgadi. In this paper we establish the extent of tectonic surface expression associated with the Makgadikgadi Rift Zone (MRZ). We identify a series of parallel, NNE-SSW, normal faults and scarps, expressing horst and graben structures linked to seven major blocks in the northern Makgadikgadi basin, using both Shuttle Radar Topographic Mission (SRTM) and geomagnetic data. Subtle expression of rifting has controlled endorheic drainage topology, replicated regional dune-field patterns and displaced the 945 m palaoelake contour since lake desiccation. These observations underscore the role of neotectonic “piano key” block movement in shaping surface landforms across a large expanse of the Kalahari region. This paper provides the first detailed map and introduction to the Makgadikgadi Rift Zone (MRZ) and its geomorphology. © 2015 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Kalahari Geomorphology Makgadikgadi Rifting SRTM ICESat
1. Introduction This study is centered on subtle neotectonic landforms in the Middle Kalahari, considered terminal branches of the East African Rift System (Scholz et al., 1976), with the Okavango Rift Zone (ORZ), an alluvial fan, being the most documented example (McCarthy and Ellery, 1998; Modisi, 2000; Gumbricht et al., 2001; Kinabo et al., 2007, 2008; Shemang and Molwalefhe, 2011; Bufford et al., 2012; Podgorksi et al., 2013). In this paper we will particularly focus on the Makgadikgadi Rift Zone (MRZ) to the east of the ORZ (Fig. 1). Research emphasis in the MRZ in contrast, has been on Late Quaternary lacustrine sandy ridges, flanking the perimeter of the current dry lake basin (Grove, 1969; Heine, 1978; Cooke, 1980; Mallick et al., 1981; Helgren, 1984; Shaw and Cooke, 1986; Shaw et al., 1997; Ringrose et al., 1999, 2005; Burrough et al., 2009; Ringrose et al.,
* Corresponding author. E-mail address: [email protected] (F.D. Eckardt).
2009; Moore et al., 2012; Riedel et al., 2014) as well as fossil dune forms (Stokes et al., 1997). Although the ORZ and MRZ both include recent and ongoing horst and graben development (Baillieul, 1979a and b; Reeves, 1972), and hosted the same regional palaeolake (Cooke, 1980), it is fair to say that the MRZ has remained comparatively understudied. With this in mind, we set out to identify, map and describe faults and related geomorphic landforms, covering a 40 000 km square area of north-eastern Botswana and western Zimbabwe. The results include the identification of seven major fault blocks, control of drainage channels flowing into the basin, the distribution of palaeo dune topography and the vertical displacement of the sandy ridges associated with the paleo 945 m lake level. 2. Methods Surface traces of faults were mapped using SRTM3 (Shuttle Radar Topography Mission, 3-arc second resolution) version 4 data obtained from the CGIAR-CSI site (Consultative Group on International Agricultural Research e Consortium for Spatial Information,
http://dx.doi.org/10.1016/j.quaint.2015.09.002 1040-6182/© 2015 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
2
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Fig. 1. Introduction of the MRZ. The Middle Kalahari of northern Botswana, home to the Makgadikgadi Rift Zone (MRZ) to the east and the Okavango Rift Zone (ORZ), (Kinabo et al., 2007) to the west, as well as the former 945 m palaeolake outline (light grey shading), extracted from SRTM data. Gidikwe Ridge and associated ridge lines (dotted lines) were obtained from Mallick et al., 1981. Preliminary MRZ faults (thick solid lines) depicted here are taken from Baillieul (1979a and b). Dunes ridges (thin solid EeW lines) were mapped from vegetation stripes in Landsat 7 (Image Source: Global Land Cover Facility).
