Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, CLIMATE SCIENCE ( (c) Oxford University Press USA, 2020. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 20 January 2020

Changes in Precipitation Over West Africa During Recent Centuries

Summary and Keywords

Water, not temperature, governs life in West Africa, and the region is both temporally and spatially greatly affected by rainfall variability. Recent rainfall anomalies, for example, have greatly reduced crop productivity in the Sahel area. Rainfall indices from recent centuries show that multidecadal droughts reoccur and, furthermore, that interannual rainfall variations are high in West Africa. Current knowledge of historical rainfall patterns is, however, fairly limited. A detailed rainfall chronology of West Africa is currently only available from the beginning of the 19th century. For the 18th century and earlier, the records are still sporadic, and an interannual rainfall chronology has so far only been obtained for parts of the Guinea Coast. Thus, there is a need to extend the rainfall record to fully understand past precipitation changes in West Africa.

The main challenge when investigating historical rainfall variability in West Africa is the scarcity of detailed and continuous data. Readily available meteorological data barely covers the last century, whereas in Europe and the United States for example, the data sometimes extend back two or more centuries. Data availability strongly correlates with the historical development of West Africa. The strong oral traditions that prevailed in the pre-literate societies meant that only some of the region’s history was recorded in writing before the arrival of the Europeans in the 16th century. From the 19th century onwards, there are, therefore, three types of documents available, and they are closely linked to the colonization of West Africa. These are: official records started by the colonial governments continuing to modern day; regular reporting stations started by the colonial powers; and finally, temporary nongovernmental observations of various kinds. For earlier periods, the researcher depends on noninstrumental observations found in letters, reports, or travel journals made by European slave traders, adventurers, and explorers. Spatially, these documents are confined to the coastal areas, as Europeans seldom ventured inland before the mid-1800s. Thus, the inland regions are generally poorly represented. Arabic chronicles from the Sahel provide the only source of information, but as historical documents, they include several spatiotemporal uncertainties. Climate researchers often complement historical data with proxy-data from nature’s own archives. However, the West African environment is restrictive. Reliable proxy-data, such as tree-rings, cannot be exploited effectively. Tropical trees have different growth patterns than trees in temperate regions and do not generate growth rings in the same manner. Sediment cores from Lake Bosumtwi in Ghana have provided, so far, the best centennial overview when it comes to understanding precipitation patterns during recent centuries. These reveal that there have been considerable changes in historical rainfall patterns—West Africa may have been even drier than it is today.

Keywords: West Africa, rainfall variability, precipitation reconstruction, historical climatology, documentary sources


Anthropogenic climate change has increased the need for a better understanding of historical climates, causing an almost exponential growth in climate-based research in recent decades. However, in comparison with Northern Hemisphere midlatitudes, such as Europe and North America, historical rainfall variability in West Africa is poorly understood (Nash & Adamson, 2014). Whereas investigations of climatic variability in Europe provide a detailed picture for the last 1,000 years (See R. Przybylak’s Poland’s Climate in the Last Millennium), detailed knowledge on West African rainfall variability covers only 250 years. Extending the rainfall chronology is vital. Besides improving current understanding of historical rainfall variability, it will also facilitate better future predictions. When it comes to the climate, understanding the past is the key for a better understanding of the future.

Due to worldwide media coverage of the many droughts that have affected the Sahel region—a semi-arid area stretching across the African continent from west to east—since the end of the 1960s, many easily associate rainfall variability in West Africa with droughts and famines. It is often forgotten that West Africa comprises not only some of the driest areas of the world, but also some of the wettest. In the Sahel, annual rainfall varies between 50 and 100 mm, whereas in the wetter southwestern parts of West Africa, annual rainfall varies between 1,600 and 2,000 mm (Nicholson, 2011). Moreover, interannual rainfall variability is high, which is why a better understanding of past rainfall variability is of great importance for the future development of West Africa.

The aim of this article is to provide the reader with a state of the art and comprehensive overview of current knowledge of historical rainfall variability during recent centuries in West Africa. A secondary aim is to provide incentive for further research. Since research started in the 1970s, it has evolved in two separate stages and at a relatively slow pace in comparison with midlatitude climate research. The first stage, which lasted until the turn of the millennium, had a particular aim of identifying wet and dry periods and historical droughts, instead of establishing interannual rainfall variability. In the second stage, methodological development and a wider use of documentary evidence facilitated a more detailed reconstruction of historical precipitation patterns. Because of this, the article has a chronological structure. Following a presentation of data availability is an overview of the first stage of research, which includes the discussions and research questions that eventually led to the creation of the first long-term climate chronologies. Thereafter follows a presentation of the second stage, including an in-depth discussion and presentation of the historical records that have been employed. The rainfall chronology for each region of West Africa is included in each respective section. Figure 1 details a map of West Africa depicting the key areas discussed and referred to throughout the paper.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 1. Map of West Africa depicting key areas, places, and regions discussed in the article.

The Rainy Season

Water availability and rainfall variability determine life in West Africa. Water is of great significance not only for the growth of natural grasslands, but also for agricultural activity and livestock keepers, especially in the semi-arid region of the Sahel (Adams & Mortimore, 2005; Mortimore, 1989). Temperature fluctuations, which are larger on a diurnal than an annual range, do not have the same impact on life as rainfall variability. Seasons in West Africa are not defined on the basis of temperature. Instead, the most perspicuous seasons are the rainy season, when the majority of the annual precipitation occurs, and the dry season, which is predominantly dry with little expectancy of rain (e.g., Buckle, 2004; Nicholson, 2011).

The general understanding of the tropical rainfall regime and climate variability in West Africa has changed considerably the last 50 years (REF to Nicholson’s ORE article “Evolving Paradigms of Climatic Processes and Dynamics Affecting Africa”). The classic explanation of the West African rainfall regime depicts the Intertropical Convergence Zone (ITCZ) as the rain bearing mechanism. However, recent research views the rainfall and its seasonality in the context of the West African monsoon (e.g., Thorncroft, Nguyen Zhang, & Peyrillé, 2011). The ITCZ is a component of the monsoon, but rainfall is actually concentrated in a region (in some sources termed the “tropical rainbelt”) that lies between two upper-jet streams (Nicholson, 2008; Nicholson, 2009). Even though the structure and function of the rainfall regime is not in focus here, it is necessary to understand the seasonality that it creates, since seasonality is critical in understanding how historical records can be effectively exploited for gaining more knowledge of past variability.

The timing and arrival of the rains, as well as the duration and intensity, are vital for the natural environment and human societies of West Africa, which makes the rainy season a sought after event each year. The rains end a long dry period, and the success of the rains determines the crop cycle and harvest output for the following year. In the agricultural cycle, the rains signal the time for planting, and they supply plants with water for the ensuing dry season (van der Geest & Dietz, 2004).

