Satellite Remote Sensing: Monitoring the Omo River Delta in East
Africa Using Remote Sensing
Changing environment conditions of the Omo River Wetland are measured using remote sensing.
By Barry Haack and Joe Messina

A problem common in environmental management, particularly in less developed countries, is the inadequacy of information on current and changing environmental conditions and the available resource base. Without accurate information managers often fail to make decisions or make incorrect decisions. One tool useful in providing current, reliable environmental and other resource information is remote sensing. Historically, this was primarily aerial photography which together with ground surveys was used to obtain earth surface information. Aerial photography is an excellent source of environmental information. It is, however, expensive, especially for large areas and in developing countries. Generally, the photography is often of inconsistent quality, out of date, or in many cases, does not exist at all.
     During the last two decades, data acquired by satellite-borne sensors have become available and been applied in many environmental studies. Advantages of spaceborne remote sensing are that it systematically and frequently acquires information for areas which are difficult to access, provides a synoptic view of large features in a manageable number of images or photographs, and maintains a permanent record of conditions at the time of acquisition. These permanent records are extremely useful for monitoring the extent, type, and location of environmental changes. Data collected by the different global satellite sensors currently in operation vary in resolutions and other characteristics, including costs.
     Globally there has been an increase in concern over environmental degradation and the need for greater environmental protection and management. Those concerns have varied in scale from global climatic change to small specific chemical waste sites. All ecosystems of the Earth are potentially affected by man's activities, but wetlands are especially fragile and often neglected. Wetlands are not well understood nor appreciated and have been increasingly under natural and human pressures in all parts of the globe and especially in Africa. Knowledge on the locations, characteristics, and changes of the wetland communities must be available to establish protection or management policies. Remote sensing and GIS are useful for assessing and monitoring these ecosystems. This article presents the results of a study which used spaceborne remote sensing together with aircraft reconnaissance flights and GIS modeling to examine recent changes in the Omo River Delta Wetland in Eastern Africa.

