The Socio-Economic Downstream Impact of Large Dams: A Case Study from an Evaluation of Flooding Risks in the Benue River Basin Downstream of the Lagdo Dam (Cameroon)

* Geographer, University of Ngaoundere (Cameroon), P.O. Box 553 Ngaoundere. E-mail: tchotsoua@yahoo.fr

** Geomatician at the Mission d’Etudes pour l’Aménagement et le Développement de la province du Nord (Cameroon). E-mail: aboubakmous@yahoo.fr

*** Geographer, University of Orleans (France). E-mail : jean-marie.fotsing@univ-orleans.fr

Abstract. Based on a combination of data from different sources in a spatial database, this article aims at evaluating flooding risks downstream of the Lagdo Dam. The compilation of this type of geographic information relies on the comparison of morpho-hydrologic data with socio-economic ones. Using this database shows that during a year with normal rainfalls, the floods reach the height of 187 m. An area of about 80,000 hectares, including 69 villages of different sizes, is flooded. In a year when rainfalls are excessive, the floods reach a height of 195 m and cover an area of 130,000 hectares, increasing the number of villages affected to 141, with a population of approximately 123,000. From an analysis of maximum daily rainfalls and the corresponding flooded surfaces, it follows that a height of 200 m can only be reached with exceptional rainfalls or the rupture of the dam. In such cases, the flooded surface would be close to 160,000 hectares, with a total of 193 villages and about one third of the town of Garoua underwater. The dry season agro-pastoral areas would equally be affected. In the event that the protective levees of the western pass are breached, the waters would cover a heavily populated area of about 950 hectares. Some important recommendations are made from the above analyses.

Keywords: Benue, Cameroon, Floods, Planning, River Basin, Risks.

Résumé. A partir d’une combinaison d’informations multi-sources dans une base de données à référence spatiale, cet article évalue les risques d’inondation en aval du barrage de Lagdo. Le montage de ce Système d’Information Géographique repose sur la confrontation des données morpho-hydrologiques avec les données socio-économiques. L’exploitation de cette base de donnée montre qu’en année de pluviométrie normale, les inondations atteignent la cote 187 m. Une surface de 80 000 hectares environ, comprenant 69 villages de tailles différentes, est inondée. En année de pluviométrie excédentaire, les inondations atteignent la cote 195 m et couvrent 130 000 hectares, avec 72 villages s’ajoutent aux précédents. La population affectée avoisine alors 123 000 âmes. De l’analyse des hauteurs de précipitations journalières et des correspondances avec les surfaces inondées, il ressort que la cote 200 m ne peut être atteinte qu’en cas de crue exceptionnelle ou de rupture du barrage. Dans ce cas, la surface inondée avoisinerait les 160 000 ha. Ainsi, un total de 193 villages et près du tiers de la ville de Garoua seraient sous les eaux. Les secteurs agropastoraux de saison humide seraient également bien affectés. En cas de rupture de la digue de protection du col ouest, les eaux couvriraient une superficie d’environ 950 ha en cours d’urbanisation. De ces analyses, quelques recommandations prioritaires sont énoncées.

Mots-clés: Aménagement, Bassin versant, Inondation, Risque, Bénoué, Cameroun, SIG.

