Abstract

River basin rainfall series and extensive river flow records are used to characterize and improve understanding of spatial and temporal variability in sub-Saharan African water resources during the last century. Nine major international river basins were chosen for examination primarily for their extensive, good quality flow records. A range of statistical descriptors highlight the substantial variability in rainfall and river flows [e.g., differences in rainfall (flows) of up to −14% (−51%) between 1931–60 and 1961–90 in West Africa], the marked regional differences, and the modest intraregional differences. On decadal time scales, sub-Saharan Africa exhibits drying across the Sahel after the early 1970s, relative stability punctuated by extreme wet years in East Africa, and periodic behavior underlying high interannual variability in southern Africa. Central Africa shows very modest decadal variability, with some similarities to the Sahel in the adjoining basins. No consistent signals in rainfall and river flows emerge across the whole of the region.

An analysis of rainfall–runoff relationships reveals varying behavior including strong but nonstationary relationships (particularly in West Africa); many basins with marked variations (temporal and spatial) in strength; weak, almost random behavior (particularly in southern Africa); and very few strong, temporally stable relationships. Twenty-year running correlations between rainfall and river flow tend to be higher during periods of greater rainfall station density; however, there are situations in which weak (strong) relationships exist even with reasonable (poor) station coverage. The authors conclude for sub-Saharan Africa that robust identification and attribution of hydrological change is severely limited by data availability, conflicting behavior across basins/regions, low signal-to-noise ratios, sometimes weak rainfall–runoff relationships, and limited quantification of the magnitude and effects of land use change.

1. Introduction

a. The significance of water resources variability in Africa

Rainfall and river flows in Africa display high levels of variability across a range of spatial and temporal scales, with important consequences for the management of water resource systems (Sutcliffe and Knott 1987; Grove 1996; Laraque et al. 2001; Conway 2002; Ogutunde et al. 2006; Hamandawana 2007). Throughout Africa, this variability brings significant implications for society and causes widespread acute human suffering and economic damage. Examples of variability include prolonged periods of high flows for rivers draining large parts of East and central Africa (Conway 2002), and multidecadal anomalies in river flow regimes in parts of West Africa where long-term mean yields of freshwater into the Atlantic Ocean fell by 18% between 1951–70 and 1971–89 (Mahé and Olivry 1999). There are many examples of the challenges posed by water resources variability in Africa: Lake Chad fisheries (Sarch and Allison 2000), reservoir management on the Senegal River (Magistro and Lo 2001), balancing supply and demand for Nile water in Egypt (Conway 2005), irrigation management in the Greater Ruaha River in Tanzania (Lankford and Beale 2007), and hydropower generation in the Kafue (Sutcliffe and Knott 1987) and Lake Victoria basins (Tate et al. 2004).

As anthropogenic climate change becomes increasingly manifest, the prospect of shifts in flows and variability underscores the need for better understanding of the drivers of variability and rainfall–runoff interactions. It is likely that extreme events are going to be the greatest socioeconomic challenge. Although sub-Saharan Africa is generally associated with drought-related influences, anecdotally there appears to be greater frequency and spatial extent of damaging floods, particularly in East Africa and Ethiopia (e.g., 2006 and 2007). Extreme floods have caused substantial socioeconomic disruption in Mozambique (2000; Christie and Hanlon 2001) and East Africa (1961, 1978, and 1997; Conway 2002), whereas smaller floods may be somewhat overlooked but locally significant, for example, in Nigeria (Tarhule 2005). Late 2006–07 saw major floods of unprecedented spatial extent (and timing) across Somalia, Ethiopia, and other parts of East Africa, which is broadly in line with projections in the Intergovernmental Panel on Climate Change’s Fourth Assessment Report for increases in autumn and winter rainfall (Christensen et al. 2007).

The main driver of much of the observed variability in river flows is—of course—rainfall, particularly at the scale of large river basins [see Hulme et al. (2001) for African overview]. Above average and sometimes extreme rainfall in East Africa tends to be associated with periodic circulation dipole events in the Indian Ocean and complex interaction with the El Niño–Southern Oscillation (ENSO), particularly during the short October–December rains (Saji et al. 1999; Webster et al. 1999). The large decline in many West African river flows is primarily related to the effects of the prolonged drying in the Sahel (late 1950s–late 1980s), with conditions still drier than during the humid 1950s (L’Hôte et al. 2002; Dai et al. 2004)—though there are exceptions (Nicholson 2005).

Despite the large influence of rainfall fluctuations on river flow variability, the response may be influenced by other factors such as changes in land use or land cover for the Sahel/West Africa (Mahé et al. 2005; Li et al. 2007; Leblanc et al. 2008) and southern Africa (Troy et al. 2007; Woyessa et al. 2006; Lørup et al. 1998). Human abstractions and reservoir construction also play a role (Vörösmarty and Sahagian 2000; Hamandawana et al. 2007) along with land surface to atmosphere feedbacks (Savenije 1996).

Depending on the actual hydrological conditions, the effects of rainfall variability on the hydrologic response will generally translate into smooth and delayed responses in lake and wetland systems, whereas semiarid river basins often exhibit low runoff coefficients and high sensitivity to rainfall fluctuations (Nemec and Schaake 1982; Li et al. 2005; McMahon et al. 2007). In addition, the nature of the land surface itself may increase the variability of river flow responses to rainfall fluctuations. Peel et al. (2001, 2004) examined differences in the temporal behavior of rainfall and river flow between continents and showed that variability of annual river flows is higher for temperate Australia, arid southern Africa, and temperate southern Africa, than for other continents with similar climatic zones. In addition to rainfall variability, they found that the distribution of evergreen and deciduous vegetation in temperate regions was a potential cause of greater river flow variability. Vörösmarty et al. (2005) found interesting features of African river systems by combining biophysical and social datasets to show that population distribution is strongly concentrated in regions exposed to high levels of interannual variability in rainfall and runoff.

b. Aims

Although much evidence exists for high interannual and decadal variability in rainfall and river flows in sub-Saharan Africa, there are few detailed studies of their spatial and temporal covariability. Previous attempts to estimate runoff at this scale have been based on the limited data available through international bodies; none have undertaken detailed rainfall–runoff analysis at the basin and subbasin scale, or taken a long historical perspective to incorporate natural variability over decadal and century time scales. Rather surprisingly, sub-Saharan Africa provides one of the best opportunities globally to do this type of analysis because many of the very large river basins possess long, relatively natural river flow records. We combine extensive databases held by international and national agencies to build a comprehensive picture of hydrometeorological variability during the twentieth century in large river basins comprising roughly 32% of sub-Saharan Africa’s area. Our overarching aim is to contribute to the scientific understanding of variability in large water resource systems. First, we characterize the spatial and temporal dimensions of rainfall and river flow covariability across the region and second, we examine the characteristics and stability of rainfall–runoff relationships over time.

2. Data sets: Identifying river basins and regions for the analysis

a. Data sources

We use river flow observations recorded over the period 1901–2002. These were provided mainly by the Institute de Recherche pour le Développement (IRD) of Montpellier for the West and central African regions and from a range of international and national sources for the East and southern African regions. The data provided by IRD come primarily from the SIEREM database (UMR HydroSciences of Montpellier; Boyer et al. 2006; available online at www.hydrosciences.fr/sierem). The SIEREM project gathers hydrological and climatic data collected by national networks, various international organizations, and research bodies such as the Food and Agriculture Association (FAO) and IRD. IRD updated the data until 1980 (Ardoin-Bardin 2004). Thereafter, updates have been obtained within the framework of the United Nations Educational, Scientific and Cultural Organization’s (UNESCO’s) Flow Regimes from International Experimental and Network Data, West and Central Africa (FRIEND–AOC) project and cooperative undertakings with national agencies. Data for East Africa come primarily from Hurst (1933) for rivers in the Nile basin [Kagera, Lake Victoria outflows, Blue Nile, Equatorial Lakes, Sobat and Atbara rivers; see Conway and Hulme (1993); Sutcliffe and Parks (1999)], UNESCO (1995) for the Tana, and Hamandawana et al. (2005) for the Okavango. Other river flow series and some recent updates to the above series were obtained from respective national agencies.