Jarvis et al., 2008), The unenhanced 90 m DSM (Digital Surface Model) data was side shaded and illuminated with threefold vertical exaggeration, using variable azimuths including 0, 45, 90, 125, 180, 225, 270 and 315 and applying sun elevation of 45 . Faults appeared as continuous or discontinuous straight features, indicative of tectonic fault block movement (Figs. 2 and 3). Fault traces were also delineated using national aeromagnetic data, supplied by the Department of Geological Surveys, Botswana (Fig. 4). The survey was flown at a height of 80 m with a line spacing of 200 m and the data converted to a 50 m grid. A range of derivative and trigonometric filters were applied in order to enhance the shallow and subtle features in the magnetic dataset. Mapping of faults from this datasets relies on a magnetic contrast between units on opposite sides of the fault, or between the fault itself and the rocks that it displaces. A fault may appear as a magnetic anomaly in its own right, presumably as a result of fluid flow and precipitation or stripping of magnetic minerals. While it may be difficult to separate dykes and faults with a high degree of confidence, the mapped faults do not coincide with the prominent Karoo-age swarm dykes (NWeSE) in the extreme SW of the dataset. Approximate Kalahari thickness data was also depicted in Fig. 4. This contour data by Haddon and McCarthy (2005) is suitable for mapping regional trends of Kalahari depth but lacks the resolution to identify relative uplift and subsidence of underlying horst and graben structures. Additionally SRTM was used to generate the 945 m contour, point heights for sandy ridges, resolve stream topology, identified fossil dune ridges and provided the topographic cross-sections (Fig. 3) and terrain view (Fig. 5). It is important to note that while SRTM generally overestimates absolute elevation in southern Africa by approximately 5 m, its relative accuracy however is better than 5 m (Rodriguez et al., 2005). For additional validation purposes we also compared SRTM elevation data against the ICESat (Ice, Cloud, and land Elevation Satellite) laser altimeter points, which has a much greater absolute vertical accuracies (Global elevation data product (GLAH06), Level-1B, laser campaigns: L1A, L2A, L2A, L2C, L3A, L3B, L3C, L3D, L3E, L3F, L3G. L3H, L3I and L3J, release 33)
(Schutz et al., 2005) for the Makgadikgadi basin specifically (spatial domain: Latitude 20.17 to 21.46, Longitude þ24.72 to þ26.61). While ICESat laser coverage is not dense enough to produce digital elevation models, it is well suited for validation purposes, since the 65 m diameter laser altimeter footprints, separated by 172 m at ground level, have returned a vertical accuracy of <2 cm (Fricker et al., 2005). For comparative purposes we ensured that all elevation data used here conformed to the WGS 84 ellipsoid. Fossil dune distribution patterns for the middle Kalahari have traditionally been inferred from the vegetation patterns detectable in Landsat satellite imagery (Mallick et al., 1981). However the vegetation continues to reveal the past dune forms, even after the dune ridges have been degraded or even substantially flattened. While mapping fossil dunes near the Gumare Fault (Fig. 1, McFarlane and Eckardt, 2007) and comparing the vegetation stripes in Landsat imagery against SRTM data, actual dune topography appeared most pronounced at the margins of, or within downfaulted graben and along incised streams, essentially in areas of lowered local base level. It was proposed that the action of water is the key to the development of linear forms visible in SRTM data, a process that was termed “dune replication”. The striking contrast between actual dune topography (“replication”) detectable in SRTM data and vegetation patterns (“ghosting”) visible in Landsat 7 (Red, Green, Blue/Bands 7, 4 and 2) data, is also observed in the MRZ. 3. Results and discussion We identified a series of normal faults to the east of the ORZ. A rose diagram indicates that the majority strike from the NNE to SSW in both the SRTM (approx. 30 , Fig. 2) and geomagnetic (approx. 35 , Fig. 4) datasets. Secondary, west to east faultsets are identifiable in the geomagnetics (90 ) that have lesser manifestion in SRTM data, where ENE to WSW faults are better expressed (70 ). Detailed microseismic studies by Scholz et al. (1976) demonstrated the current activity of the north-northeast striking faults. Focal mechanism solutions indicate that earthquake slip is mostly
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
Fig. 2. Surface Morphology of the MRZ. Major faults (black lines), scarps (multi-line) and seven fault blocks (indicated by numbers) mapped from side shaded SRTM data using a variety of illumination angles (note: only one view depicted here). The modern 945 m contour (dashed line), and shoreline ridge point elevations were obtained using SRTM. The river knickpoints (solid triangles), by Nash and Eckardt (2015) were extracted from river long profiles using SRTM also. SRTM only depicts dune ridge topography, the flat dune vegetation stripes depicted as dashed EeW lines are seen in Landsat 7 imagery only. The cross-sections (Fig. 3) AeB and CeD are indicated by the heavy black lines. The MRZ was home to a magnitude 4.2 earthquake on May 26th in 1984. Its epicenter is marked by a star (Source: United States Geological Survey database). Previously unidentified shore features to the west, are demarcated by “þ” around the modern 960e980 m contour.