Omo River Wetland
The largest closed-basin lake in the East African Rift is Kenya's Lake Turkana (Figure 2). Lake Turkana, formerly known as Lake Rudolph, is 250 km long and has a maximum width of 50 km. The surface area of the lake is about 7,500 sq. km.
     The lake is famous for its color. It is sometimes referred to as the Jade Sea, which is largely due to the presence of blue-green algae in the phytoplankton community. Today, however the lake color, especially in the north, is brown because of sedimentation. The drainage basin for the Omo River is about 146,000 sq. km and is primarily within Ethiopia. There is little overland surface run-off into the lake from the surrounding watershed and no outlet from the lake. Most of the influx into the lake is from three rivers: Kerio, Turkwel, and Omo. The Omo River flows 1,000 km from Mt. Amara in the Ethiopian Highlands west of Addis Ababa into the northern end of Lake Turkana. At the Omo River entrance to Lake Turkana, there has developed a highly complex and spatially fluctuating floodplain and delta.
     The delta was extensively studied in the 1960s by Karl Butzer and a team of colleagues from the Department of Geography at the University of Chicago, but has been isolated due to regional political instability and access issues.
     In 1971, Butzer estimated that the lake has varied by 20m during the last 100 years from his base line in the late 1960s. It was 15m higher at the end of the 19th century and 5m lower in the 1950s. These changes were climatic and to a lessor extent geomorphic in origin. This study extends the work of Butzer from the late 1960s. The changes since then are more dramatic and more likely to be anthropogenic.
     Understanding the current conditions and changes of the Omo Delta Wetland is significant for a number of reasons. Of greatest immediacy may be effects on the local human populations and changing biodiversity. An increase or decrease in natural vegetation and perhaps wildlife populations may require some level of management and protection. Of broader interest, however, may be what is causing deltaic changes and what are the environmental implications for those causes. Certainly, the delta growth may be an indicator of regional or global climatic changes. Spaceborne remote sensing was used as a first indicator of deltaic changes subsequently enhanced by low altitude aircraft overflights.
     Three dates of Landsat MSS images for the Omo Delta area were initially obtained. Those dates were Feb. 1, 1973, Jan. 1 1979, and March 1, 1989. Of the four spectral wavelengths in which these data can be acquired, the longest wavelengths, which are in the near-infrared, 0.8-1.1 micrometers, are very effective for separating land from water. In these wavelengths, water has virtually no reflectivity so it is very dark on an image and green vegetation or dry soils reflect very highly and are light toned. For this study, this single band black and white imagery provided an appropriate data set at minimal costs. This is Band 7 on Landsats 1, 2, and 3 and renumbered as Band 4 on Landsats 4 and 5.
     The single band imagery was obtained as 18.5 cm film positives at a scale of 1:1,000,000. A subscene, of about 5 cm per side, was identified on each image for the Omo Delta. Those subscenes were enlarged 20 times to produce photographic prints at a scale of 1:50,000. On these images, a delineation was made between land and water to map the extent of the delta. Figures 4 and 5 are copies of the 1973 and 1989 single band subscenes.
     A GIS was used to determine the size and changes of the delta over time. Each map of the delta from the three Landsat dates was digitized into a GIS. One of the difficulties was the absence of clearly identifiable points, such as cultural features, in the imagery to register them either to a geographic grid or to each other. Portions of the rivers and the lake shore were used to register the maps. This provided a set of three spatially registered GIS layers of the delta for analysis.
     Table 1 contains the growth of the delta from the information extracted from the Landsat imagery and measured with the GIS. The growth is tremendous, mostly occurring since 1979. Figure 3 is a GIS derived map illustrating the areas of expansion.
     The Landsat imagery provided information on the extent of the deltaic growth but fails to provide clarification as to what are the land covers and land uses on these new areas. Examination of a variety of spaceborne imagery including hand held shuttle photographs, Landsat MSS color composites, and a very recent (1994) SPOT panchromatic image illustrated the growth to be a function of both changing water levels in the lake and increased sedimentation.
     The spaceborne data only provided generalized, speculative information on conditions within the delta. To obtain more detailed information, a low altitude reconnaissance flight was conducted in November 1994. This flight confirmed that much of the delta is green vegetation and not floating or submergents but grasses, some on the order of 1 to 2m in height. There are also trees and shrubs evident near the river distributaries, perhaps on slightly elevated natural levees. As speculated from the spaceborne imagery, there are areas of dry exposed soils on the older portions of the delta and large areas of open water. Figures 6 and 7 are low altitude airphotos of portions of the delta.
     The most important observed feature was the obvious utilization of the delta by indigenous tribes. The primary tribe in the delta is the Dasanech, also known as Merille. The economy of the Dasanech is based on stock-raising, agriculture, and to a lessor extent, fishing.
     An important question is what has caused the growth of the delta? The lack of scientific data collection in this area makes it difficult to document answers to this question. Basically, a delta grows as a function of either a decrease in lake level or an increase in sedimentation or, as it appears to be the case in the Omo, a combination of these two factors. Another important consideration is to what extent is the delta growth anthropogenic and to what extent due to natural factors, or what combination of these factors.
     The growth of the delta appears to be primarily a result of decreased lake levels. This may be a result of increased aridity and/or decreased flow into the lake. Both a decrease in precipitation within the catchment basin and an increase in temperature can be contributing factors. An increase in temperature would increase evaporation rates from the surface of the lake and elsewhere in the region, decreased soil moisture would lessen run-off. The decline in lake level is also a result of decreased river input caused by cultural practices. Along the Omo River in Ethiopia, there has been an extensive increase in small irrigation schemes diverting water from Lake Turkana. Alexander (1990) estimates the Omo River discharge has been reduced by 50 percent because of these activities.
     The lack of a scientific infrastructure together with rapid changes in the region makes establishment of trends and causes difficult. A recent visit to the lake in November 1994 made it very clear there have been lake level declines. There are visible changes to the shoreline. Fergusons Gulf on the west shore is now an enclosed lake because lowered lake levels have exposed the inlet bottom. Discussions with residents of the area did, however, indicate that the lake level has risen by 1.5m over the last two years. In parallel to the decrease in lake level, there has also been an increase in siltation for some of the same reasons. Increased population pressures in the drainage basin of the Omo River in Ethiopia has caused increased soil erosion by removal of the vegetative cover for fuel and conversion to agricultural lands. Imagery from Landsat and other spaceborne sensors clearly show the heavy sedimentation flow into the lake.
     The Omo River Delta has grown extensively in the last 15 years. The causes of the growth are complex and to some extent uncertain. The expansion can be viewed from two different perspectives and provides both reactive and proactive opportunities. Certainly, this can be considered a negative environmental situation. It may be an indication of regional climatic changes and/or the result of rapid and perhaps destructive changes in land use patterns in the drainage basin. However, the expansion of the delta may create some unique environmental opportunities and may have some important positive results. In a world where wetlands have been rapidly decreasing, here a large wetland has been created in an otherwise arid landscape. The expansion of the delta wetland is potentially maintaining or increasing the biodiversity of fauna and flora, both locally and regionally. However, the good soils, lush vegetation, and availability of water are also attracting permanent human populations, most likely in conflict with the fauna and flora. Much more information needs to be obtained for this area including measurements of precipitation, lake levels, river flows, populations, both human and animal, and land use/land cover. This information is needed to make effective decisions on management and protection options. The delta may be a possible location for a wildlife preserve or sanctuary but that protection will require a joint management strategy recognizing the needs and existing presence of the indigenous human populations. The creation and implementation of management strategies for the delta is complicated by its international location. The delta has historically been considered to be Ethiopian but recent growth has extended it into Kenya. However, land use practices in Ethiopia greatly affect the delta and must be considered in any management perspective. Appropriate environmental policy requires accurate information on current conditions and the locations and rates of change. Spaceborne remote sensing and GIS techniques are very effective methods to obtain, analyze, and display such information. Unfortunately, good information does not always result in good policies, those decisions are controlled by many factors and pressures, but good information is a necessary beginning to good policies.

Acknowledgments:
The data acquisition and analysis conducted by Jeanette Baier while a graduate student at George Mason University is greatly appreciated. Derek Dames and Allen Binks were extremely helpful in providing the aircraft overflights over the Omo Delta. Halewijn Scheuerman of Nairobi shared many observations on the delta and Ken Ferguson of Teleos International generously provided the 1994 Landsat TM image of the delta downloaded at their new satellite receiving station in Nairobi.

About the Authors:
Barry Haack is professor of geography at the George Mason University in Fairfax, Va. He may be reached at 703-993-1210. Joe Messina is a GIS/RS scientist for SPOT Image Corp. in Reston, Va. He may be reached at 703-715-3135.

Back