For rainfall, we use the University of East Anglia’s Climate Research Unit (CRU) TS 2.1 0.5° resolution time series for 1901–2001 from New et al. (2001) and updated in Mitchell and Jones (2005). River basin boundaries upstream of gauging stations were delineated using the Shuttle Radar Topography Mission (SRTM) from the National Aeronautics and Space Administration (NASA)/U.S. Geological Survey (USGS) as the digital elevation model, and ESRI’s Digital Chart of the World drainage files to delineate the catchments. Basin rainfall series were calculated as the average of all 0.5°CRU TS 2.1 grid boxes within the basin boundary. Detailed information on CRU TS 2.1 quality control and notes on data interpretation can be found in relevant publications, but we note here that Africa has generally poor spatial and temporal coverage of rainfall stations (Hulme 1996; Nicholson 1996), and this is true for the CRU TS 2.1 data—particularly before the 1930s and after about 1980. We use gridded time series of the number of stations within range of a grid box (Mitchell and Jones 2005; available online at www.cru.uea.ac.uk/~timm/grid/stns.html) and calculate a basin average series from all grid boxes in the basin. Range is defined as the correlation decay length (450 km for rainfall), so that the series represent the average number of stations with data upon which the grid boxes in the basin may draw to calculate rainfall anomalies. Because these series do not record the actual number of stations that have been used to generate rainfall values, we concentrate on changes in their relative—rather than their absolute— number.

b. River flow records in major African river basins

The study concerns sub-Saharan Africa, divided for means of presentation and analysis into four regions: west, central, east, and south. Observations are used for 26 gauging stations in total, irregularly spaced between and within the regions (Fig. 1). The period of record and details of river gauge and basin characteristics are summarized in Table 1. The river basins range in size from ∼30 200 to 3 475 000 km2. We use three loosely applied criteria to select river basins for analysis, which are in decreasing order of priority: availability of long, verifiably good quality river flow record; large in area (>10 000 km2); and spatial coverage across sub-Saharan Africa. In some situations, we relaxed the criteria to maximize spatial coverage. Figure 1 shows that in general, the gauges represent the key upstream contributing areas of these large river basins and in some situations, combinations of upstream and downstream gauges (Niger and Ogooué). In all situations at annual time scales, the river flow records are—for the most part—unaffected by human influences in the form of upstream dams and major abstractions. Because of the construction of the Owen Falls dam in 1954, the outflows from Lake Victoria have been regulated to follow the natural relationship between lake level and outflows (an “agreed curve”; Tate et al. 2004). After a rise in the lake in the late 1990s, outflow departed from the agreed curve to alleviate flooding around Lake Kyoga downstream (Goulden 2006; Sutcliffe and Petersen 2007), but we only use data up to 1989. There is a major dam on the Senegal River, the Manantali dam, so we use reconstructed natural discharges for the downstream series at Bakel, Senegal (Bader 1990, 1992). The Niger River has some dams but these have only minor effects on the annual time scales used in this anlaysis. The discharge series of the Senshi Halcrow gauging point is influenced by the Akosombo dam mainly during the first years of filling (1964–67). Its interannual variability remains similar to that of neighboring rivers, but the monthly regime has been modified as a result of regulation and the total discharge has been reduced because of evaporation (Moniod et al. 1977).

Fig. 1.

The main drainage basins analyzed and location of the gauging stations (refer to Table 1).

Fig. 1.

The main drainage basins analyzed and location of the gauging stations (refer to Table 1).

Table 1.

General characteristics of the river basins included in the analysis (locations in Fig. 1).

General characteristics of the river basins included in the analysis (locations in Fig. 1).
General characteristics of the river basins included in the analysis (locations in Fig. 1).

Interestingly, West and East Africa are well served by long river flow series primarily as a result of French, English, and Anglo-Egyptian interests in water resources development from the early colonial period in Africa (circa 1880s) to independence (1960s onwards). Since independence, western and central African countries have tended to receive greater support for coordinated data collection [particularly through the Office de la Recherche Scientifique et Technique d’Outre-Mer (ORSTOM), now IRD]. In the Nile River basin, Egyptian and Sudanese interests have maintained extensive hydrological records, although conflict in southern Sudan has undermined these efforts since 1983. There are only a few large international river basins in East Africa that drain into the Indian Ocean because the region is dominated by the complex internal hydrology of the Rift Valley Lakes system. Extensive, reliable lake level records exist for many of these lakes and have been described in detail by Nicholson (1998, 1999). Rather surprisingly, southern Africa has few long duration records for its larger river basins, partly because the effects of human influence (dams and abstractions) very early on in the Limpopo and Orange Rivers have restricted the compilation of such records. Our coverage is, therefore, limited to the Zambezi (measured at Livingstone, upstream of Lake Kariba) and Okavango Rivers. We also analyzed the Olifants, a major tributary of the Limpopo, but decided the flow record was too difficult to naturalize. Finally, the Horn of Africa, Ethiopia, and Somalia are poorly represented because of limited data availability, especially in eastern and southern Ethiopia where some very large river basins (e.g., Omo and Wabe Shebelle) remain sparsely instrumented and understudied.

3. Methods of analysis

We characterize rainfall and runoff for different periods; World Meteorological Organization (WMO) normals (1901–30, 1931–60, and 1961–90) and periods before and after notable breakpoints were identified using statistical tests. Means, coefficients of variation (CV), and selected indicators of trend (based on linear regression) and temporal variability are presented. Breakpoints in time series are identified using Khronostat 1.0 software (Lubès-Niel et al. 1998a) for nonparametric tests and segmentation tests, including Hubert’s segmentation, Pettitt, Lee and Heghinian, and Buishand tests (Lubes-Niel et al. 1998b). The relationship between rainfall and river flow is examined using plots of 20-yr moving average correlations. Annual rainfall–runoff plots are used to identify shifts and spatial differences in relationships. The runoff coefficient represents the ratio (expressed as a percentage) of rainfall to runoff—that is, the fraction of total rainfall that becomes river flow.

4. Rainfall and river flow variability

a. Long-term conditions

Table 2 shows descriptive statistics for rainfall and river flows based on the 1961–90 WMO period along with percent differences from the previous 30-yr period from which data are available. The West African rivers are mainly strongly seasonal and humid with fairly modest interannual rainfall variability. In all cases, river flows show much greater CVs mainly because of the heterogeneity and nonlinear response of runoff to changes in rainfall, especially to variations in rainfall intensity. Groundwater interactions also contribute: variability of groundwater levels is linked to cumulative rainfall anomalies and can affect runoff over prolonged periods independently of the rainfall anomaly of a specific year (Mahé et al. 2000). Runoff coefficients are fairly low and show considerable variation, ranging from around 4% to 27%. The period is marked by the large negative trend in rainfall and river flows, which occurs in all the West African rivers and has been widely documented (Janicot 1992; Paturel et al. 1997, 1998; Mahé and Olivry 1999; Mahé et al. 2001; Leduc et al. 2001; Le Barbé and Lebel 1997). Time series show this event is characterized by a shift rather than a trend. Proportionally, the shift is much greater in river flows (from −13% to −51%) than rainfall (from −7% to −14%). West Africa shows strong intraregional homogeneity, with all rivers studied exhibiting broadly similar temporal behavior. Changes in rainfall and runoff between 1901–30 and 1931–60 (not shown) are modest, with rainfall ranges from −2% to 8% and river flow from −1% to 2% (only five rivers have data for both periods).

Table 2.