In West Africa, the primary rainy season occurs in the boreal summer months; but whereas the coastal region experiences two rainy seasons, the north is dominated by a single rainy season. On the Guinea Coast (south of 10°N), the primary rainy season is commonly in its strongest phase in June or July, only to subside in August, when the area experiences a short dry season. This is when the rains migrate northwards to the Sahel area (north of 10°N), which usually experiences maximum rainfall in August. The coastal areas then receive a weaker secondary rainy season in the autumn months, when the rains retract southwards.

The onset of the monsoonal rains, and their intensity and duration, show high interannual and interdecadal variations. For example, the results of a study that investigated the onset of springtime rainfall on the Guinea Coast, between 1979 and 2009, showed that the onset of the rainy season varied from April 6 to June 10, and that the length of the rainy season varied between 3 and 14 pentads (Nguyen, Thorncroft, & Zhang, 2011). Rainfall is also spatially highly erratic. In 1998, for example, two stations, only 80 kilometers apart, but in a similar environment, recorded 450 mm and 1,050 mm of rainfall respectively (Le Barbé, Lebel, & Tapsoba, 2002). Consequently, it is the regularity of the seasonal rains that facilitates the use of historical documents to gain further knowledge of past rainfall variability; however, paradoxically, it is also the irregularity of the rainfall regime that creates distinct challenges in regard to proxy data and the analysis of historical records.

Data Availability

Investigating historical precipitation patterns in West Africa differs considerably from investigating historical precipitation patterns in temperate regions. Besides different climatic parameters, there is a considerable contrast in the availability of reliable and usable records. Whereas the introduction of national meteorological networks provides Europe with meteorological data for at least the last 160 years (Brázdil, Pfister, Wanner, von Storch, & Luterbache, 2005), meteorological data in West Africa barely cover the last 100 years (Nicholson, Klotter, & Dezfuli, 2012a).

Reliable natural proxy-data (e.g., tree-rings, speleothems, and pollen) are as scarce as direct meteorological observations (Vershuren, 2004). The most valuable natural proxy-data that have been obtained are from Lake Bosumtwi on the Guinea Coast (e.g., Talbot & Delibrias, 1977; Shanahan et al., 2009) and Lake Chad in the Sahel (e.g., Maley, 1973; Maley, 1993), which were investigated for the first time in the 1970s. Later investigations of the same lakes have refined the results from the 1970s and provided an even more detailed picture of wet and dry eras. For example, by measuring oxygen isotopes in laminated sediments in Lake Bosumtwi, Shanahan et al. (2009) indicated that West Africa, in the 18th century, was even drier than it was in the 20th century and, furthermore, the drought that has persisted in the Sahel since the end of the 1960s is not anomalous. However, even though data from Lake Bosumtwi provide near annual resolution of rainfall variability, they have not been extrapolated or cross-referenced to data obtained from historical records.

Tree-rings have been used extensively when investigating past climates and can provide good interannual information in temperate regions. In Europe the longest tree-ring chronology covers several thousand years (Pilcher, Baillie, Schmidt, & Becker, 1984), but in West Africa, it only extends back to the mid-1800s, and then only for parts of the Guinea Coast region (Schöngart, Orthmann, Hennenberg, Porembski, & Worbes, 2006). One of the many challenges is that the growth dynamics of tropical trees are poorly understood (Worbes, 2002). Different types of trees grow in different ecological zones, and some are more drought resistant than others. Some generate clearly identifiable growth rings, while other do not. On the Guinea Coast, studies have shown that some tree species show a better climate-growth relationship than others (Schöngart et al., 2006). In the Western Sahel, on the other hand, the relationship between tree growth and climate is unidentified (Nicolini, Tarchiani, Saurer, & Cherubini, 2010). Many tree species in West Africa are still under-investigated, and additional studies are needed before dendrochronological studies could be employed effectively to determine interannual rainfall variability. Hence, the general paucity of reliable natural proxy-data presently renders the use of historical records as the best way to study historical rainfall variability.

Historical Records: A General Overview

The “historical” period of West Africa can be said to have started with the arrival of Europeans in the 15th century (Fage, 1965). Before this, due to the prevailing oral traditions, only a fraction of the region’s history was recorded in writing (Parker & Rathbone, 2007). Until the mid-19th century, most of the available records originate from European settlements on the coast, with climatic information from the interior parts of West Africa mostly found in Arabic chronicles. Compared to other types of historical records, chronicles are compiled in retrospect, often comprising second-hand knowledge derived from the collective memory and passed from one generation to the next by word of mouth before being recorded many years later. Chronicles, therefore, lack detailed information on weather, but they do contain descriptions of large-scale societal impacts from droughts, floods, and/or famines. Because they were compiled in retrospect, there is, occasionally, an uncertainty between cause and effect. For instance, the author of Tedzkiret en-Nisian mentions starvation, but not the cause of it. It is unknown, therefore, whether or not the crisis was caused by rainfall variability or other factors, such as war (Weiss, 2008).

Historical accounts containing descriptions of climate and weather in West Africa are closely linked to the political history of the region. As such, the records can be divided into three categories. The first comprises documents from the slave-trading era or the pre-colonial period, which covers the period from the 16th to roughly the mid-19th century. Documents belonging to this category are mainly confined to the coastal areas of West Africa, where European slave traders built their settlements. Weather and climate descriptions from this era mostly consist of eyewitness descriptions of meteorological extremes and information related to the onset, duration, and intensity of the summer rains (Norrgård, 2015). They might contain, on rare occasions, meteorological data. Thomas Winterbottom (1803), for example, made diligent measurements of rainfall and temperature while stationed in Sierra Leone in the 1790s.

The second category comprises governmental documents from the colonial period, roughly covering the period from the early to mid-19th to the mid-20th century. Spatially, these records cover a much larger area than the previous category, as the Europeans established permanent settlements across West Africa. Precipitation records from this era comprise three types. First are official records from stations that occasionally continue to the modern day. Measurements from the 19th century are, in this case, more sporadic in West Africa than in North and southern Africa. Second, regular reports came from stations created by the colonial governments, many of which were abandoned when nations gained their independence in the mid-1900s. The final type of records from the colonial period consists of observations made by explorers, missionaries, and travelers who undertook inland expeditions for discovering the unknown (Nicholson et al., 2012a).

The third category consists of meteorological data obtained from meteorological stations governed by the independent states in West Africa. Unfortunately, many national meteorological agencies struggle with funding, while in the Sahel, political unrest also affects data availability and reliability.