Introduction

The last century saw a rapid increase in the construction of large dams. By 1949, about 5,000 large dams had been built worldwide, three-quarters of them in industrialized countries. By the end of the 20th century, there were over 45,000 large dams in over 140 countries. The period of economic growth following the Second World War saw a phenomenal rise in the global dam construction rate, lasting well into the 1970s and 1980s. At its peak, nearly 5,000 large dams were built worldwide in the period from 1970 to 1975. The decline in the pace of dam construction over the past two decades has been equally dramatic, especially in North America and Europe where most technically attractive sites are already developed. The average large dam today is about 35 years old (Kader Asmal, 2000). Dams have been promoted as an important means of meeting perceived needs for water and energy and as long-term, strategic investments, which have many additional benefits. Some of these additional benefits are typical of all large public infrastructure projects, while others are unique to dams and specific to particular projects. Regional development, job creation, and fostering an industry base with export capability are most often cited as additional considerations for building large dams. Other reasons include creating income from export earnings, either through direct sales of electricity, or by selling cash crops or processed products from electricity-intensive industries such as aluminum refining. Water-rich countries such as Cameroon have developed large dams for hydropower generation where suitable sites were available. The majority of large dams in Cameroon are for power generation and irrigation and indeed, single-purpose hydropower dams are most common in the country. While dams can contribute to economic growth, the services they provide may come at a cost. This dilemma is exemplified by the Lagdo Dam. Thanks to the technical and financial contribution of the People’s Republic of China, Cameroon built the Lagdo Dam on the River Benue in 1982 (fig. 1). This dam made possible electricity production and irrigated rice-growing downstream. Before its construction, the central part of the valley, which was very flat, used to be affected by the natural rhythm of flooding between July and October each year. The population had exploited this situation to elaborate a system of land use which alternated subsistence agriculture (growing sorghum) and livestock-raising between the flooded lowlands and the highlands, according to a specific hydro-agricultural calendar. This system of production has now been disturbed mainly by the modifications of the hydrological cycle caused by the Lagdo Dam as well as by demographic pressures. Furthermore, during periods of heavy rainfall, the release of water associated with the Mayo Kébi, a tributary of the Benue River, causes flooding which sometimes has an enormous impact on crops and infrastructures (dwellings, granaries, bridges, wells, and roads) as well.

Figure 1. The major Benue River bed downstream of the Lagdo Dam (Cameroon)

The objective of this study is to evaluate the risks of flooding. In this case, risk is defined as the combination of hazard–the probability of causing damage to both property and humans—and vulnerability—the extent of the damage or losses resulting from the occurrence of a natural phenomenon (Flageollet, 1989, quoted in Bechler-Carmaux et al., 2000)—of a given site and at a given moment in time (ISMAP, 1993). The authors used spatialised data (maps, satellite images and field data collected with the Global Positioning System [GPS]) in conjunction with socio-economic information. They then present the context of these studies and explain the syntheses of methods used to combine spatialised data originating from different sources in a GIS in order to evaluate the real risks of flooding in the Benue River Valley.

Context of the study

1.2. Climatic and morpho-hydrolic characteristics

The Benue River in its Cameroonian and Chadian course drains a basin slope of 97,000 km², 60,000 of which are in the southern part of the Garoua parallel; the remainder lies in the North and the East. Its total length from the rise to the junction with the Faro River is 350 km, 220 km of which are upstream of the dam and 130 km are downstream. Its major river bed is an extended flood plain constituted by lateral basins upstream of Garoua, and then of a long, narrower flood plain between Garoua and Malape. In addition, a new flood plain has formed at the junction of the Benue and Faro Rivers. The average slope downstream of the Lagdo Dam is so gentle that the river meanders extensively between sandy and poorly-stabilised beaches. This flat topography is partly responsible for large accumulations of alluviums, which are essentially argillaceous (Muller and Gavaut, 1996). The climate of the region is characterized by two strongly contrasting seasons. One is a dry season, which runs from November to May, and the other a shorter rainy season which goes from May to October (Olivry, 1986). Moreover, the area is generally characterized by exceptional years, which vary between extensive rainfalls and drought, and this particularly influences the hydrologic situation of the Benue River and its environs. In 1997, rainfall was very low (690 mm); field enquiries reveal that vast stretches of fertile flooding plains remained above the water during the whole season. Only the lowest parts of the valley were flooded. That dry year was followed by a year of heavy rains (1350 mm) during which the filling of the dam reached the height of 216 m for the first time. The dam water releases associated with the floods of the intermediary basins and of the Mayo Kébi River have dangerously affected the infrastructures downstream. The report by the Bureau d’Etudes Consultants Associés has distinguished three different kinds of damage amounting to 200 millions of F. CFA (13 000 €) in repair costs. Breaches on the embankment, regressive erosion of the downstream wall of the main dam, and destruction of infrastructure (roads and houses) (Consultants Associés, 1997).