Long-term conditions for rainfall and river flows: annual mean, CV, runoff coefficient (RC), and percentage change in rainfall and river flow between 1931 and 1960, and between 1961 and 1990. Refer to Table 1 and Fig. 1 for locations.

Long-term conditions for rainfall and river flows: annual mean, CV, runoff coefficient (RC), and percentage change in rainfall and river flow between 1931 and 1960, and between 1961 and 1990. Refer to Table 1 and Fig. 1 for locations.
Long-term conditions for rainfall and river flows: annual mean, CV, runoff coefficient (RC), and percentage change in rainfall and river flow between 1931 and 1960, and between 1961 and 1990. Refer to Table 1 and Fig. 1 for locations.

Central Africa is dominated by the Congo River and its major tributary, the Bangui. Rainfall and river flows are fairly stable from year to year, and annual means show little variation between 1931–60 and 1961–90, except for the decrease of flows of the Bangui.

East and southern Africa show greater heterogeneity, both within and between regions. Interannual variability tends to be highest in the drier basins—higher than in West African basins with an equivalent annual rainfall. River flows are generally less variable than in West Africa, with the exception of the Atbara. Rainfall and river flows in all basins show decreasing trends from 1961–90. Runoff coefficients range considerably—from 3% to 63%—because of the effects of high evaporative demand and transmission losses (Atbara River), transmission losses (Sobat River and Sudd swamps), and lake evaporation (Lake Victoria). Changes between 1931–60 and 1961–90 are mixed: three basins show modest decreases in rainfall, three show almost no change, and two show modest to large increases. These trends are associated with some very large river flow responses, not easily explained by the rainfall changes and are explored in more detail in section 5. The two southern African rivers possess slightly different climatic conditions: the upper Zambezi is humid seasonal and the Okavango is closer to semiarid seasonal, and both rivers have modest interannual rainfall and river flow variability and quite low runoff coefficients. Mean rainfall between 1931–60 and 1961–90 is fairly stable as is the river flow in the Zambezi, whereas a 14% increase was recorded in the Okavango.

b. Decadal and interannual variability

A sample of rainfall and river flow records for West Africa is shown in Fig. 2. These highlight strong regional homogeneity in temporal behavior because all the series show the marked downturn in rainfall and river flow around 1970 that characterizes the climate of the Sahel during the last century. Many series also show humid conditions during the 1950s and 1960s. These features have been analyzed in detail, for example, rainfall by Lamb (1982), Nicholson (1983), and Hulme (1992); river flow by Sircoulon (1976, 1985); and both rainfall and river flow by Mahé and Olivry (1999). Unfortunately, it has not been possible to update most of these series beyond the 1990s; however, recent studies by L’Hôte et al. (2002) and Dai et al. (2004) note that rainfall in the Sahel has not returned to those conditions prior to the early 1970s. Rainfall and river flows certainly reached their lowest point in the mid-1980s and have stabilized, and in most cases recovered somewhat.

Fig. 2.

Annual rainfall (black line) and river flow (gray line) series for West Africa stations (river in parentheses): Douna (Niger), Dapola (Volta), Koulikoro (Niger), Oualia and Bakel (Senegal), and N’Djamena (tributary of Lake Chad). Note different vertical scales and record lengths for river flow (refer to Table 1) and standard record lengths for rainfall (1901–2002).

Fig. 2.

Annual rainfall (black line) and river flow (gray line) series for West Africa stations (river in parentheses): Douna (Niger), Dapola (Volta), Koulikoro (Niger), Oualia and Bakel (Senegal), and N’Djamena (tributary of Lake Chad). Note different vertical scales and record lengths for river flow (refer to Table 1) and standard record lengths for rainfall (1901–2002).

The Chari and Logone Rivers are the main tributaries of Lake Chad, and they join at N’Djamena, Chad, to form the N’Djamena River just before entering the lake. Although it is the most easterly of this group of rivers, it shows similar temporal behavior, with its decrease leading to the dramatic lowering and shrinking of Lake Chad since the 1970s (Lemoalle 2004). Under present climatic conditions, these two rivers contribute ∼90% of the lake’s water, with the remaining 10% coming from local rainfall and deliveries by the El Beid and Komadougou Yobe Rivers (Birkett 2000). Severe droughts occurred during the 1970s and 1980s, leading to widespread sinking of boreholes and to the development of a large-scale irrigation system that promoted water loss through evaporation (Birkett 2000). The more “recent” reduction in the lake’s surface area is the outcome of sustained decrease in river flow as a result of persistent rainfall failures, human-induced increases in evaporation losses, and poor management of irrigation water.

The two longest river flow records for central Africa are shown in Fig. 3. The region is dominated by the Congo and by the Bangui, which forms its main northern tributary—draining large parts of the Central African Republic, south of Lake Chad (in fact, the decrease in rainfall and flows post-1970 is similar to the West African region). The Congo’s rainfall and river flows have been quite stable, with only the 1960s (wet) and 1980s (dry) showing any decadal patterns of variability. The other river basins (series not shown) exhibit quite marked interannual variability (high CVs) and modest decadal or trend-like patterns.

Fig. 3.

Same as Fig. 2 but for central Africa: Kinshasa (Congo) and Oubangui (Bangui).

Fig. 3.

Same as Fig. 2 but for central Africa: Kinshasa (Congo) and Oubangui (Bangui).

Figure 4 shows four examples from East Africa, again chosen for display because of their long river flow records. Temporal behavior in river flows—and to a lesser extent rainfall—is regionally less homogeneous. The Blue Nile (and Atbara, but not shown) display some features similar to West Africa: humid 1950s and dry 1970s and 1980s, but it has recovered more than West African river flows have. The Sobat River is more stable and shows intermediate behavior to the Blue Nile and Atbara to the north and Lake Victoria and other rivers to the south. The marked rise in outflows from Lake Victoria after 1961 has been explained using the lake’s water balance to show it resulted from a series of extremely wet years in the 1960s and a slight increase in the short rains after the 1960s, combined with lake storage effects (Piper et al. 1986; Sene and Plinston 1994; Conway 2002). Many other East African lakes show marked increases in level in 1961 (and in other years such as 1968, 1978, 1982, and 1997), but these increases have been much shorter in duration. The smaller rivers such as Tana (Fig. 4) and Kagera show the short-lived effects of major rainfall extremes that produce high levels of interannual variability with modest decadal variability. The two rivers in southern Africa (Fig. 5) have stable rainfall series but quite high decadal variability in river flows, which is discussed below.

Fig. 4.

Same as Fig. 2 but for East Africa: Blue Nile, Sobat, Lake Victoria outflows and Tana.

Fig. 4.

Same as Fig. 2 but for East Africa: Blue Nile, Sobat, Lake Victoria outflows and Tana.

Fig. 5.

Same as Fig. 2 but for southern Africa: Zambezi and Okavango.

Fig. 5.

Same as Fig. 2 but for southern Africa: Zambezi and Okavango.

Across the whole of sub-Saharan Africa, maximum 20-yr trends expressed as percent of long-term means have ranged from ±3% per annum for rainfall and from −15% to +11% per annum for river flow (not shown).

5. Rainfall–runoff relationships

a. Regional relationships

Figure 6 shows the strength of regression relationships between annual rainfall and runoff during 1961–90 and 1931–60. Rivers in West Africa generally display very strong relationships, with rainfall accounting for around 60%–70% of river flow variability. In central Africa, relationships are slightly weaker but still quite robust (around 50% variance explained). Relationships in East and southern Africa are substantially weaker, with the exception of the Blue Nile, which drains much of central and northern Ethiopia. To help explain some of these differences, Fig. 7 shows the nature of the relationships that are typical of each region. The Niger at Koulikoro, Mali (upstream of the Niger Inland Delta), is fairly typical of West Africa. Nearly all the rivers show much weaker (and generally lower gradient) relationships during the period 1901–30, with stronger relationships (and generally higher gradient) across both 1931–60 and 1961–90. In central Africa the results are less consistent: fairly weak relationships in the Ogooué River of Gabon (two flow series have wide scatter and no obvious patterns; plots not shown) and stronger but highly unstable relationships in the Congo and Bangui. Both of these rivers only produced strong relationships during 1961–90; Fig. 7 shows almost random patterns for the time prior to that period, although a moderately positive relationship exists for the Congo. The other rivers in central Africa generally show reasonable linear relationships but data are only available for 1961–90, so it is not possible to comment on their temporal stability.