Stage 1

Identifying Historical Drought Periods

The need to gain more and detailed knowledge on historical rainfall variability and climatic change in West Africa was prompted by the need to understand the drought crisis and famines that started in the Sahel in 1968 (McCann, 1999). After several decades of “good rains” had favored the region both environmentally and demographically, rainfall levels gradually weakened causing a famine and drought crisis of unseen proportions. Rainfall was between 20 to 40% lower in comparison to the 30-year average (Nicholson, Some, & Kone, 2000), affecting millions of people. Crop production fell drastically, and millions of cattle and other domestic animals perished (FAO, 1973). Solving the situation was of essence, and understanding the causes of famine and drought was seen as a large part of the solution.

Initially, both historians and climatologists/meteorologists took an active part in resolving the situation, trying to determine the causes and long-term effects. Even though information was used across disciplines, researchers often conducted investigations based on the perspectives of their respective specialism. Whereas historians, to some extent, stressed that climatic variability had always been a part of living in the Sahel, and that local societies had adapted their way of living to climatic variability, climatologists were more interested in explaining the drought as a consequence of either atmospheric changes or environmental degradation. Attached to this discussion, was the issue of desert encroachment or desertification—the belief that the Sahara desert was expanding into the Sahel region—and whether it was caused by overgrazing and vegetation degradation or by climatic change (Grove, 1974; for a review see Nicholson, Tucker, & Ba, 1998). The potential causes generated much controversy, as did the question of whether or not there was a long-term trend towards more arid conditions (Lamb, 1974). If the drought continued, there was the choice of continuing with international aid or relocating several millions of people southward towards wetter and more sustainable areas; this, presumably, would have caused great stress on states further south.

In 1973, Reid A. Bryson was the first to suggest that it was the increase of carbon dioxide in the atmosphere, largely from the burning of fossil fuels, which intensified the greenhouse effect; and that this in turn altered wind circulation patterns and prevented tropical rains from migrating northwards into the Sahel (Bryson, 1973). Bryson therefore considered it erroneous to assume that the drought would end, and that the monsoonal rains would return to “normal,” that is, to the conditions that had prevailed earlier in the 20th century. Two years later, Jules Charney challenged Bryson’s hypothesis, suggesting a biogeophysical feedback model whereby the degradation of vegetation in the Sahel, mostly due to human activity (i.e., overgrazing), increased the albedo (reflectivity) effect, which in turn caused decreased rainfall (Charney, 1975). This model was later labeled the “Charney hypothesis,” and although confidence in the model has declined, it has remained topical (e.g., Giannini, Saravanan, & Chang, 2003; Hulme, 2001). Hence, the controversial question was whether the drought that started in the late 1960s was a consequence of human-induced climate change and its effect on climatic parameters on a wider scale, or whether it was caused by local activity in recent decades, namely overgrazing and land degradation (Seifert & Kamrany, 1974).

At that time, it seemed natural to assume that human activity in the Sahel would have had an impact on the area’s environment and thus impact upon the rainfall regime. The rains had been favorable throughout most of the first half of the 20th century (Nicholson et al., 2000), and this had a positive impact on regional development. Crop yields increased, as did cattle and population levels (FAO, 1960; Caldwell & Okonjo, 1968). Subsequently, when the drought started at the end of the 1960s, it was only natural to postulate that human activity, in any form, was a natural part of both the problem and the solution. A critical issue for the future development of the region was whether sedentary farmers and nomadic herdsmen should be able to continue their way of living after the drought.

The historians Paul E. Lovejoy and Stephen Baier provided a new interpretation of the situation in the Sahel, not by denying the fact that rainfall had diminished, but by questioning the effects and direct impact of human activity. In a study of the desert-side economy of central Sudan, the authors showed how economies in earlier centuries had adapted to droughts, but more importantly, they also showed that long-term droughts had occurred in the past (Lovejoy & Baier, 1975). Their analysis of chronicles from Agadez, Kano, and Bornu, resulted in a chronology that showed a recurring pattern of major droughts since the 16th century. Major droughts, the authors explained, were largely unheard of because the climate, since the 19th century, had been favorable and even wetter than in the 20th century. The results, especially those referring to wide-spread droughts in the mid-18th century, supported Jean Maley’s study of lake level sediments in Lake Chad (Maley, 1973) and added to Philip D. Curtin’s studies of economic development in Senegambia (Curtin, 1975). Lovejoy and Baier, argued, therefore, that the drought that began in the late 1960s was only part of a climatic cycle. They concluded that “unless the most pessimistic forecasters are correct that long-range climatic changes related to air pollution in the northern hemisphere will continue to produce drought conditions in the [S]ahel, the area will gradually recover” (Lovejoy & Baier, 1975; p. 580). Baier later argued that it was colonial policies earlier in the 20th century that hindered people from adapting to the drought in the early 1970s (Baier, 1976). Jeremy Swift (1977) made a similar argument, thereby stressing the need for historical context when studying historical climates in West Africa.

Most of the research in the 1970s was drought-driven. The purpose was not so much to investigate rainfall variability, as to show that large-scale droughts had recurred throughout Sahelian history. While multi-year droughts were apparently absent in the 19th century, researchers identified several in the 18th century; for example between 1738 and 1756; between 1770 and 1774, and finally in the 1790s (Curtin, 1975; Lovejoy & Baier, 1975; Maley, 1973). However, the need for a long-term climate chronology was evident. A comprehensive reconstruction would reveal whether large-scale droughts were part of a cyclic phenomenon; if they were human induced; if there was a long-term trend towards aridity, and whether the drought was in proportion to the diminished rainfall.

The First and Second Climate Chronologies

Based on her unpublished doctoral thesis from 1976, the meteorologist Sharon E. Nicholson became the first to construct a comprehensive climate chronology for West Africa (Nicholson, 1976, 1979). Nicholson’s chronology combined several types of information gained from natural proxy-data, landscape descriptions, maps, and historical records such as travel journals. Nicholson’s chronology was in many regards a synthesis of previous investigations; however, this was the first time that all previous research was combined to create a long-term climatic timeline.

Nicholson’s approach was purely climatological in the sense that she did not include the region’s historical development in the analytical approach. However, the aim was not only to depict the climate, but also to discuss methodological dilemmas that arise when investigating past climates. Historical climatology—the interdisciplinary marriage between climatology and environmental history (e.g., Brázdil et al., 2005)—was at this stage only starting to position itself as a discipline (Ingram, Underhill, & Wigley, 1978), and Nicholson’s article lay out the basic methodology for using historical records when investigating past climate change in West Africa.

Nicholson stressed that one of the many dilemmas that arise when working with past climates is that descriptions of weather and climate are both relative and subjective—eyewitness descriptions are products of their own time. The researcher must, therefore, avoid climatic anachronism and not imply present climate characteristics on past climates. For example, a dry year in the 18th century would by present standards not necessarily be described as dry, but may be normal or wet.