Photo 1. Flooding of the main embankment near Garoua Bridge.

MEADEN’negative, October 1999

Photo 2. Houses destroyed in Bamé village by the flood of October 1999.

Ngounou Ngatcha’negative March 2004

The year 1999 was also one with heavy rainfall, with rains reaching a total of 1200 mm. A report written by the Lagdo sub-district officer helps us to better appreciate the extent of the disaster:

From 9th to 12th, September 1999, 805 x106 sq m of water had to be released with the only worries to limit the height of the water in the dam, due to the fact that very large amounts of water were recorded in the dam: 1304 x 106 m³ for that period. If that quantity had not been released, the amount of water stocked in the dam would have been 5038 x 106 m³ on 21 September 1999, which corresponds to the Maximal Alarm Zone (MAZ) of 216.65 m, by far above the nominal value of 216.00 m. The opening of the floodgates up to 10 m and of the tunnel up to 7.16 m permitted the evacuation of a total flow of 2316 m3/s, which stabilised the height upstream at 217.04 m.

At the Garoua Bridge, the embankment was completely flooded (photo 1). At Kokoumi-Dargala, Pitoa, Padang, Bamé, Pabla, Ndamboutou, and Taipée, the floods caused the loss of crops (corn, sorghum), destruction of houses (photo 2), and displacement of people away from the river. We have been able to reconstitute the height reached by those floods thanks to the GPS written landmarks provided by the populations of the villages which suffered these disasters.

These transformations (flooding, drought, and erosion) upstream of the Lagdo Dam are narrowly related to the construction of that work as was underlined by the Executive Governor of the Adamawa State (Nigeria), His Excellency Alhadji Saleh Michika, in a speech given in May 1993 during a friendship visit to the Northern Province of Cameroon. He noted that “Since the beginning of the repercussions in the Adamawa State: periodic flooding of the cultures, the silting of the riversides, the erosion of the banks, having as a consequence the destruction of whole villages which are along the river” (Mengue Mbom, 1995). All these transformations necessitate the establishment of an evaluation method of flooding in the area for improved follow-up of the phenomenon.

2.1 Data and methods

To determine the sectors prone to risk requires gathering large quantities of complex data. The management of these data, which are collected on changing environments and which are continually shaped by resourceful human inhabitants change, can be done only with the help of GIS.

The data used for one such study come from different sources:

• one topographic map of Garoua at 1/200,000 ;
• eight topographic maps at 1/50,000;
• three ASTER images made in January 2001;
• GPS positions of villages and socio-economic infrastructures;
• socio-economic data obtained by field interviews.

This information is assembled in a GIS database (fig. 2). In order to analyze images, a combination of methods including oriented classification and algorithms of a maximum likelihood were used. Classification was based on the acquaintance with the land and was done by printing the different plots using a coloured coding system. This classification was applied to all the channels for whichimages were considered.

Figure 2. Channel of collection and data treatment

The coloured coding consisted of a combination of three channels for the synthesis of a maximum of contrast. In so doing, a primary colour (Blue, Green, Red) was assigned to the display of the content of each channel according to the synthetic representation of the additive model trichrome RGB (Red, Green, Blue). This process allows for the identification of the main spatial components of the primary Benue River bed environment. The colour-coded map was superposed to the topographic map to determine the contour lines of the areas reached by the waters, as indicated by people in the field. This technique permitted the calculation of the maximum surface which would probably be covered by the waters at the maximal amount of the rainfall in each of the years considered.

The GIS database we used is made of layers of information. Each layer is made of many entries to which alphanumerical data are connected. Thus the database consists of two groups of data depending on:

1. The use and location of land (village, town, rainy season agro-pastoral spaces, dry season agro-pastoral spaces, orchards, road networks, open, uncultivated spaces, and/or spaces made water-repellent, etc.) extracted from the satellite images completed by GPS surveys and field enquiries.
2. The physical milieu (relief, hydrographical network, stagnant pools and flooding zones) extracted from the topographic map, which was completed with GPS surveys.