Fig. 6.

Strength of regression relationships (R2) between rainfall and runoff for the 30-yr periods 1931–60 and 1961–90 (results shown for all rivers with >20 yr flow data).

Fig. 6.

Strength of regression relationships (R2) between rainfall and runoff for the 30-yr periods 1931–60 and 1961–90 (results shown for all rivers with >20 yr flow data).

Fig. 7.

Rainfall–runoff relationships for up to three different 30-yr periods, with data permitting for West Africa, Niger; central Africa, Congo and Bangui; East Africa, Lake Victoria and Blue Nile: and southern Africa, Zambezi and Okavango.

Fig. 7.

Rainfall–runoff relationships for up to three different 30-yr periods, with data permitting for West Africa, Niger; central Africa, Congo and Bangui; East Africa, Lake Victoria and Blue Nile: and southern Africa, Zambezi and Okavango.

In East Africa, many of the rivers possess complex drainage systems as a result of the Rift Valley and other physiographical features. For example, the Sobat River and Equatorial Lakes experience, respectively, the nonlinear effects of overbank losses and lake level–area effects on inflow–outflow relationships, resulting in weak relationships with significant nonstationary behavior (see below). Outflows from Lake Victoria show a very weak relationship to basin rainfall—even during periods of stationary conditions. Lake outflows are constricted so that their response to wet years is attenuated, leading to a smoothed response. Difficulties in estimating lake rainfall have been identified in efforts to model the lake’s water balance (the lake has an area of about 78 000 km2 out of a basin area of 258 000 km2; Piper et al. 1986; Ba and Nicholson 1998). Rivers with less complex hydrology show better results such as the Blue Nile (Fig. 7) and Tana (not shown). The two rivers in southern Africa are shown in Fig. 7 to have very weak relationships as a result of their complex runoff response to rainfall. Further exploration of the basin’s physiography may shed light on this; although, data quality for the Zambezi may be a critical factor. We have been unable to trace any metadata for this river flow record and, therefore, its accuracy is open to question (see next section). In addition, the upper basin, which drains parts of Angola, is unlikely to have had many rain gauges present throughout the whole period.

b. Analysis of nonstationary behavior

It is clear from the previous section that significant shifts in rainfall, runoff, and rainfall–runoff relationships have occurred across sub-Saharan Africa. To explore this in more detail, Table 3 lists breakpoint years and whether series show evidence of breakpoints using four statistical tests (described in section 2). The shift/discontinuity in West African rainfall and river flow series around 1970 has been previously documented (Paturel et al. 1998; Mahé and Olivry 1999; Mahé et al. 2001) and is common to all the series presented here. What this analysis identifies is the possible existence of changes in rainfall–runoff relationships as highlighted by large differences between intercepts (smaller differences in slope), which are very clear in the cases of stations Douna (Fig. 8), Niamey, Dire, Bakel, and N’Djamena (not shown). Some rivers show a greater change in the slope factor such as the Niger (Koulikoro; Fig. 8), Makurdi, and Oualia. In the Sudano–Guinean area, the prolonged rainfall decline since the beginning of the 1970s led to a persistent deepening of groundwater levels. The percentage of baseflow in the annual discharge of all rivers in West Africa is, therefore, correspondingly lower because the drought exacerbates the effects on river flows (except for Sahelian rivers, where the groundwater contribution to surface runoff is insignificant; Mahé et al. 2005). This is visible via the increase of the depletion coefficient (Bricquet et al. 1997; Orange et al. 1997; Olivry et al. 1998; Mahé et al. 2000), which means that the groundwater resources have declined, rapidly draining out since the 1970s. These shifts are likely to primarily reflect nonlinear dynamics in runoff response; however, because of the prolonged duration of the change in rainfall patterns, they may also incorporate the effects of land cover change.

Table 3.

Breakpoint years identified using four statistical tests on rainfall (top) and river flow series (bottom). Here, A = test is accepted, the series are “stationary”; R = test is rejected, the series are not “stationary”; “—” = no break detected; and NA = test is not applied (series nonnormal or gap/missing data).

Breakpoint years identified using four statistical tests on rainfall (top) and river flow series (bottom). Here, A = test is accepted, the series are “stationary”; R = test is rejected, the series are not “stationary”; “—” = no break detected; and NA = test is not applied (series nonnormal or gap/missing data).
Breakpoint years identified using four statistical tests on rainfall (top) and river flow series (bottom). Here, A = test is accepted, the series are “stationary”; R = test is rejected, the series are not “stationary”; “—” = no break detected; and NA = test is not applied (series nonnormal or gap/missing data).
Fig. 8.

Rainfall–runoff relationships before and after breakpoint years identified from rainfall series for West Africa, Niger (Koulikoro and Douna); Central Africa, Congo and Bangui; East Africa: Lake Victoria, Atbara, and Blue Nile; and Southern Africa, Zambezi.

Fig. 8.

Rainfall–runoff relationships before and after breakpoint years identified from rainfall series for West Africa, Niger (Koulikoro and Douna); Central Africa, Congo and Bangui; East Africa: Lake Victoria, Atbara, and Blue Nile; and Southern Africa, Zambezi.

Figure 9 shows time series of 20-yr running correlations between rainfall and river flows for three of the West African rivers and the total number of rainfall stations within range of all the grid boxes in the basins. The temporal pattern for the Niger at Koulikoro is similar to that of the Senegal River at Bakel, showing highest correlations between the 1950s and 1980s, which is the period with the best rainfall station coverage. Correlations tend to strengthen moderately from the 1920s to the 1950s (as station density rises), and most records show a rapid (step like) decay roughly around 1980, at which point data from the 1990s would just begin to feed into the relationships while data from 1970 are removed. This behavior most likely reflects the dramatic decline in rainfall stations after 1990, with numbers in most instances dropping from more than 50 to fewer than 10 between the late 1980s and the early 1990s.

Fig. 9.

Twenty-year running correlation between rainfall and river flow (black line) and the total number of rainfall stations (gray line) within range of grid boxes in the basin for West Africa, Niger (Koulikoro and Douna); central Africa, Congo and Bangui; East Africa, Atbara and Blue Nile; and Southern Africa, Zambezi and Okavango.

Fig. 9.

Twenty-year running correlation between rainfall and river flow (black line) and the total number of rainfall stations (gray line) within range of grid boxes in the basin for West Africa, Niger (Koulikoro and Douna); central Africa, Congo and Bangui; East Africa, Atbara and Blue Nile; and Southern Africa, Zambezi and Okavango.

The time series of rainfall stations in central Africa (Fig. 8; Congo and Bangui) shows their extreme scarcity for this region during most of the last century. The Congo shows a rainfall break point in 1980 (1981 river flows) that is associated with a modest difference in the slope and a marked reduction in the strength of the rainfall–runoff relationship. The running correlation is highly unstable until 1940 because no rainfall stations contributed to the basin until 1930; from then on, only five rainfall stations contributed to this vast basin. Considering the paucity of data, the relationship with river flow is remarkably strong from the 1950s to the 1980s. Similar observations hold for the Bangui, where the absolute low numbers of rainfall stations and their change over time are likely to account for shifts in the nature of their rainfall–runoff relationships.