Nicholson also emphasized that historical famines are relative conditions. Claims of famines only imply that the harvest had failed or that food was unavailable. Famines should, therefore, not be presumed as evidence of drought. Famine-like situations may also occur because of war, pestilence, intense or heavy rainfall, or bad agricultural management. Accordingly, Nicholson argued that historical droughts cannot be directly compared to the droughts that affected the Sahel in the 1970s; not knowing the standard of living, the amount of food consumed, or the size of the population in historical societies, prevent comparisons across time.

From an interdisciplinary point of view, Nicholson emphasized the need to include climatological parameters when interpreting historical records. For example, recorded water scarcity in one area should not be interpreted as evidence of diminished rainfall in the same region. Areas in the Sahel and around Lake Chad are partially dependent on the water level in Lake Chad and the rivers feeding in to it, which in turn are dependent on the success of the rains further south in, for example, Cameroon. Therefore, a drought in one area may only imply reduced water availability, not that the rains had failed.

Nicholson’s chronology focused on establishing changes between wet and dry periods while also identifying short-term rainfall anomalies. The chronology spanned several thousand years, commencing with the late Pleistocene (35,000 bp), but its primary focus was on climatic changes in historic times. Nicholson emphasized the Sahel because, as a semi-arid region, its environment is more sensitive to rainfall variations. Information from the wetter Guinea Coast region was in general scarcer, which is why the chronology reflects the changes in the Sahel more so than further south.

The constructed timeline revealed that West Africa experienced several epochs of wet and dry periods between the 8th and 19th centuries. According to Nicholson’s chronology (1979), West Africa experienced a wetter phase from approximately the 8th to the 14th century. There is only a limited amount of information for this period, but the caravan trade across the El Djouf desert seems to have flourished, which would have been impossible in drier conditions. A study of Lake Bosumtwi (Talbot & Delibrias, 1977) indicated that wetter conditions also prevailed on the Guinea Coast until at least the 13th century; however, a trend towards more arid conditions then commenced, especially in the Sahel.

The Sahel, Nicholson argued, was wetter in the 16th and 18th centuries, and probably more so than in the 20th century. Contrary to this, the Guinea Coast experienced somewhat drier conditions in comparison to the 20th century. Indications of a wetter period in the Sahel were strongly supported by sediment and pollen data from Lake Chad (Maley, 1973), whereas drier conditions on the Guinea Coast were supported by a study of Lake Bosumtwi (Talbot & Delibrias, 1977). Information derived from biographies and travels journals written, for example, by Willem Bosman (1705), Thomas Winterbottom (1803), and John Matthews (1788) also indicated that the Guinea Coast was drier. Some of the key events that Nicholson employed to reconstruct the climate in the 18th century are shown in Figure 2.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 2. Some of the key climatic indicators used in Nicholson’s chronology for the 18th century. Evidence from the Sahel is on the top of the bar, whereas evidence from the Guinea Coast is presented below the bar.

Shortly thereafter, Nicholson (1980) extended her chronology into the 19th and 20th centuries. Several droughts were distinguished in the Sahel in the first half of the 19th century, which was at the same time that the Guinea coast experienced increased rainfall. The conditions then reversed in the early 20th century, when large areas of West Africa experienced decreased rainfall.

Nicholson’s chronologies were followed by further a climatic reconstruction produced by the historian George E. Brooks in 1986 (Brooks, 1986). Like Nicholson, Brooks’ chronology also had a long-term perspective. Beginning in 9000 BC, Brooks’ timeline continued well in to the last quarter of the 19th century.

Brooks’ spatial approach was different to Nicholson’s. First, Brooks defined West Africa as the area west of the inland delta of the Niger River and Bandama River, which excluded the eastern parts of the Guinea region. Second, Brooks implemented a historical perspective similar to that of Lovejoy and Baier (1975) in their study of the desert-side economies ten years earlier. He chose to reconstruct historical rainfall variability by following its impact on people, cattle and vector borne diseases within zones defined by the shifting patterns of the isohyets.

The basis for Brooks approach was that water availability governs human life and activity, thereby enabling a reconstruction of the climate solely by following the mobility of people. In other words, the ecological zones of West Africa are principally defined by the amount of rain that they receive, and as such they are suited for different types of socioeconomic activities. Where some activities thrive, other would perish. For example, the dry desert camel zone in the north was too arid for agricultural activity, and cattle did not thrive there as working animals. The camels had better carrying capacity and managed better without water than cattle. Conversely, camels were not as economically lucrative—measured in the value of produced meat, milk, and hides—in the wetter cattle zone further south. However, each zone was highly transitional due to rainfall variability, meaning that the boundaries that defined the economic activity shifted according to rainfall. Wetter periods made it possible for the trade routes to exist in otherwise arid regions by creating water holes. Conversely, weakened rains facilitated the use of horses in areas where they otherwise would have died due to vector borne diseases, which was an important advantage in war campaigns. Reversed conditions provided pasture for cattle in otherwise arid areas, and prohibited the use of horses, as water provided breeding grounds for vector borne diseases. For example, increasing aridity in the 18th century would have meant that the boundaries of the wetter ecological zones retracted towards the coast, which resulted in migration of people and animals. Hence, by studying the mobility of people between these demarcation zones, defined by the economic activity within them, Brooks managed to reconstruct historical rainfall variability.

Brooks’ explanatory model showed how the history of West Africa could be divided with regard to the effects of ecological and political stress factors. However, from a climatological perspective, it is necessary to bear in mind that tropical rain does not fall evenly, neither within a zone nor between two neighboring zones. Furthermore, the causality between climatic anomalies and human reaction to these events is not linear. Both drier and wetter periods might have affected food supply, but also resulted in the migration of people and cattle. Still, Brooks’ investigation elaborately shows the possibilities for establishing rainfall variability in areas that lack historical document, which is symptomatic for the interior parts of West Africa.

Brooks’ results, which also depicted centennial changes between wet and dry, diverged from Nicholson’s results. For instance, Nicholson suggested that the Sahel experienced a wetter phase in the 18th century, whereas Brooks suggested the opposite (see Figure 3). However, Brooks suggested that it was a dry period interspersed with abundant rainfall, whereas Nicholson suggested a wetter period intermixed with droughts and famines. Consequently, the dichotomy between the two chronologies might be more an issue of semantics than a product of analytical discrepancies.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 3. This figure shows the dichotomy between Sharon E. Nicholson and George E. Brooks’ climatic periodizations of the Sahel in West Africa. Whereas, for example, Nicholson argued that the early 18th century was wetter than the 20th century, Brooks argued that it was drier.