3
4
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
Fig. 3. Topographic cross sections. Transects AeB and CeD (both 150 km in length) as indicated in Fig. 2, with emphasis on “piano key” block depiction and major associated landforms. AeB less pronounced fault blocks hosting both Nxai and Ntwetwe Pan. CeD transect, capturing most notable blocks, scarps and replicate dune patterns evident alongside major horsts and river incisions also depicted in Fig. 5.
normal, with a minor component of dextral oblique motion. Fault dip indicated by the best located earthquakes is ~60 NW, and earthquake locations correlate well to the surface traces of the most topographically pronounced faults, suggesting these are active faults with more conspicuous surface expression. We interpret these faults as low-displacement horst and graben structures, expressed in local topographic relief (Figs. 2, 3 and 5), which control the flow of the Nunga, Lememba and Nata Rivers into the northern Makgadikgadi basin. Topologies of the endorheic MRZ drainage channels are aligned parallel to these faults, in contrast to the Okavango River, which drains perpendicularly across the ORZ faults (Fig. 1). Recent work by Nash and Eckardt (2015) has shown that some of the knickpoints associated with the MRZ drainage appear
to be controlled by the faults identified here. The mapped faults and associated forms clearly depict neotectonic processes in the MRZ. This appears to be the case even in an area where the Kalahari cover, according to Haddon and McCarthy (2005) may exceed 200 m in thickness (Fig. 4). Underlying calcrete, silcrete and hybrids thereof, as well as relatively cohesive saprolite, could have enabled fault expression through to the surface sands, as also occurs along the Gumare fault (McFarlane et al., 2005, 2010). According to SRTM data, crest heights along the contiguous Gidikwe ridge (Figs. 1 and 3), along the western rim of the basin, range in elevation from 935 to 950 m, centered along the 945 m contour. To the northeast, on the other hand, sandy ridges appear both stepped and discontinues, range in height from 930 to 965 m
Fig. 4. Geomagnetics of the MRZ. Shallow fault traces delineated in geomagnetic data provided by Department of Geological Survey, Botswana. Contour depth of Kalahari cover was obtained from Haddon and McCarthy (2005) and Council for Geoscience (2000). Rose diagram depicts fault orientations for both SRTM (Fig. 2) and magnetic data. Magnetic data for Zimbabwe not available and not included in the rose diagram. The location of knickpoints from Nash and Eckardt (2015) are indicated by solid triangles. Present day pans are shown by the grey shading.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
5
Fig. 5. Terrain view of MRZ. Extreme vertically exaggerated terrain view using SRTM and Landsat 7 drape, covering the area shown in Fig. 2 and including most of the transect CeD. Area is approximately 130 by 70 km in size, view looking north. Individual fault blocks are evident as is Lememba River valley and the result of dune replication in river valley and on the dune horst between major blocks 4 and 5. The extent of vegetation dune stripes with little topographic expression is evident in Landsat 7 imagery. Note “piano key” block offsets between 945 m lake level and sandy shore ridge also depicted in Fig. 2.