East Africa has had better rainfall station coverage than central Africa, although not as good as West Africa. Figures 8 and 9 show that the Blue Nile has had a remarkably stable and robust relationship over time, which is somewhat surprising given the low number of rainfall stations contributing to the rainfall series and the complex large basin (Conway 2000). The Atbara, just to the north and with a similar number of contributing stations, has a weaker and much less stable relationship and a breakpoint in river flows that occurred in 1964 when runoff decreased. This phenomenon may be related to the construction of the Khasm el Girba reservoir upstream from the river gauge in the early 1960s. The Sobat and Kagera both show very weak rainfall–runoff relationships probably because of the combined effects of low rainfall station density and substantial internal storage in wet years because of their complex wetland hydrology. The relationship between Lake Victoria outflows and rainfall (Figs. 8 and 9) clearly shows the nonstationary behavior (also present in the Equatorial Lakes and Kagera River series) associated with the dramatic rise in lake level between 1961 and 1964.

The previous section highlighted the weak relationships between rainfall and runoff in both of our southern African basins, where rainfall station coverage is poor and running correlations between rainfall and river flows are generally weak (Fig. 9). For the Zambezi, before ∼1960 the correlation is very low (with a short peak caused by a couple of years “in phase” during the 1930s) and after ∼1960 it slowly increases. A quite marked shift in the rainfall–runoff relationship exists before and after the 1945 breakpoint in river flow, with substantially higher runoff after this point being possibly related to the integrity of the river flow series (Fig. 9). The early part of the Okavango record produces reasonable correlations but from the 1950s, these decrease to random behavior. Neither series shows a clear relationship with station density, a result that is somewhat surprising and difficult to explain.

6. Conclusions

a. Hydrometeorological data in Africa

The collection, quality control, and regular update of datasets used in this analysis results from the long-term collective efforts of many individuals and funding sources. The compilation and quality assurance of the hydrological records have been time consuming and even with extensive contacts across Africa, we have not been able to update many records. Our choice of river basins and hydrological series represents most of the key extensive, high-quality records for the large international rivers in sub-Saharan Africa. These series are a valuable scientific resource and should be recognized as benchmark stations for studying environmental change. The decline in the overall number of sites, their frequency of reporting, and—in some cases—quality of measurements are major concerns for the understanding of environmental change in sub-Saharan Africa. Although this is part of a global phenomenon with hydrological data (Vörösmarty et al. 2001), the situation is particularly bad in sub-Saharan Africa. The massive decline in rainfall stations used in CRU TS 2.1 from the 1980s onward severely constrains efforts to accurately monitor climate variability and confidently model biophysical systems. This is part of Africa’s wider financial, political, and institutional challenges to undertaking climate research (Washington et al. 2006).

b. Rainfall–runoff relationships

Our results show a complex pattern of behavior that includes strong but nonstationary relationships, with most examples in West Africa; a large group from across Africa with marked variations in strength, often but not always, showing the influence of rainfall station density; weak almost random behavior (particularly in southern Africa but examples occur in all other regions); and very few examples of strong, temporally stable relationships.

For some basins, limited spatial coverage of rainfall stations leads to weak rainfall–runoff relationships. In many cases, this limits the ability to establish robust relationships very far back in time (generally prior to the 1950s). However, there are situations in which weak relationships exist throughout the period of analysis, even with reasonable station coverage. There are no obvious reasons for this, particularly because some basins produce good results with relatively few rainfall stations. We surmise that the most likely reasons are the combination of data coverage and quality, possibly exacerbated by local physical conditions beyond the scope of this basin-scale analysis, for example, the impact of geology on the regime of the right-bank tributaries of the Congo River (Laraque et al. 2001) and the Niger River (Mahé et al. 2000). It is not easy to explain why some basins show robust stable relationships, with rainfall series composed of relatively few gauges (e.g., the Blue Nile).

Overall, the best period for analysis is broadly 1961–90, with the strongest relationships occurring in West Africa, reasonable relationships in central (even though station densities are very low), highly variable relationships in East Africa because of the Rift Valley’s complex hydrology, and very weak relationships in southern Africa. These variations deserve further study (see below); however, one important implication is that for many of the basins analyzed here macroscale modeling of variability using these data will be limited in accuracy.

c. Land use and land cover change

The high variability and weak rainfall–runoff relationships means that it is difficult to identify and attribute changes in runoff to particular causes, such as climate change, or land use or land cover change (LUCC). Recent work on global precipitation variability has identified a climate change signal that includes Sahel drying (Zhang et al. 2007). However, no other clear patterns have emerged across Africa. Gedney et al. (2006) identify a direct effect of CO2 on transpiration and global runoff patterns that they postulate has contributed to a recent increase in global runoff. Our results identify marked changes in runoff ratios (especially in the Sahel after 1970) but these integrate changes in data, nonlinearities in the runoff response, and the effects of LUCC. We find that in sub-Saharan Africa, robust identification and attribution of hydrological change is, therefore, severely limited by conflicting behavior across basins/regions, low signal-to-noise ratio, sometimes weak rainfall–runoff relationships, and limited assessment of the magnitude and potential effects of LUCC or other anthropogenic influences. An important area deserving further research that we have not looked at is the role of evaporation in rainfall–runoff analyses.

d. Future climate change

The high levels of variability found in the historical records provide excellent opportunities to better understand their societal effects and adaptive responses (Glantz 1992; Adger et al. 2003). The areas with reasonable climate model convergence show runoff increases in East Africa and reductions in southern Africa but no clear signal for the Sahel and central Africa (Milly et al. 2005; Christensen et al. 2007). The main and most understood climate drivers of interannual and decadal rainfall variability are Atlantic (and other) Ocean SST patterns (West Africa and the Sahel), ENSO behavior (West, southern and East Africa), and Indian Ocean dynamics (East Africa and southern Africa). However, the underlying drivers of variability in these factors and their African teleconnections are not well captured by climate models, and model simulations of future climate do not show clear tendencies in their behavior (e.g., ENSO; Merryfield 2006; Indian Ocean; Conway et al. 2007). Improvements in physical understanding and modeling capability will hopefully improve confidence and lead time in seasonal forecasts and climate projections, although nonstationary behavior in teleconnections may affect progress toward these goals (Richard et al. 2000).

e. Overall conclusions

The conclusions are based around our aims, which for sub-Saharan Africa were to characterize the spatial and temporal dimensions of rainfall and river flow variability. The region is possibly unique in its possession of so many large, relatively undisturbed river basins in which to study long-term hydrometeorological behavior. Rainfall records from a global high-resolution product and extensive river flow records from nine major international river basins provide the basis for the analysis. The early and latter decades of last century generally show very sparse coverage of rainfall stations. In most cases, we are confident the hydrological series are reliable and possess thorough supporting information, although the Zambezi record is an exception to this.

Our findings confirm that rainfall variability in the region is high, but also that rainfall provides the dominant control along with river basin physiography and human interventions on interannual and interdecadal variability in river flows and hence surface water availability. River flows in major basins show clear examples of significant variability that challenge the effective management of water resources and result in huge socioeconomic costs. Although this work has concentrated on biophysical variability, we recognize an important need to link this understanding with the institutional and policy context of water resources management in Africa. More effective management of variability (the foundation for adaptation) is contingent upon operational capacity, which is weak in many parts of Africa.

The main findings of this analysis are presented below.

  • Station densities in CRU TS 2.1 (and other similar products) for Africa before ∼1930 and after ∼1980 are very low, and some basins have very low densities throughout their records (e.g., Congo: maximum of five gauges across 3.5 × 106 km2). Care is required in the interpretation of time series for these periods and regions.

  • Trends in rainfall and river flows have been large during the twentieth century. Rainfall (river flows) have displayed 20-yr moving trends of up to ±3% (−15%/+11%) of annual means per year. Changes in rainfall are magnified in the runoff response. This level of variability presents significant challenges to water resources management.