The Third Chronology

Ten years after Brooks’ chronology, the historian James L. A. Webb (1995) applied Brooks’ approach in a new attempt to shed light on the development of West Africa’s historical climate. Webb considered that the best approach to resolving the dichotomy between Brooks’ and Nicholson’s chronologies was to abandon the artificial construct of wet and dry periods. Webb asserted that the Sahelian and Savanna climates were more humid in the 17th and 18th centuries than in the 19th and 20th centuries. Webb based his conclusions on historical maps and travel accounts depicting the existence of rivers and lakes in western parts of Sahara. For example, whereas late 17th and early 18th century maps depicted rivers at Cap Timiris in Western Sahara (Mauretania), they had vanished from maps later in the 18th century. However, Webb also suggested that a long-term trend towards aridity began in the late 16th or early 17th century and continued into the mid-19th century. Identifying the increasing aridity was a challenging task, Webb argued, as rainfall variability and the increasing aridity undoubtedly was uneven between the ecological zones of semi-arid Sahel and the savanna grasslands. This explains why historical records do not register droughts in conjunction even though the evidence, in general, signaled a marked trend towards aridity across the region.

Webb’s explanation has a structured basis, and illustrates a long-term change in the mean rainfall while considering regional variability. It presents a solution that might explain the dichotomous results from Nicholson and Brooks. Nonetheless, these early chronologies were largely climatic generalizations that did not depict interdecadal or interannual rainfall variations.

The investigations conducted by Nicholson, Brooks, and Webb were the only ones that aimed specifically at reconstructing past climate. However, there have been other historical investigations that have included a climatic perspective in their analysis of the socioeconomic and political history of the region (e.g., Barry, 1998). As historians, their intentions have not been to reconstruct the climate; however, they have contributed to the general understanding of climatic variability by validating some events and providing valuable information on other previously unregistered droughts and famines. Nonetheless, the investigations by Nicholson, Brooks, and Webb provided the best overviews of historical rainfall variability in West Africa in the 17th, 18th and 19th century.

Stage 2

Identifying Historical Rainfall Variability

In 2012, Nicholson et al. (2012a) and Nicholson, Dezfuli, and Klotter (2012b) published the first rainfall chronology covering the last two centuries in West Africa. The new rainfall sequences were significantly different compared to Nicholson, Brooks, and Webb’s early climate chronologies. By including never before employed historical accounts and new methodological approaches the new chronologies were able to reconstruct rainfall variability on an interannual level.

For the 19th century, rainfall data from West Africa are fragmentary, interspersed, and sporadic, but long-term precipitation records exist from various areas governed by both the British and the French (see Table 1).

Table 1. West African Stations with Long Precipitation Records and the First Year of Regular Coverage (after Nicholson, Dezfuli, & Klotter, 2012b)


First year

St. Louis, Senegal


Banjul, Gambia


Freetown, Sierra Leone


Accra, Ghana


Aburi, Ghana


Lagoa, Nigeria


In addition to the stations noted in Table 1, there are multi-year records available from other locations. For example, there are data from near the Danish Castle of Christiansborg in Accra, Ghana, that cover the period from 1829 to 1834, and again between 1839 and 1842. Similar records can also be obtained from Senegal, near the French castle at Bakel, between 1856 and 1862 and, furthermore, from St Louis between 1830 and 1831 (Nicholson et al., 2012b). Most of the recordings were made at stations near the coastline and are generally sporadic. The reliability of the readings cannot always be verified and validating the records is a challenging feat in itself, as the readings cannot be calibrated and cross-referenced to other data. In addition, the data are, due to the nature of the tropical rains, not always a good indicator of precipitation over large areas.

In the 19th century, Europeans ventured on inland expeditions to explore unknowns part of West Africa. Many did so for a specific reason, such as searching for the fabled city of Timbuktu. In their quest for discovery, travelers sometimes made elaborate description of both climate and weather. A majority describe the occurrence of droughts and floods, but some records, such as those made by the German Heinrich Barth (1859), who crossed the Sahara desert and kept day-by-day records of weather and temperature. Barth also made detailed descriptions of Lake Chad, which facilitated Maley’s (1973) reconstruction of lake levels.

Nicholson et al. (2012a, 2012b) employed records such as these to construct a seven-class wetness index ranging from -3 to +3, wherein each year was labeled according to its wetness. In this index, 0 denoted normal rainfall conditions, whereas negative numbers indicated below normal rainfall and positive numbers suggested above normal rainfall conditions. For example, a dry year was denoted as -1, whereas -2 required that the sources described and explicitly mentioned droughts. The last value was reserved for years manifesting severe droughts with noticeable societal impacts, such as famines or migrations. At the other end of the spectrum, +3 indicated very wet conditions, such as floods and intense rainfall events that greatly impacted daily life.

The key in the interannual chronology created by Nicholson et al. (2012a, 2012b) was the use of a large number of geographical regions, each of which was shown to be homogeneous with respect to internal variability. This demanded detailed algorithms to weigh the qualitative historical and quantitative station records with respect to the given information. For a more detailed explanation on how the final values were obtained, and how the data from a specific region was evaluated, see Nicholson et al. (2012a, 2012b). Connecting the data from the 19th century with meteorological data from the 20th century (e.g., Lamb, 1985; Le Barbé, Lebel, & Tapsoba, 2002; Nicholson et al., 2000) facilitated the reconstruction of the first precipitation record that extends well back into the 19th century for the Sahel and the Guinea Coast (see Figure 4). The results of these studies, which detail rainfall variability in the Sahel and on the Guinea Coast starting from the 1800s, are presented below.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 4. The Sahel and Guinea Coast regions, to which the wetness index for the 19th and 20th century refers (after Nicholson, Klotter, & Dezfuli, 2012a).

Rainfall Variability in the 19th and 20th Century

The seven-class wetness index for the Sahel region in West Africa shows that arid conditions dominated in the Sahel in the beginning of the 19th century, thereby extending the dry period that commenced in the 1790s (Nicholson, 1979; see Figure 5). The rains regained some of their strength in the 1820s, followed by exceptionally arid conditions that lasted until the mid-1800s. The latter half of the 19th century show interchangeable conditions between wetter and drier periods. Nonetheless, more stable conditions with favorable rains dominated from the mid-1870s and the long interdecadal droughts that dominated in the early part of the century are absent.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 5. Sahel wetness index in the 19th century (after Nicholson, Klotter, & Dezfuli, 2012a).

On the Guinea Coast, rainfall fluctuated more than in the Sahel, and interannual variability was greater. Intermixed with periods of below average rainfall were, as can be seen in Figure 6, significantly wetter years. Interannual variations are high throughout the century, but long-term drought periods such as those in the Sahel are nonexistent.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 6. Guinea Coast wetness index in the 19th century (after Nicholson, Klotter, & Dezfuli, 2012a).

