Figure 1. General geographic map of Ethiopia (African Studies WWW, U. Penn).
Naomi Lubick, Lamont-Doherty Earth Observatory, Columbia University, New York, NY 10027; email: nbl11@columbia.edu; Submitted 12/22/98
Abstract
Introduction
General History of the Region
Project
Results
Problems
Conclusions
Future Work
Literature Cited
The central eastern Ethiopian Plateau is a volcanic surface that is being eroded into by streams that flow west, feeding into the Blue Nile. This project culminates in the creation of three-dimensional (3D) images using a Landsat Thematic Mapper (TM) image and a digital elevation model (DEM). The results I report here are minimal, but the process holds promise for contributions to the study of bedrock-bottomed streams.
Bedrock erosion processes have been little considered in geomorphology, which, in general, has focused on alluvial processes (Tinkler & Wohl, 1998). This study focuses on the stream systems of the Ethiopian Plateau, which sits to the west of the Ethiopian rift zone (see Figure 1). These streams are eroding bedrock that consists of a volcanic surface, overlying sediments. The volcanic surface can be considered a "clean slate" for the regional geomorphology, which should allow inferences as to the initiation of erosion in a stream area and the migration of the knickpoints of the streams into the headlands of the plateau.
Figure 1. General geographic map of Ethiopia (African Studies WWW, U. Penn).
I use two remotely sensed datasets from this region, a Landsat Thematic Mapper (TM) image (Figure 2) and a digital elevation model (DEM; Figure 3), to look at the streams and their gorges. Using images from the Landsat TM and DEM, three-dimensional (3D) images can be made of the streams that are flowing to the west, feeding in to the Blue Nile river system. The general idea of this project is to examine the propagation of the streams into the eastern upraised portion of the central eastern Ethiopian Plateau.
The central eastern portion of the Ethiopian Plateau bounds the East African Rift system and Afar, and the specific region studied sits to the northeast of Addis Ababa (see Figure 4). The Ethiopian rift was initiated approximately 30 m.y.a., and has continued to evolve, with several episodes of volcanism (George et al., 1998; WoldeGabriel et al., 1990). The central eastern Ethiopian Plateau surface is composed of remnant volcanics that are Oligocene-Miocene flood basalts (WoldeGabriel et al., 1990). The region experienced continental lithosphere rebound during extension, and the nearly uneroded plateau is preserved by this uplift, which was approximately 1000 m (Weissel et al., 1995). Because the age of the surface can be constrained, this area is particularly appropriate for stream erosion studies.
The regional slope of this area is now controlling the eastward migration of the heads of streams, and the knickpoints of these streams are migrating headwards. The migration of the knickpoints is controlled by the bedrock, which serves as a base level for stream profiles (i.e., upstream erosion is controlled by the elevation of the bedrock surface downstream; Tinkler & Wohl, 1998). The area receives 1,000 mm of rain per year, and undergoes drought often (information from www.africanet.com; see full address below). The vegetation in the images used is concentrated around the streams, and there are wide arid areas that do not seem to have stream channels yet eroded into them.
The goal of this project was to obtain 3D images to examine the streams feeding into the Blue Nile. The main images used are the Landsat TM image (path 168, row 53), collected in January 1986, and a NASA DEM of the East African Rift (April 1990; from D. Harding). In order to "warp" the Landsat TM image (bands 7, 4, and 3) to the DEM of the region, I chose approximately 60 ground control points (GCPs), using the application ENVI. The recognizable features of the landscape that I used as GCPs were mostly stream intersections.
When the GCPs were collected, I warped the TM image to the DEM, which proved unsatisfactory because of the loss of resolution (the DEM image has a resolution of 30 m per pixels and the TM has a resolution of 20 m). In order to maintain the higher resolution of the TM image, I switched the GCPs in order to use the Landsat as the base image and warped the DEM to the TM image. Once the GCPs were finalized, I draped sections of the Landsat TM over the DEM to obtain 3D images of the region (Figures 5, 6, 7).
Figure 5. Upper stream sections from the northeast portion of the TM image, shown at a 75 degree tilt from horizontal. Note how the stream tips are eroding into a relatively flat region in the upper right of the image. The elevation range is about 1200 to 3700 meters, and the image is 1000x1000 pixels. (Click here for full screen image.)
Figure 6. View of the stream heads and plateau in the lower left corner of the TM image. The elevation range is about 0 to 150 m, and the image is 950x950 pixels. (Click here for full screen image. For a plan view of the image, click here.)
Figure 7. View of the streams in the upper central portion of the TM image. The elevation range is about 1000 to 3600 m, and the image is 800x800 pixels. (Click here for full screen image.)
The main results of this project are the 3D images. These renditions of the plateau should allow qualitative judgements of the stream catchments and how they seem to propogate. Unfortunately, the resolution is not high enough to determine anything conclusively. One can see gorge tips that may be knickpoints in several streams, but without data regarding the times of exposure and lacking time sequence images on a long time scale, it is not possible to state how quickly the knickpoints and gorge heads are propagating with any detail or assurance.
The problems associated with this project are as follows. The warped DEM images are difficult to rotate and manipulate within ENVI, and despite the retention of the higher resolution of the Landsat TM image, it is still difficult to resolve detail in the images. Stretches that were applied to the images (linear, Gaussian, decorrelation) did not improve the ability to discern lithology. Unfortunately, obtaining ground truth was not feasible for this project. Because of this, it is hard to tell whether the 3D images are accurate, and hard to make judgements as to the character of the streams and the bedrock.
The central eastern region of the Ethiopian Plateau, which sits to the west of the Ethiopian rift zone, has a surface that is constrained by volcanism and tectonics, which should allow us to infer initiation of erosion in a stream area. Using DEM and Landsat TM images, 3D images can be made of the region for a qualitative view of the streams. However, real results require ground truthing and quantitative modeling of the system.
The warped DEM would be an excellent tool to determine the stream flow patterns of the catchment system in the images. Densities of flow could be determined using the ENVI application, and stream profiles could be made using the DEM. Quantitative analysis of stream profiles is necessary in order to contribute to the study of bedrock-eroding streams.
African Studies WWW, U. Penn, http://www.sas.upenn.edu/African_Studies/AS.html
George, R., Roger, N., Kelley, S. 1998. Earliest magmatism in Ethiopia: evidence for two mantle plumes in one flood basalt province. Geology 26(10):923-926
Tinkler, K.J., Wohl, E.E. 1998. A primer on bedrock channels. In Rivers Over Rock: Fluvial Processes in Bedrock Channels, ed. K.J. Tinkler, E.E. Wohl, pp. 1-18. Washington, DC: American Geophysical Union
Weissel, J.K., Malinverno, A., Harding, D.J., Karner, G.D. 1995. Erosional development of the Ethiopian Plateau of Northeast Africa from a fractal analysis of topography. In Fractals in Petroleum Geology and Earth Processes, ed. C.C. Barton, P.R. La Pointe, pp. 127-142. New York: Plenum Press
WoldeGabriel, G., Aronson, J.L., Walter, R.C. 1990. Geology, geochronology, and rift basin development in the central sector of the Main Ethiopian Rift. Geological Society of American Bulletin, 102:439-458
http://www.tanzania-web.com/www.africanet.com/africanet/country/ethiopia/climate.htm