  • On decadal time scales, sub-Saharan Africa is characterized by drying across the Sahel after the early 1970s, relative stability punctuated by extreme wet years in East Africa (sometimes spreading into the Congo basin, e.g., 1961), and periodic behavior underlying high interannual variability in southern Africa. Central Africa shows very modest decadal variability, with some similarities to the Sahel in adjoining basins. A subregion of East Africa, the Horn, shows drying in the 1970s and 1980s similar to the Sahel but has recovered substantially during the 1990s.

  • Runoff coefficients tend to increase with increasing annual rainfall; they show a widespread decrease in West Africa after the 1970s drought but no consistent patterns elsewhere.

  • Overall, the best period for robust rainfall–runoff relationships analysis is broadly 1961–90. The strongest relationships occur in West Africa, reasonable relationships in central (even though station densities are very low), highly variable relationships in East Africa because of the Rift Valley’s complex hydrology, and very weak relationships in southern Africa. During the period 1961–90 (1931–60), rainfall explains 60%–80% (40%–60%) of the variability in river flows in West Africa. Equivalent approximations for other regions are 40%–70% (insufficient data) in central Africa, 5%–65% (5%–65%) in East Africa, and only 5%–20% (5%–20%) in southern Africa.

  • For some basins, limited spatial coverage of rainfall stations leads to weak rainfall–runoff relationships. However, there are examples in which weak relationships exist throughout the period of analysis, even with reasonable station coverage and examples of robust stable relationships with rainfall series comprising relatively few gauges. In basins with weak relationships, macroscale modeling using these data will be of limited success without considering data and subbasin scale conditions.

  • Our results identify marked changes in runoff ratios and nonstationary rainfall–runoff relationships (especially in the Sahel after 1970), which integrate changes in data, nonlinearities in the runoff response, and the effects of LUCC. We conclude that for sub-Saharan Africa, robust identification and attribution of hydrological change is severely limited by data limitations, conflicting behavior across basins/regions, low signal-to-noise ratios, sometimes weak rainfall–runoff relationships, and limited assessment of the magnitude and potential effects of LUCC or other anthropogenic influences.

Acknowledgments

The authors wish to acknowledge John Sutcliffe, Dick Grove, and Jacques Sircoulon for their pioneering work on this topic. We also thank the many African hydrological agencies for the information and data that supported this analysis. The United Kingdom–France collaboration was facilitated by a British Council Alliance/EGIDE award (2004–06). Helpful comments from three reviewers improved this manuscript.