The beginning of the 20th century was clearly wetter than the same period in the 19th century. In the Sahel, the five first decades of the 1900s detail great interannual variations, but in general, the Sahel enjoyed good rains. The two most striking anomalies (see Figure 7) in the 1900s are a) the wet 1950s, and b) the dry period that started in 1968 and has prevailed ever since. There was some recovery of the rains between 1997 and 2003, not shown in Figure 7, even though the recovery was uneven between the western and central parts of the West African Sahel (Nicholson, 2005). Rainfall has, after the turn of the millennium, been highly variable, and generally remained below the average between 1920 and 2003 (Nash et al., 2016). Nonetheless, the 20-year pre-dominantly wet period during the 1950s and 1960s, serves as a radical contrast to the droughts that followed.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 7. Sahel wetness index for the 20th century (after Nicholson, Klotter, & Dezfuli, 2012a).

Although droughts have not troubled the Guinea Coast to the same extent as the Sahel, rainfall levels clearly diminished after the 1960s, especially when compared to the wetter period that the Guinea Coast experienced between the 1950s and 1970s (see Figure 8). On the Guinea Coast, the period prior to the 1950s showed more variation than in the Sahel, but data availability is somewhat better from the southern parts of West Africa, which may have some effect on the general presentation.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 8. Guinea Coast wetness index in the 20th century (after Nicholson, Klotter, & Dezfuli, 2012a).

One of the most striking characteristics of rainfall variability in West Africa is high interannual variability and the occurrence of interdecadal periods of wetter and drier conditions. Researchers have also identified a rainfall pattern that distinguishes the regions below and above latitude 10°N, referred to as the Sahel (north of 10°N) and the Guinea Coast (south of 10°N). The pattern, known as the “dipole” (opposite anomalies in the respective regions) and “non-dipole” (same anomalies in respective regions) perspectives, are beyond the scope of this entry (see e.g., Nicholson, 2008; Nicholson, 2011; Wagner & Da Silva, 1994 for more detail), but the figures presented here clearly show the prevalence of coherent precipitation patterns between the Sahel and the Guinea Coast. For example, both areas show increased rainfall between the 1950s and 1970s, as well as diminished rainfall in the post-1970s period. Indications of similar patterns are visible in the 18th century (Norrgård, 2015), but more research is needed before the existence of this precipitation pattern could be established also for this period.

Investigating Rainfall Variability in the 18th Century

Of the periods presented in this synopsis, the 18th century is the least investigated. The most detailed 18th-century rainfall chronology was published by the historian Stefan Norrgård, in 2015, in an article based on his doctoral thesis (2013). Norrgård relied on much of the basic methodology created by Nicholson (1979; 2012b), whereas he, like Brooks (1986) and Webb (1995), included a historical narrative and stressed the use of implementing historical development and context—the world, era, or milieu where the document was constructed—as an important part of the analysis. Historical context is essential as there are no records available from the interior parts of West Africa in the 18th century. Subsequently, records written by European slave traders (e.g., British, Danish, Dutch, and French) provide the best means for investigating rainfall variability. For general description of the climate in the 18th century, there are a number of travel journals and biographies written by slave-traders and adventurers. These include descriptions such as those provided by the British trader Archibald Dalzel (1793) and the Danish clerk Fredrik Rømer (1760). Many of these have been employed in previous studies (e.g., Barry, 1998; Nicholson, 1979; Webb, 1995;), but they need to be revisited as the methods improve and new climatic indicators are identified.

Spatiotemporally, Norrgård’s approach was significantly different from the investigations of Nicholson, Brooks, and Webb. Whereas previous investigations had focused on the Sahel region or West Africa in general, Norrgård focused on rainfall variability on the Gold Coast, which was a part of the larger region known as the Guinea Coast, in particular. Moreover, only the latter half of the 18th century was studied, which enabled a detailed comparison of the extracted information. Norrgård mostly relied on reports written by the Chief Governor, who resided at Cape Coast Castle (Cape Coast, Ghana) and was formally responsible for the day-to-day running on all British settlements. In his reports to the investors in London, the Chief Governor sporadically described the intensity and duration of the rains, but also its effects on British and other European forts and castles in the vicinity of Cape Coast Castle. The records contained no quantitative information, which was why the main indicators employed in Norrgård’s analysis, besides direct descriptions of the rains, were the impact of rainfall variability on the coastal settlements and the infrastructure (i.e., structural damages on buildings and roads), the trading in general, and the health of the British soldiers and other personnel. The advantage gained by employing the Chief Governors reports were that they were written by persons who held the same formal position within the slave trade, thus creating a continuity of specific topics and observations that previous investigations lacked.

The slave traders’ settlements—poorly constructed mud walls that were easily torn down during prolonged or intense rains—were built on the coastal dry-zone, a section of the coast with lesser rainfall than the surrounding areas. Extending across almost the entire coast of Ghana into Benin, this corridor was not suited for large-scale agricultural production (Owusu & Waylen, 2009). The crop-producing regions were in the wetter regions further inland (the Guinea Coast in general), while the villages and cities around the European settlements were mainly urban centers concerned with trading commodities (Kea, 1982).

Climatologically, the dichotomy between the wetter crop-producing regions in the inlands, and the coastal dry-zone, had Norrgård (2015) dividing rainfall in West Africa into three different regions: namely the Sahel, the “inlands,” and the coastal dry zone. This differs from the geographical division by Nicholson et al. (2012b), which is why the chronologies are not depicted on the same timeline in Figures 8 and 9; moreover, Norrgård’s wetness index is also based on a qualitative comparison of the provided data, not statistical algorithms, which is why Figures 8 and 9 are not entirely comparable. One of Norrgård’s arguments for his spatial subdivision was that the sources contained, despite any signs of ongoing wars, claims of famine further inland although the rains on the coastal dry-zone indicated anything from normal to wetter conditions. In contrast, years with weakened rainfall on the coast did not seem to affect the inlands by causing droughts leading to large-scale famines. Because the impacts of weaker rainfall were less distinct, years with heavy or excessive rainfall, which destroyed the settlements and made travelling by land and sea impossible, were more easily identified.

Historically, the coastal dry-zone, in contrast to the wetter inland areas, generated a distinct dilemma in Norrgård’s assessment of drought related famines. First, food availability depended on the crops further inland. Food shortages meant that the storages closed their doors, food rations became smaller, and food prices incremented. Food prices were only sporadically recorded, but restricted food availability meant that coastal communities started to prepare themselves for the onset of a famine. However, because the British did not have access to reliable information from the inlands, they could not always specify what caused famines, food shortages, or raised crop prices. Second, the British, unlike the slaves and the local free population, were not entirely dependent on the success of the local crops. Local crops were mainly used for feeding slaves, whereas the British imported food and beverages from Europe. The British could also, through their networks and use of ships, resupply by applying for food from other European settlements in West Africa. Consequently, droughts and famines, as caused by weaker rains, affected the coastal populations differently and were based on a hierarchical structure. Slaves were the most vulnerable; when food scarcity was imminent, they were the first to be left without food. Thereafter were the local free population, who were dependent on food availability through local food storages, and finally the slave traders, who through their network and contacts were the last to suffer from food shortages. It is common, therefore, that British accounts described famines or food scarcities that affected the slaves, and/or the surrounding villages and those living further inlands, but not themselves.