In this global methodological diagram, interactive cartography is prioritized as a tool for the acquisition of knowledge and management of floods, and means for the prescription and information of riverside dwellers and the larger public.. This approach constitutes a more efficient framework which integrates economic, social, and ecological dimensions into the evaluation of the options through which planning and implementation of the project may take place. It is the very notion of flooding risks which necessitates the making of a secure corridor along the Benue River downstream of the Lagdo Dam.

3.1 Primary results and discussion

The results of this analysis are presented in simple maps which will facilitate ongoing discussions by users of the water from the Benue River. Maps with zones of flooding risks are established on the basis of 1999 rainfall, as compared with the normal rainfall of 2001. Three main types of flooding risks are then established in relation with their main causes. One risk is related to breaches of the western embankment. As for the risks of floods related to the rainfalls, an equation permitting us to predict the flooded surface along the Benue River in terms of the daily rainfalls was established. Flooding is being considered here as the consequence of the accumulation of daily rainfalls. During a year, rain is accumulated daily in the main collectors and when the rains reach a critical level, the river beds overflow and invade the main river bed. The critical levels, the period (day and month), and the flooded surface have been reconstituted for the years 1999 and 2001. The figures below present the monthly accumulated records of daily rainfall for those two years.

Figure 3. Accumulation of monthly rainfall in 1999

Figure 4. Accumulation of monthly rainfall in 2001

On October 13, 1999, an accumulation of daily rainfall of 1.187 mm was recorded in Garoua, which corresponds to a reconstituted flooded surface of 130,000 hectares. In October 2001, the maximum accumulated daily rainfall was 870 mm and this amount caused flooding.

Table 1. Consequences of flooding in terms of heights reached by the water surface

Designation	Year 1987 (Normal rainfall)	Year 1999 (Heavy rainfall)	Simulation	Total
Contour line reached by water (in meters)	187	195	200
Flooded surface (in hectares)	80,000	130,200	159,700
Villages affected by the floods	69	72	53	194
Population affected by the floods	67,637	58,898	173,553	300,088
Schools affected	30	36	17	83
Dispensaries affected	6	9	6	21
Water spots affected	35	18	17	70
Markets affected	40	45	34	119

Source: Field enquiries, GPS survey and treatment of satellites images of 2001 and 2004

Figure 5. The main Benue River bed: Risk zone of flooding in case of normal annual rainfall

Sources: Topographic Map of Garoua NC-33-VIII (1973), Satellite Images (2001), GPS data collected and field enquiries (April–May 2004). Compilation: M. Tchotsoua, July 2004.

a) During a year of normal rainfall, floods reach the height of 187 meters (fig. 5). An area of 80,000 hectares is flooded, containing 69 villages of different sizes and with a total population of 67,637 (table 1). In villages situated on hillocks and islets, the houses and socioeconomic infrastructure are not affected. While the waters are judiciously exploited for out-of-season cultivation and planting, there may be damage due to flooding, even in situations of normal rainfall. Such is the case with the Benue public primary school in Garoua, where, every year since 1986, some classrooms are not accessible due to high water levels and can only be occupied at the end of November.

Figure 6. Zone with flooding risks and erosion at the height of 195 m in the main Benue River bed.

Sources: Garoua topographic Map. N.C 33- VIII (1973), satellite images (2001), GPS surreys and field enquiries (April–May 2004). Compilation : M. Tchotsoua, July 2004.

b) During years with heavy rainfall such as 1999, flood waters reached the height of 195 m (fig. 6), and covered about 130,200 hectares, with 72 villages (table 1), bringing their number to 141 affected villages. The population exposed to the risk was approximately 126,535 individuals. At this level, some houses situated on hillocks within the contour wells are the most numerous and are filled up with mud. Market attendance is greatly affected, which has a negative impact on the economy and, indirectly, on the fight against poverty and underdevelopment in the region.