and adhere much less to the 945 m contour. We attribute these anomalies to the neotectonic “piano key” movement of seven major fault blocks separated by approximately 10e20 km (Figs. 2 and 3). These blocks form the core of the MRZ. However, faults and scarps appear less pronounced in the west (Block 1, 2 and 3, Figs. 2 and 3, transect AeB), than the east (block 4, 5, 6 and 7, Figs. 2 and 3, transect CeD). Examining the blocks in turn, block 1 to the west is among the lowest and also hosts Nxai Pan at its southern end. The Nunga River is in part controlled by its eastern faults. Block 2 appears to have dropped less than block 1, based on the stepped shape of the 945 m contour. For block 1 and 2 the 945 m contour is devoid of major sandy ridges which we attribute to block movement since lake emplacement. Lake shore ridges for block 1 and 2 lie south of the modern 945 m contour at an elevation of approximately 930e940 m. Block 3, a horst, appears among the most elevated relative to its surrounds, with its sandy mound clearly bound by the 945 m contour. Block 4 is lower than 3 but above 5, 6 and 7. To the north, block 4 is distinctly incised by the Lememba River, whose course is controlled by both replicated dune straats and faults. To the south, block 4, accommodates the Ntwetwe pan handle (Fig. 2). Block 5 is slightly lower than 4. Its southern end is bound by a sandy ridge and the 945 m contour has “spilled” onto the block surface (Figs. 2 and 5). Block 6 is higher than both 5 and 7, bounded by distinct faults and scarps (~40 m). Finally, we propose that block 7 is a discrete graben, which accommodates the smooth and broad Nata River valley. The sandy ridges formerly associated with the 945 m lake contour appear vertically displaced by the horst and graben development, suggesting block movement to have incurred in the MRZ since paleolake desiccation. Sandy ridges in blocks 1, 2 and 5 appear below the 945 m contours, in blocks 3, 4 and 6 the ridge crests appear above the contour. At this point it would be appropriate to consider the accuracy of SRTM data. Compared to 45 291 ICESat elevation points, SRTM overestimates absolute height in the MRZ, on average by 4.6 m, supporting the earlier estimates for southern African by Rodriguez et al. (2005). We therefore acknowledge that the orthometric elevation of features will need to be subjected to more accurate height estimation in future. However, considering the standard deviation of 1.8 m, we trust that the
data has sufficient fidelity to support our mapping effort of both faults and blocks. Furthermore, we noted variation in the surface morphology of the individual fault blocks as depicted in shaded SRTM data (Fig. 2). In particular the east-west running replicated dune ridges are predominantly observed in the incised Lememba River valley of block 4 and on the elevated horst blocks such as 6 and to some extend 5. This is distinctly different from the lower lying blocks 1, 2 and 7, which feature no replicated dune ridges or incised rivers. Both of these also appear to be absent on the former lake floor below the 945 m contour. The replicated ridges are particularly well exemplified on the narrow horst (approximately 30 4 km in size), extending northward from the Ntwetwe Pan between blocks 4 and 5 (Fig. 5). On the other hand blocks 2 and 5, and to some extend 4 and 6, feature dunes that are mostly identifiable as vegetation stripes in Landsat imagery only. The contrast between replicated dunes detectable in SRTM data and ghost dune vegetation patterns visible in Landsat (Fig. 5), is analogous to the fault and drainage controlled ridge topography first identified in the ORZ (McFarlane and Eckardt, 2007). Here again, we propose that the degraded dunes are replicated from semi-consolidated sands responding to changes in the local base level. Blocks that have risen relative to the surrounding terrain in particular, appear to have the necessary piezometric gradient to potentially achieve suffosion acting on interdune straats, which in part could also account for the pitted and karst like surface textures of blocks 5 and 6. While the proposed process of suffosion needs verification, it appears that vertical transport processes, by faulting and possibly leaching, play a role in shaping the topographic ridges in the Middle Kalahari. 4. Conclusion This is the first study to identify extensive surface expression of the MRZ in the context of Kalahari surface geomorphology, and link these subtle landscape patterns to underlying fault systems. The geomorphic landforms of the Makgadikgadi basin, including its shorelines (Ringrose et al., 2005; Burrough et al., 2009; Ringrose et al., 2009) and fossil dunes to the north east (Stokes et al., 1997) have to date solely been interpreted as palaeo