One indicator that was used to determine years with excessive rainfall even though rainfall descriptions were vague, was the health of the castles’ personnel. Due to high mortality amongst the Europeans within the first year of arrival, West Africa has been referred to as the “White Man’s Grave” (Curtin, 1990). Of those arriving in West Africa, between 25% and 75% died within the first year. The condition of the personnel is, therefore, a recurring topic in many of the British accounts, even during years when descriptions of the seasonal rains are absent. Most of the men died during the “sickly season,” which was synonymous with the rainy season and especially during years with heavy rains. Norrgård therefore suggested that intense rains increased the spread of vector borne diseases, thereby affecting the British and increasing the death toll. This approach was based on an investigation by Collins (1997) on the eastern parts of the Ghanaian coastline, which showed that the prevalence of malarial infection grew from 20% in the dry season to 36.6% in the rainy season. This radical increase in malaria infections during the rainy season in the 1990s might explain increasing deaths during wetter years in the 18th century. A direct link between vector borne diseases and increased deaths is indeterminable; however, records indicate a correlation between heavy rainfall and increasing deaths.

Information on rainfall variability in the 18th century is heavily affected by the slave trade itself. During the 18th century, the slave trade grew almost exponentially (Lovejoy, 2000). Caught in the areas inland of the West African coast, many of the slaves were marched southwards, where they awaited embarkation before being shipped across the Atlantic Ocean. All slaves had to be fed and watered, which also affected both food and water availability. Hence, as the number of slaves passing through the “gate of no return” grew throughout the 18th century, so did the use of food and water. An excessive use of water and food, especially during years with diminished rainfall, might have caused food shortages and depleted wells. For example, records show that the wells at Cape Coast Castle were depleted in the late 1770s, but whether this was a result of diminished rainfall or an excessive use of water is, due to scarcity of information, difficult to establish with certainty.

In conclusion, there are several uncertainties that affect the analysis when assessing interannual rainfall variability in the 18th century. Due to the undeterminable impact of the forced migration of millions of enslaved Africans and the paucity of information from the interior parts of West Africa, the challenges of interpreting 18th century records differ considerably from those in the 19th and 20th century. In the Sahel, rainfall variability has not yet been investigated. The droughts and famines that stressed the Sahel during the 18th century are only identified because of research conducted by Curtin (1975) and Lovejoy and Baier (1975).

The Gold Coast and the Sahel in the 18th Century

There is scarce evidence from the 1750s; however, the rains were relatively stable between the 1750s and the late 1770s (see Figure 9). There were periods with intense rainfall during the 1750s, especially in 1752, 1753, and 1758, and again in the 1760s. A famine stressed parts of the Guinea Coast at the end of the 1760s, which was at the same time that the rains caused massive deaths on the coast. Besides the famine, there are no records indicating drought like conditions near the British settlements. The rains were favorable until the late 1770s, when the first signs of weaker rains appear (Norrgård, 2015). Ponds and wells dried up, and the Dutch in the neighboring castle Elmina, had to apply for water from the British, thereby indicating that the rains diminished over a larger area.

Changes in Precipitation Over West Africa During Recent Centuries

Figure 9. Wetness index for the coastal dry-zone (after Norrgård, 2013; Norrgård, 2015; Nicholson, 2017).

The beginning of the 1780s were drier, with recorded famines in either 1780 or 1782. The dating varies between some sources; however, a British governor, who had spent several years on the coast, described the rains at Cape Coast Castle in 1782 as the worst in many years. The rains in the 1780s show high interannual variability, but they gradually regained their intensity, and wetter conditions prevailed after the latter part of the 1780s and throughout most of the 1790s (Nicholson & Norrgård, forthcoming; Norrgård, 2015).

Between the 1750s and 1800s, the onset of the rains varied between April and June, whereas the rainy season lasted between May and June, often subsiding in early July (Norrgård, 2015). There are some uncertainties regarding the secondary rainy season on the coast. The accounts from Cape Coast Castle never mention a secondary rainy season, while some travel journals do and others do not. The secondary rainy season may have been significantly weaker in the 18th century, thus explaining the discrepancy between the historical records.

Rainfall variability in the Sahel or interior parts of West Africa is very sporadic before the mid-18th century and cannot be reconstructed on an interannual level. A period of intermittent droughts troubled the Senegambia regions and Niger bend in the Sahel between 1738 and 1756. A general famine affected the Sahel region between 1747 and 1750. Rainfall was clearly intermittent across the region. Senegambia experienced a famine in 1752 (Barry, 1998), whereas Senegal may have been more humid (Nicholson, 1979). Records from Senegal suggest that the rains were heavy in this region in 1754 (Webb, 1995). The rains probably gained their strength after the mid-1750s and continued favorable until late 1760s. The rains then gradually weakened, and the first signs of desiccating conditions appear in 1768, when the rains probably failed over a larger area in the Sahel (Norrgård, 2015). Thereafter, famines were registered in Timbuktu, 1770–1771, and in Trarza, 1771–1775 (Nicholson, 1976). The 1780s appear to have been wetter. The water level in Lake Chad was clearly higher in the 1780s, which was at the same time that Agadez (Niger) was distressed by heavy floods. The conditions then changed rapidly in the 1790s. Droughts forced Agadez to be evacuated (Nicholson, 1979), and there are indications of droughts in Kano (Northern Nigeria) between 1793 and 1795 (Nicholson, 2012b), and food shortages in the western Sahel (Norrgård, 2015).

The most noteworthy changes in precipitation in the 18th century were the rapid changes that occurred in the 1790s in both the Sahel and the coastal dry zone. Historical accounts are insufficient to determine whether West Africa, as a region, was wetter or drier in comparison to later periods. However, investigations of Lake Bosumtwi (Shanahan et al. 2009), indicate that the 18th century may have been even drier than today.

Future Research

Research has shown that the droughts that started in the Sahel in the late 1960s were meteorological, and possibly amplified by human activity (Giannini, Saravanan, & Chang, 2003). However, whether they were caused by the burning of fossil fuels, as suggested by Bryson (1973), is still unanswered. Nonetheless, the investigations presented in this article clearly show that the Sahel has experienced several drought periods throughout history and, furthermore, that similar rainfall patterns have prevailed during recent centuries. Investigations conducted since 2010 have significantly increased the climatological community’s understanding of historical rainfall variability in West Africa; however, in comparison to other regions of the world, it is still limited. Extending the rainfall record is still of considerable importance, and there is a great variety of sources that could facilitate detailed reconstructions. Combining and cross-referencing historical records from different European archives and, furthermore, comparing with lake level data, and improving the three-ring chronology would greatly increase our understanding of past rainfall variability in West Africa.