Figure 7. Zone with flooding and erosion risks in case of exceptional floods or bursting of the dam in the main Benue River bed.

Sources: Garoua topographic Map NC-33-VIII (1973), satellites images (2001), GPS surveys and field enquiries (April–May 2000). Compilation: M. Tchotsoua, July 2004.

c) From the analysis of maximum daily rainfall and its correspondence with flooded areas, the height of 200 m can only be reached in the cases of excessively heavy rainfall or the bursting of the dam. In such cases the flooded surface would be 159,700 hectares (fig. 8). Thus, 53 more villages, with an approximate population of 173 553, are also affected (table 1). Thus, there will be a total of 194 villages and a population of 300,088 persons whose infrastructures and/or farms will be exposed to flooding. It must be specified that certain villages situated out of the flood plain are also impacted because their fields are located in the flood-prone corridor. Nearly one-third of the Garoua town could be flooded (fig. 7).

There are three release galleries at the level of the Lagdo Dam. Three distinct embankments exist. These include: 1) the main one which opens directly in the valley; 2) the eastern one which also opens in the Benue River but with a diversion for irrigation purposes in the rice-growing zones around Gounougou, with a maximum capacity of 14 m³/s; and 3) the western one which opens on a channel stretching over about 5 km before rejoining the Benue River valley at the level of Pitoaré.

Among these three release galleries, the most dangerous is the western one. It is made of earth (predominantly clay), and is approximately 1,257 meters long, with a maximum height of 6.5 meters. If this embankment were to burst, water would flow toward the major Benue River bed passing through Pitoaré, thus covering a surface of about 932 hectares. The Lagdo town neighborhoods are built in that risky area, chiefly Madagascar Bobi Lagdo, Moundang, Ouro Doole, Mefere Lagdo, the public high school neighbourhoods, and Pitoaré (fig. 7). These neighborhoods will be literally engulfed. Considering the fact that the bursting of the dam cannot be dissociated from rainfall, the contour line of 200 meters could be reached downstream, but only for a brief period since not all the water in the dam will need to be released. At the level of the embankment the maximum water height that will have to be released from the dam is estimated at 2 meters.

Conclusion

While many have benefited from the services which the Lagdo Dam provides, its construction and operation have had considerable negative societal and environmental consequences. The adverse effects on populations include displaced families, host communities where families are resettled, especially those downstream of the dam, whose livelihood and access to resources are affected in varying degrees by altered river flows and ecosystem fragmentation.

This study of risks of flooding of the main Benue River bed clearly shows the physical characteristics of the environmental, human, and socioeconomic factors at stake. Flooding is related to larger than normal water flow and, as a consequence, we are more concerned with natural phenomena than with that of human management. Thus, periodic flooding is mainly explained by the return of heavy rains after years of drought during which time the valley was attractive to different groups of migrant for agricultural and pastoral occupations. Land cleared because of intensive agricultural and pastoral activities favours surface flows of water which contribute to flooding, particularly for the peasant communities, which periodically suffer from the downstream overflows of the Lagdo Dam.

To manage such risks does not just mean simply consulting the tables and making comparisons, or applying mathematical formulae. The risky zones are defined as presented on the maps, and then must be explicitly treated in terms of activities, which include obtaining opinions of the populations, and recording infrastructures in the major river bed. These risks and activities are determined after a combination of concepts and tools which many scientific disciplines—such as hydrology, hydraulics, ecology, study of soils, agropastoralism, and geomatics—contribute in order to develop the ways and means for protection of this zone against flooding.

These results allow the population to better organize their activities in terms of seasonal agricultural work. They also help decision-makers and different actors, chiefly the AES-SONEL, to manage the future of this highly important valley for the development of the Northern Province, in particular, and of Cameroon, in general. Finally, this research makes available permanent spatial information, including both qualitative and quantitative data, about the environment in its different forms.