REFERENCES

REFERENCES
Adger
,
W. N.
,
S.
Huq
,
K.
Brown
,
D.
Conway
, and
M.
Hulme
,
2003
:
Adaptation to climate change in the developing world.
Prog. Dev. Stud.
,
3
,
179
195
.
Ardoin-Bardin
,
S.
,
2004
:
Hydroclimate variability and impacts on water resources of large hydrological catchments in the Sudanese–Sahelian area (in French). Ph.D. thesis, University of Montpellier II, 437 pp
.
Ba
,
M. B.
, and
S. E.
Nicholson
,
1998
:
Analysis of convective activity and its relationship to the rainfall over the Rift Valley lakes of East Africa during 1983–90 using the Meteosat infrared channel.
J. Appl. Meteor.
,
37
,
1250
1264
.
Bader
,
J. C.
,
1990
:
Homogenization and completion of hydrometric data in Senegal catchment upstream of Bakel (in French).
ORSTOM Rep., 18 pp
.
Bader
,
J. C.
,
1992
:
Study of Manantali’s dam impacts on hydrological regime of the Senegal river at Bakel (in French).
ORSTOM Rep., 56 pp
.
Birkett
,
C. M.
,
2000
:
Synergistic remote sensing of Lake Chad: Variability of basin inundation.
Remote Sens. Environ.
,
72
,
218
236
.
Boyer
,
J. F.
,
C.
Dieulin
,
N.
Rouche
,
A.
Cres
,
E.
Servat
,
J. E.
Paturel
, and
G.
Mahé
,
2006
:
SIEREM: An environmental information system for water resources.
Climate Variability and Change—Hydrological Impacts, S. Demuth et al., Eds., IAHS Publication 308, IAHS/AISH, 19–25
.
Bricquet
,
J. P.
,
F.
Bamba
,
G.
Mahé
,
M.
Toure
, and
J. C.
Olivry
,
1997
:
Water resource variations of the Atlantic river basins of Africa: The long-term effects of rain shortage (in French).
Revue Sci. Eau
,
3
,
321
337
.
Conway
,
D.
,
2000
:
The climate and hydrology of the Upper Blue Nile River.
Geogr. J.
,
166
,
49
62
.
Conway
,
D.
,
2002
:
Extreme rainfall events and lake level changes in East Africa: Recent events and historical precedents.
The East African Great Lakes: Limnology, Palaeolimnology and Biodiversity, E. O. Odada and D. O. Olago, Eds., Advances in Global Change Research Series, Vol. 12, Kluwer, 63–92
.
Conway
,
D.
,
2005
:
From headwater tributaries to international river: Observing and adapting to climate variability and change in the Nile basin.
Global Environ. Change
,
15A
,
99
114
.
Conway
,
D.
, and
M.
Hulme
,
1993
:
Recent fluctuations in precipitation and runoff over the Nile subbasins and their impact on main Nile discharge.
Climatic Change
,
25
,
127
151
.
Conway
,
D.
,
C. E.
Hanson
,
R.
Doherty
, and
A.
Persechino
,
2007
:
GCM simulations of the Indian Ocean dipole influence on East African rainfall: Present and future.
Geophys. Res. Lett.
,
34
,
L03705
.
doi:10.1029/2006GL027597
.
Christensen
,
J. H.
, and
Coauthors
,
2007
:
Regional climate projections.
Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 849–940
.
Christie
,
F.
, and
J.
Hanlon
,
2001
:
Mozambique and the Great Flood of 2000. International African Institute and James Currey, 176 pp
.
Dai
,
A. D.
,
P. J.
Lamb
,
K. E.
Trenberth
,
M.
Hulme
,
P. D.
Jones
, and
P.
Xie
,
2004
:
The recent Sahel drought is real.
Int. J. Climatol.
,
24
,
1323
1331
.
Gedney
,
N.
,
P. M.
Cox
,
R. A.
Betts
,
O.
Boucher
,
C.
Huntingford
, and
P.
Stott
,
2006
:
Detection of a direct carbon dioxide effect in continental river runoff records.
Nature
,
439
,
835
838
.
doi:10.1038/nature04504
.
Glantz
,
M. H.
,
1992
:
Global warming and environmental change in sub-Saharan Africa.
Global Environ. Change
,
2
,
183
204
.
Goulden
,
M.
,
2006
:
Livelihood diversification, social capital and resilience to climate variability amongst natural resource dependent societies in Uganda. Ph.D. thesis, School of Environmental Sciences, University of East Anglia, 421 pp
.
Grove
,
A. T.
,
1996
:
African river discharges and lake levels in the twentieth century.
The Limnology, Climatology and Paleoclimatology of the East African Lakes, T. C. Johnson and E. O. Odada, Eds., Gordon and Breach, 95–100
.
Hamandawana
,
H.
,
2007
:
The desiccation of southern Africa’s Okavango Delta: Periodic fluctuation or long-term trend?
Past Global Changes
,
15
,
12
13
.
Hamandawana
,
H.
,
F.
Eckardt
, and
R.
Chanda
,
2005
:
Linking archival and remotely sensed data for long-term environmental monitoring.
Int. J. Appl. Earth Obser. Geoinf.
,
7
,
284
298
.
Hamandawana
,
H.
,
R.
Chanda
, and
F.
Eckardt
,
2007
:
The role of human factors in the degradation of natural resources in and around the Okavango Delta.
Int. J. Environ. Stud.
,
64
,
589
605
.
Hulme
,
M.
,
1992
:
Rainfall changes in Africa: 1931–1960 to 1961–1990.
Int. J. Climatol.
,
12
,
685
699
.
Hulme
,
M.
,
1996
:
Climate change within the period of meteorological records.
The Physical Geography of Africa, W. M. Adams, A. S. Goudie, and A. R. Orme, Eds., Oxford University Press, 88–102
.
Hulme
,
M.
,
R.
Doherty
,
T.
Ngara
,
M.
New
, and
D.
Lister
,
2001
:
African climate change: 1900–2100.
Climate Res.
,
17
,
145
168
.
Hurst
,
H. E.
,
1933
:
The Nile Basin.
Vol. IV and Supplements 1–13, Ten-Day Mean and Monthly Mean Discharges of the Nile and its Tributaries, Government Press, 291 pp
.
Janicot
,
S.
,
1992
:
Spatiotemporal variability of West African rainfall. Part I: Regionalization and typings.
J. Climate
,
5
,
489
497
.
Lamb
,
P.
,
1982
:
Persistance of subsaharan drought.
Nature
,
299
,
46
47
.
Lankford
,
B. A.
, and
T.
Beale
,
2007
:
Equilibrium and non-equilibrium theories of sustainable water resources management: Dynamic river basin and irrigation behaviour in Tanzania.
Global Environ. Change
,
17
,
168
180
.
Laraque
,
A.
,
G.
Mahé
,
D.
Orange
, and
B.
Marieu
,
2001
:
Spatiotemporal variations in hydrological regimes within Central Africa during the XXth century.
J. Hydrol.
,
245
,
104
117
.
Le Barbé
,
L.
, and
T.
Lebel
,
1997
:
Rainfall climatology of the HAPEX-Sahel region during the years 1950–1990.
J. Hydrol.
,
188–189
,
43
73
.
Leblanc
,
M.
,
G.
Favreau
,
S.
Massuel
,
S.
Tweed
,
M.
Loireau
, and
B.
Cappelaere
,
2008
:
Land clearance and hydrological change in the Sahel: SW Niger.
Global Planet. Change
,
61
,
135
150
.
Leduc
,
C.
,
G.
Favreau
, and
P.
Schroeter
,
2001
:
Long-term rise in a Sahelian water-table: The continental terminal in south-west Niger.
J. Hydrol.
,
243
,
43
54
.
Lemoalle
,
J.
,
2004
:
Lake Chad: A changing environment.
Dying and Dead Seas: Climatic Versus Anthropic Causes, J. C. J. Nihoul, P. O. Zavialov, and P. P. Micklin, Eds., NATO Science Series, Vol. 36, Kluwer, 321–340
.
L’Hôte
,
Y.
,
G.
Mahé
,
B.
Some
, and
J. P.
Triboulet
,
2002
:
Analysis of a Sahelian annual rainfall index updated from 1896 to 2000: The drought continues.
Hydrol. Sci. J.
,
47
,
563
572
.
Li
,
K. Y.
,
M. T.
Coe
, and
N.
Ramankutty
,
2005
:
An investigation of hydrological variability in West Africa using land surface models.
J. Climate
,
18
,
3173
3188
.
Li
,
K. Y.
,
M. T.
Coe
,
N.
Ramankutty
, and
R.
De Jong
,
2007
:
Modeling the hydrological impact of land-use change in West Africa.
J. Hydrol.
,
337
,
258
268
.
Lørup
,
J. K.
,
J. C.
Refsgaard
, and
D.
Mazvimavi
,
1998
:
Assessing the effect of land use change on catchment runoff by combined use of statistical tests and hydrological modelling: Case studies from Zimbabwe.
J. Hydrol.
,
205
,
147
163
.
Lubès-Niel
,
H.
,
J. E.
Paturel
,
J. F.
Boyer
, and
E.
Servat
,
1998a
:
Khronostat: Statistical analyses software of chronological series (in French), version 1.0.
IRD-ORSTOM
.
Lubès-Niel
,
H.
,
J. M.
Masson
,
J. E.
Paturel
, and
E.
Servat
,
1998b
:
Variablite climatique et statistique. Etude par simulation de la puissance et de la robustesse de quelques tests utilises pour verifier l’homogeneïte de chroniques.
Revue Sci. Eau
,
11
,
383
408
.
Magistro
,
J.
, and
M.
Lo
,
2001
:
Historical and human dimensions of climate variability and water resource constraint in the Senegal River Valley.
Climate Res.
,
19
,
133
147
.
Mahé
,
G.
, and
J. C.
Olivry
,
1999
:
Assessment of freshwater yields to the ocean along the intertropical Atlantic coast of Africa.
C. R. Acad. Sci.
,
328
,
621
626
.
Mahé
,
G.
,
J. C.
Olivry
,
R.
Dessouassi
,
D.
Orange
,
F.
Bamba
, and
E.
Servat
,
2000
:
Surface water and groundwater relationships in a tropical river of Mali (in French).
C. R. Acad. Sci.