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002
6
F.D. Eckardt et al. / Quaternary International xxx (2015) 1e6
environmental proxies. Although tectonic activity in the MRZ was identified over thirty years ago (Baillieul, 1979a and b; Reeves, 1972) the various fault blocks and their effects on landforms have not been previously recognized. This study demonstrates the strength of combining SRTM and geomagnetic data analyses to identify subtle horst and graben topography, particularly to resolve sub-40 m fault scarps in the central and eastern MRZ. The results point to post-lake, surface modifications, as supported by the patches of replicated dune patterns preserved on elevated horsts. We also identify tectonic movement, revealed by drainage incision and the lateral separation of the modern 945 m contour, from the palaeolake shore form. It is also evident that the fault structures are more complex and extensive than we have depicted, because they not only continue south across the lake basin but also north through the Zambezi valley, and they possibly form a continuum with the ORZ to the west. The Makgadikgadi has long been associated with a palaeolake environment and to a lesser degree dunefield processes, but this study also points to the landscape's significant tectonic history, which has entailed both exogenic and endogenic controls acting on surface processes. The formeprocesses relationships that have shaped the replicated dune patterns require further examination as does post sedimentary modification of landforms associated with the palaeolake. These findings have implications for the palaeoenvironmental interpretation of landforms in the Middle Kalahari. Acknowledgments We would like to thank Richard Kahle for the preparation and analyses of the geomagnetic data and Andy Moore for his reviewer comments. References Baillieul, T.A., 1979a. Makgadikgadi pans complex of central Botswana: summary. Geological Society of America Bulletin 90, 133e136. Baillieul, T.A., 1979b. Makgadikgadi Pans complex of central Botswana. Geological Society of America Bulletin 90, 289e312. Bufford, K.M., Atekwana, E.A., Abdelsalam, M.G., Shemang, E., Atekwana, E.A., Mickus, K., Moidaki, M., Modisi, M.P., Molwalefhe, L.M., 2012. Geometry and faults tectonic activity of the Okavango Rift Zone, Botswana: evidence from magnetotelluric and electrical resistivity tomography imaging. Journal of African Earth Sciences 65, 61e71. Burrough, S.L., Thomas, D.S., Bailey, R.M., 2009. Mega-Lake in the Kalahari: a Late Pleistocene record of the Palaeolake Makgadikgadi system. Quaternary Science Reviews 28, 1392e1411. Cooke, H.J., 1980. Landform evolution in the context of climatic change and neotectonism in the Middle Kalahari of north-central Botswana. Transactions of the Institute of British Geographers 80e99. Council for Geoscience (South Africa), Erasmus, D.M., Roos, H.M., 2000. Isopach Map of the Kalahari Group. SADC by the Council for Geoscience. Fricker, H.A., Borsa, A., Minster, B., Carabajal, C., Quinn, K., Bills, B., 2005. Assessment of ICESat performance at the Salar de Uyuni, Bolivia. Geophysical Research Letters 32 (21). Grove, A.T., 1969. Landforms and climatic change in the Kalahari and Ngamiland. Geographical Journal 191e212. Gumbricht, T., McCarthy, T.S., Merry, C.L., 2001. The topography of the Okavango Delta, Botswana, and its tectonic and sedimentological implications. South African Journal of Geology 104, 243e264. Haddon, I.G., McCarthy, T.S., 2005. The MesozoiceCenozoic interior sag basins of Central Africa: the Late-CretaceouseCenozoic Kalahari and Okavango basins. Journal of African Earth Sciences 43, 316e333. Heine, K., 1978. Radiocarbon chronology of Late Quaternary lakes in the Kalahari, southern Africa. Catena 5, 145e149.
Helgren, D.M., 1984. Historical geomorphology and geoarchaeology in the southwestern Makgadikgadi Basin, Botswana. Annals of the Association of American Geographers 74, 298e307. Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole-filled SRTM for the Globe Version 4. CGIAR-CSI SRTM 90m Database. http://srtm.csi.cgiar.org. Kinabo, B.D., Atekwana, E.A., Hogan, J.P., Modisi, M.P., Wheaton, D.D., Kampunzu, A.B., 2007. Early structural development of the Okavango rift zone, NW Botswana. Journal of African Earth Sciences 48, 125e136. Kinabo, B.D., Hogan, J.P., Atekwana, E.A., Abdelsalam, M.G., Modisi, M.P., 2008. Fault growth and propagation during incipient continental rifting: insights from a combined aeromagnetic and shuttle radar topography mission digital elevation model investigation of the Okavango Rift Zone, northwest Botswana. Tectonics 27 (3), TC3013. Mallick, D.I.J., Habgood, F., Skinner, A.C., 1981. Geological Interpretation of Landsat Imagery and Air Photography of Botswana. Overseas Geology and Mineral Resources No. 56. Institute of Geological Sciences, London. McCarthy, T.S., Ellery, W.N., 1998. The Okavango delta. Transactions of the Royal Society of South Africa 53, 157e182. McFarlane, M.J., Eckardt, F.D., Ringrose, S., Coetzee, S.H., Kuhn, J.R., 2005. Degradation of linear dunes in Northwest Ngamiland, Botswana and the implications for luminescence dating of periods of aridity. Quaternary International 135, 83e90. McFarlane, M.J., Eckardt, F.D., 2007. Palaeodune morphology associated with the Gumare fault of the Okavango graben in the Botswana/Namibia borderland, a new model of tectonic influence. South African Journal of Geology 110, 535e542. McFarlane, M.J., Eckardt, F.D., Coetzee, S.H., Ringrose, S., 2010. An African surface weathering profile in the Kalahari of North West Ngamiland, Botswana: processes and products. Zeitschrift für Geomorphologie 54, 273e303. Modisi, M.P., 2000. Fault system at the southeastern boundary of the Okavango Rift, Botswana. Journal of African Earth Sciences 30, 569e578. Moore, A.E., Cotterill, F.P.D., Eckardt, F.D., 2012. The evolution and ages of Makgadikgadi palaeo-lakes: consilient evidence from Kalahari drainage evolution. South African Journal of Geology 115, 385e413. Nash, D., Eckardt, F.D., 2015. Drainage development, neotectonics and base-level change in the Kalahari Desert, southern Africa. South African Geographical Journal 1e13. Podgorski, J.E., Green, A.G., Kgotlhang, L., Kinzelbach, W.K., Kalscheuer, T., Auken, E., Ngwisanyi, T., 2013. Paleo-megalake and paleo-megafan in southern Africa. Geology 41, 1155e1158. Reeves, C.V., 1972. Rifting in the Kalahari ? Nature 237, 95e96. Riedel, F., Henderson, A.C., Heußner, K.U., Kaufmann, G., Kossler, A., Leipe, C., Shemang, E., Taft, L., 2014. Dynamics of a Kalahari long-lived mega-lake system: hydromorphological and limnological changes in the Makgadikgadi Basin (Botswana) during the terminal 50 ka. Hydrobiologia 739, 25e53. Ringrose, S., Downey, B., Genecke, D., Sefe, F., Vink, B., 1999. Nature of sedimentary deposits in the western Makgadikgadi littoral zone. Journal of Arid Environments 32, 281e299. Ringrose, S., Huntsman-Mapila, P., Kampunzu, A.B., Downey, W., Coetzee, S., Vink, B., Matheson, W., Vanderpost, C., 2005. Sedimentological and geochemical evidence for palaeo-environmental change in the Makgadikgadi subbasin, in relation to the MOZ rift depression, Botswana. Palaeogeography, Palaeoclimatology, Palaeoecology 217, 265e287. Ringrose, S., Harris, C., Huntsman-Mapila, P., Vink, B.W., Diskins, S., Vanderpost, C., Matheson, W., 2009. Origins of strandline duricrusts around the Makgadikgadi Pans (Botswana Kalahari) as deduced from their chemical and isotope composition. Sedimentary Geology 219, 262e279. Rodriguez, E., Morris, C.S., Belz, J.E., Chapin, E.C., Martin, J.M., Daffer, W., Hensley, S., 2005. An Assessment of the SRTM Topographic Products. Technical Report JPL D-31639. Jet Propulsion Laboratory, Pasadena, California, p. 143. Shaw, P.A., Cooke, H.J., 1986. Geomorphic evidence for the Late Quaternary palaeoclimates of the middle Kalahari of northern Botswana. Catena 13, 349e359. Shaw, P.A., Stokes, S., Thomas, D.S.G., Davies, F.B.M., Holmgren, K., 1997. Palaeoecology and age of a Quaternary high lake level in the Makgadikgadi Basin of the Middle Kalahari, Botswana. South African Journal of Science 93, 273e276. Shemang, E., Molwalefhe, L., 2011. Geomorphic landforms and tectonism along the eastern margin of the Okavango Rift Zone, Northwestern Botswana, as deduced from geophysical data. In: Sharkov, E. (Ed.), New Frontiers in Tectonics Research, General Problems, Sedimentary Basins and Island Arcs. InTech, pp. 169e182. Scholz, C.H., Koczynski, T.A., Hutchins, D.G., 1976. Evidence for incipient rifting in southern Africa. Geophysical Journal International 44, 135e144. Schutz, B.E., Zwally, H.J., Shuman, C.A., Hancock, D., DiMarzio, J.P., 2005. Overview of the ICESat mission. Geophysical Research Letters 32 (21). Stokes, S., Thomas, D.S.G., Shaw, P.A., 1997. New chronological evidence for the nature and timing of linear dune development in the southwest Kalahari Desert. Geomorphology 20, 81e93.
Please cite this article in press as: Eckardt, F.D., et al., Mapping the surface geomorphology of the Makgadikgadi Rift Zone (MRZ), Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.09.002