The Benue River bed downstream of the Lagdo Dam remains an underexploited agro-pastoral area, because all present planning is concerned with less than 30.000 hectares, whereas the useful area could be increased fourfold with new equipment and improved hydraulic infrastructure. It is important to rehabilitate this agro-pastoral resource, which is all the more important as the climatic changes (e.g., rainfall in certain years and repeated droughts during others), the demographic pressure, and techniques of production not adapted to the new hydro-ecologic situation) have generated a profound ecological imbalance. This situation has contributed to phenomena including silting up of the river bed, destruction of aquatic habitats, hydrological erosion, dumping of refuse, and pollution by urban centres.

Such rehabilitation may be accomplished through:

1. The placement of boundary-markers indicating sectors with flooding or erosion risks along the main Benue River bed, from the Lagdo Dam to the Nigerian frontier;
2. The use of water in the Lagdo Dam according to controlled filling contour lines;
3. The increase of the number of pluviometric and hydrometric stations for the systematic remote collection of data and follow-up in the Benue basin;
4. The installation of an alert system providing radio broadcasts in the local language at the Benue FM and at Radio Garoua;
5. The development of a system of participatory water management and the supervision of agro-pastoral spaces in the Benue valley; and
6. The promotion and popularization of the practice of out-of-season cultivation, taking into account the hydro-agricultural calendar.

These recommendations represent a compromise between what is socially desirable, economically viable, technically possible, and ecologically acceptable. It is not a matter of a new intellectual imperialism which imposes its own perspectives and solutions on those who experience the reality of daily life in the area.

References

Bechler-Carmaux N., Mietton M., Lamotte M., 2000. Le risque d’inondation fluviale à Niamey (Niger). Aléa, vulnérabilité et cartographie. Ann. Géo., n° 612, pp. 176–187, Armand Colin.

Consultants Associés, 1997. Les aménagements hydroagricoles dans la vallée supérieure de la Bénoué. Analyse dignostic. Rapport provisoire, MINAT/MEAVSB/FED.

Gavaud M. 1975. Les sols de la vallée de la Bénoué de Lagdo au confluent du Faro. Tome A, Orstom, 60 p.

IRAD – Garoua, 2001. Etude diagnostique des aménagements hydroagricoles de Lagdo, 77 p.

ISMAP, 1993 ISMAP, 1993. Cartographie des zones à risques. Projet EUREKA EU, 479, Phase de définition.

Kader Asmal, 2000. Dams and Development: A New Framework for Decision-Making. The report of the world commission on dams. Earthscan publications ltd, 404 p.

Muller J.P., and Gavaud M., 1976. Conception et réalisation d’une carte d’aptitudes culturales de la vallée de la Bénoué. Cah. Orstom, sér. Pédol., Vol. XIV, n° 2, 131-159.

Naah E., 1981. Profil de la Bénoué en aval de Lagdo. Rapport DGRST/IRGM/CRH, 57 p.

Olivry J. C. 1986. Fleuves et rivières du Cameroun. Collection « Monographie Hydrologiques ORSTOM » n° 9, Paris

Pornon H., 1998. Système d’information géographique, pouvoir et organisations. Géomatique et stratégies d’acteurs. IETI et L’Harmattan, 255 p.

SCERTAGRI-SOGREAH, 1984. Plan directeur de l’aménagement de la vallée supérieure de la Bénoué, Phase 1, annexe 1, Milieu physique, 146 p.

TACTEBEL ENGINEERING, 1992. Gestion des eaux du Barrage de Lagdo. Expertise pour la définition d’un cadre de gestion, 38 p.

Tchoue G., 1983. Propagation de l’onde de crue de la Bénoué en aval de Lagdo. Délégation Générale de la Recherche Scientifique et Technique. Institut de Recherches Géologiques et Minières. Centre de Recherches Hydrologiques. 174 p.

Tillemant B., 1966. Etude hydrologique de la région du confluent Bénoué-Kébi, Bureau de l’Eau de Garoua, 19 p.