,
330
,
689
692
.
Mahé
,
G.
,
Y.
L’Hôte
,
J. C.
Olivry
, and
G.
Wotling
,
2001
:
Trends and discontinuities in regional rainfall of West and Central Africa: 1951–1989.
Hydrol. Sci. J.
,
46
,
211
226
.
Mahé
,
G.
,
J. E.
Paturel
,
E.
Servat
,
D.
Conway
, and
A.
Dezetter
,
2005
:
Impact of land use change on soil water holding capacity and river modelling of the Nakambe River in Burkina-Faso.
J. Hydrol.
,
300
,
33
43
.
McMahon
,
T. A.
,
M. C.
Peel
,
R. M.
Vogel
, and
G. G. S.
Pegram
,
2007
:
Global streamflows—Part 3: Country and climate zone characteristics.
J. Hydrol.
,
347
,
272
291
.
Merryfield
,
W.
,
2006
:
Changes to ENSO under CO2 doubling in a multimodel ensemble.
J. Climate
,
19
,
4009
4027
.
Milly
,
P. C. D.
,
K. A.
Dune
, and
A. V.
Vecchia
,
2005
:
Global pattern of trends in streamflow and water availability in a changing climate.
Nature
,
438
,
347
350
.
doi:10.1038/nature04312
.
Mitchell
,
T. D.
, and
P. D.
Jones
,
2005
:
An improved method of constructing a database of monthly climate observations and associated high-resolution grids.
Int. J. Climatol.
,
25
,
693
712
.
Moniod
,
F.
,
B.
Pouyuad
, and
P.
Sechet
,
Eds.
1977
:
The Catchment of the Volta River (in French). Hydrological Monogr., No. 5, ORSTOM, 513 pp
.
Nemec
,
J.
, and
J.
Schaake
,
1982
:
Sensitivity of water resource systems to climate variation.
Hydrol. Sci. J.
,
27
,
327
343
.
New
,
M.
,
M. C.
Todd
,
M.
Hulme
, and
P.
Jones
,
2001
:
Precipitation measurements and trends in the twentieth century.
Int. J. Climatol.
,
21
,
1889
1922
.
Nicholson
,
S. E.
,
1983
:
Sub-Saharan rainfall in the years 1976–80: Evidence of continued drought.
Mon. Wea. Rev.
,
111
,
1646
1654
.
Nicholson
,
S. E.
,
1996
:
Environmental change within the historical period.
The Physical Geography of Africa, W. M. Adams, A. S. Goudie, and A. R. Orme, Eds., Oxford University Press, 60–87
.
Nicholson
,
S. E.
,
1998
:
Historical fluctuations of Lake Victoria and other lakes in the northern Rift Valley of East Africa.
Environmental Change and Response in East African Lakes, J. T. Lehman, Ed., Kluwer, 7–35
.
Nicholson
,
S. E.
,
1999
:
Historical and modern fluctuations of lakes Tanganyika and Rukwa and their relationship to rainfall variability.
Climatic Change
,
41
,
53
71
.
Nicholson
,
S. E.
,
2005
:
On the question of the “recovery” of the rains in the West African Sahel.
J. Arid Environ.
,
63
,
615
641
.
Ogutunde
,
P. G.
,
J.
Friesen
,
N.
van de Giesen
, and
H. H. G.
Savenije
,
2006
:
Hydroclimatology of the Volta River Basin in West Africa: Trends and variability from 1901 to 2002.
Phys. Chem. Earth
,
31
,
1180
1188
.
Olivry
,
J. C.
,
J. P.
Bricquet
, and
G.
Mahé
,
1998
:
Water Resources Variability in Africa during the XXth Century. IASH Publication 252, IAHS/AISH, 189–197
.
Orange
,
D.
,
A.
Wesselink
,
G.
Mahé
, and
C.
Feizoure
,
1997
:
The effects of climate changes on river baseflow and aquifer storage in Central Africa.
Sustainability of Water Resources under Increasing Uncertainty, E. Servat et al., Eds., IAHS Publication 240, IAHS, 113–123
.
Paturel
,
J. E.
,
E.
Servat
,
B.
Kouame
,
H.
Lubes
,
M.
Ouedraogo
, and
J. M.
Masson
,
1997
:
Climatic variability in humid Africa along the Gulf of Guinea. Part II: An integrated regional approach.
J. Hydrol.
,
191
,
16
36
.
Paturel
,
J. E.
,
E.
Servat
,
M. O.
Delattre
, and
H.
Lubes-Niel
,
1998
:
Analysis of rainfall long series in non-Sahelian West and Central Africa within the context of climate variability (in French).
Hydrol. Sci. J.
,
43
,
937
946
.
Peel
,
M. C.
,
T. A.
McMahon
,
B. L.
Finlayson
, and
F. G. R.
Watson
,
2001
:
Identification and explanation of continental differences in the variability of annual runoff.
J. Hydrol.
,
250
,
224
240
.
Peel
,
M. C.
,
T. A.
McMahon
, and
B. L.
Finlayson
,
2004
:
Continental differences in the variability of annual runoff—Update and reassessment.
J. Hydrol.
,
295
,
185
197
.
Piper
,
B. S.
,
D. T.
Plinston
, and
J. V.
Sutcliffe
,
1986
:
The water balance of Lake Victoria.
Hydrol. Sci. J.
,
31
,
25
37
.
Richard
,
Y.
,
S.
Trzaska
,
P.
Roucou
, and
M.
Rouault
,
2000
:
Modification of the southern African rainfall variability/ENSO relationship since the late 1960s.
Climate Dyn.
,
16
,
883
895
.
Saji
,
N. H.
,
B. N.
Boswami
,
P. N.
Vinayachandran
, and
T.
Yamagata
,
1999
:
A dipole made in the tropical Indian Ocean.
Nature
,
401
,
360
363
.
Sarch
,
M-T.
, and
E. H.
Allison
,
2000
:
Fluctuating fisheries in Africa’s inland waters: Well-adapted livelihoods, maladapted management.
Proc. 10th Int. Conf. of the Institute of Fisheries Economics and Trade, Corvallis, OR, U.S. National Marine Fisheries Service. [Available online at http://osu.orst.edu/dept/IIFET/2000/papers/sarch.pdf.]
.
Savenije
,
H. H. G.
,
1996
:
Does moisture feedback affect rainfall significantly?
Phys. Chem. Earth
,
20
,
507
513
.
Sene
,
K. J.
, and
D. T.
Plinston
,
1994
:
A review and update of the hydrology of Lake Victoria.
Hydrol. Sci. J.
,
39
,
47
63
.
Sircoulon
,
J.
,
1976
:
Hydropluviometric data for the recent drought in intertropical Africa. Comparison with the droughts of 1914 and 1940 (in French).
Cah. ORSTOM, Ser. Hydrol.
,
13
,
75
174
.
Sircoulon
,
J.
,
1985
:
The drought in West Africa. Comparison between the years 1982–1984 and the years 1972–1973 (in French).
Cah. ORSTOM, Ser. Hydrol.
,
21
,
75
86
.
Sutcliffe
,
J. V.
, and
D. G.
Knott
,
1987
:
Historical variations in African water resources.
The Influence of Climate Change and Climatic Variability on the Hydrologic Regime and Water Resources, S. I. Solomon, M. Beran and W. Hogg, Eds., IAHS Publication 168, IAHS/AISH, 463–476
.
Sutcliffe
,
J. V.
, and
Y. P.
Parks
,
1999
:
The Hydrology of the Nile. IAHS Special Publication, Vol. 5, IAHS, 179 pp
.
Sutcliffe
,
J. V.
, and
G.
Petersen
,
2007
:
Lake Victoria: Derivation of a corrected natural water level series.
Hydrol. Sci. J.
,
52
,
1316
1321
.
Tarhule
,
A.
,
2005
:
Damaging rainfall and flooding: The other Sahel hazards.
Climatic Change
,
72
,
355
377
.
Tate
,
E.
,
J.
Sutcliffe
,
D.
Conway
, and
F.
Farquharson
,
2004
:
Water balance of Lake Victoria: Update to 2000 and climate change modelling to 2100.
Hydrol. Sci. J.
,
49
,
563
574
.
Troy
,
B.
,
C.
Sarron
,
J. M.
Fritch
, and
D.
Rollin
,
2007
:
Assessment of the impacts of land use changes on the hydrological regime of a small rural catchment in South Africa.
Phys. Chem. Earth
,
32
,
984
994
.
UNESCO
,
1995
:
Discharge of Selected Rivers of Africa (in French). Studies and Reports in Hydrology, Vol. 52, UNESCO, 166 pp
.
Vörösmarty
,
C. J.
, and
D.
Sahagian
,
2000
:
Anthropogenic disturbance of the terrestrial water cycle.
Bioscience
,
50
,
753
765
.
Vörösmarty
,
C. J.
, and
Coauthors
,
2001
:
Global water data: A newly endangered species.
Eos, Trans. Amer. Geophys. Union
,
82
,
54
.
Vörösmarty
,
C. J.
,
E. M.
Douglas
,
P. A.
Green
, and
C.
Revenga
,
2005
:
Geospatial indicators of emerging water stress: An application to Africa.
Ambio
,
34
,
230
236
.
Washington
,
R.
, and
Coauthors
,
2006
:
African climate change: Taking the shorter route.
Bull. Amer. Meteor. Soc.
,
87
,
1355
1366
.
Webster
,
P. J.
,
A. M.
Moore
,
J. P.
Loschnigg
, and
R. R.
Lebden
,
1999
:
Coupled ocean–atmosphere dynamics in the Indian Ocean during 1997–98.
Nature
,
401
,
356
360
.
Woyessa
,
Y. E.
,
E.
Pretorius
,
P. S.
van Heerden
,
M.
Hensley
, and
L. D.
van Rensburg
,
2006
:
Impact of land use on river basin water balance: A case study of the Modder River Basin, South Africa.
Comprehensive Assessment of Water Management in Agriculture Research Rep. 12, 37 pp. [Available online at http://www.iwmi.cgiar.org/Assessment/files_new/publications/CA%20Research%20Reports/CARR12-low.pdf.]
.
Zhang
,
X.
,
F. W.
Zwiers
,
G. C.
Hegerl
,
F. H.
Lambert
,
N. P.
Gillett
,
S.
Solomon
,
P. A.
Stott
, and
T.
Nozawa
,
2007
:
Detection of human influence on twentieth-century precipitation trends.
Nature
,
448
,
461
465
.
doi:10.1038/nature06025
.

Footnotes

Corresponding author address: Declan Conway, School of Development Studies, University of East Anglia, Norwich NR4 7TJ, United Kingdom. Email: d.conway@uea.ac.uk