Adams, W. M., & Mortimore, M. J. (2005). Working the Sahel. London: Routledge.Find this resource:

    Baier, S. (1976). Economic history and development: Drought and the Sahelian economies of Niger. African Economic History, 1, 1–16.Find this resource:

      Barth, H. (1859). Travels and discoveries in North and Central Africa, 1849–1855, III. London: Longman.Find this resource:

        Barry, B. (1998). Senegambia and the Atlantic slave trade. Cambridge, U.K.: Cambridge University Press.Find this resource:

          Bosman, W. (1705). A new and accurate description of the coast of Guinea, Divided into the Gold, the Slave and the Ivory Coasts. Cambridge, U.K.: Cambridge University Press.Find this resource:

            Brázdil R., Pfister, C., Wanner, H., von Storch, H., & Luterbacher, J. (2005). Historical climatology in Europe: The state of the art. Climatic Change, 70(3), 363–430.Find this resource:

              Brooks, G. E. (1986). A provisional historical schema for Western Africa based on seven climate periods. Cahiers d’Études Africaines, 23(101–102), 43–62.Find this resource:

                Bryson, R. A. (1973). Drought in Sahelia: Who or what is to blame? Ecologist, 3(10), 366–371.Find this resource:

                  Buckle, C. (2004). Weather and climate in Africa. Harlow, U.K.: Longman.Find this resource:

                    Caldwell, J. C., & Okonjo, C. (1968). The population of tropical Africa. Studies in Family Planning, 1(29), 10–12.Find this resource:

                      Charney, J. (1975). Dynamics of deserts and drought in the Sahel. Quarterly Journal of Royal Meteorological Society, 101(428), 193–202.Find this resource:

                        Collins, K. (1997). Malaria-related beliefs and behaviour in Southern Ghana: Implications for treatment, prevention, and control. Tropical Medicine and International Health, 2(5), 488–499.Find this resource:

                          Curtin, P. D. (1975). Economic change in precolonial Africa: Senegambia in the era of the slave trade. Madison: University of Wisconsin Press.Find this resource:

                            Curtin, P. D. (1990). The end of the “white man’s grave”? Nineteenth-century mortality in West Africa. Journal of Interdisciplinary History, 21(1), 63–88.Find this resource:

                              Dobson, M., & Ziemann, B. Introduction. In M. Dobson & B. Ziemann (Eds.), Reading primary sources (pp. 1–18). London: Routledge.Find this resource:

                                Fage, J. D. (1965). Some thoughts on migration and urban settlements. In H. Kuper (Ed.), Urbanization and migration in West Africa. Berkley: University of California Press.Find this resource:

                                  Food and Agriculture Organization of the United Nations (FAO). (1960). The state of food and agriculture 1960. Rome. Italy.Find this resource:

                                    Food and Agriculture Organization of the United Nations (FAO). (1973). The state of food and agriculture 1973. Rome. ItalyFind this resource:

                                      Giannini, A., Saravanan, R., & Chang, P. (2003). Oceanic forcing of Sahel rainfall on interannual to interdecadal time scales. Science, 302(5647), 1027–1030.Find this resource:

                                        Grove, A. T. (1974). Desertification in the African environment. African Affairs, 73(291), 137–151.Find this resource:

                                          Hulme, M. (2001). Climatic perspectives on Sahelian desiccation: 1973–1998. Global Environmental Change, 11(1), 19–29.Find this resource:

                                            Ingram, M. J., Underhill, D. J., & Wigley, T. M. L. (1978). Historical climatology. Nature, 276(23), 329–334.Find this resource:

                                              Kea, R. A. (1982). Settlements, trade and polities in the seventeenth-century Gold Coast. Baltimore, MD: Johns Hopkins University Press.Find this resource:

                                                Lamb, H. H. (1974). Last year’s weather. drifting towards drought. Geographical Magazine, 46, 455.Find this resource:

                                                  Lamb, P. J. (1985). Rainfall in subsaharan West Africa during 1941–83. Zeitschrift Für Gletscherkunde und Glazialgeologie, 21, 131–139.Find this resource:

                                                    Le Barbé, L., Lebel, T., Tapsoba, D. (2002). Rainfall variability in West Africa during the Years 1950–90. Journal of Climate, 15(2), 187–202.Find this resource:

                                                      Lovejoy, P. E., Baier, S. (1975). The desert-side economy of the Central Sudan. The International Journal of African History, 8(4), 551–581.Find this resource:

                                                        Lovejoy, P. E. (2000). Transformations in slavery. A history of slavery in Africa (2d ed.). Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                          Maley, J. (1973). Mechanisme des changements climatiques aux basses latitudes. Palaeogeography, Palaeoclimatology, Palaecology, 14, 193–227.Find this resource:

                                                            Maley, J. (1993). Chronologiecalendaire des principales fluctuations du lac Tchad au cours du dernier millenaire. In D. Barreteau & C. V. Graffenreid (Eds.), Datation et Chronologie dans le Bassin du Lac Tchad (pp. 161–163). Paris: Orstom.Find this resource:

                                                              Matthews, J. (1788). Voyage to the River Sierra Leone on the Coast of Africa (1785–87). London: B. White.Find this resource:

                                                                McCann, J. C. (1999). Causation and climate in African History. The International Journal of African Historical Studies 32(2/3), 261–279Find this resource:

                                                                  Mortimore, M. (1989). Adapting to drought. Farmers, famines, & desertification in West Africa. Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                                    Nash, D. J., & Adamson, G. C. D. (2014). Recent advances in the historical climatology of the tropics and subtropics. Bulletin of American Meteorological Society, 95(1), 131–146.Find this resource:

                                                                      Nash, D. J., De Cort, G., Chase, B. M., Vershuren, D., Nicholson, S. E., Shanahan, T. M., et al. (2016). African hydroclimatic variability during the last 2000 years. Quaternary Science Reviews, 154(15), 1–22.Find this resource:

                                                                        Nguyen, H., Thorncroft, C. D., & Zhang, C. (2011). Guinean coastal rainfall of the West African Monsoon. Quarterly Journal of the Royal Meteorological Society, 137(660), 1828–1840.Find this resource:

                                                                          Nicholson, S. E. (1976). A climatic chronology for Africa: Synthesis of geological, historical, and meteorological information and data. (Doctoral dissertation), University of Wisconsin.Find this resource:

                                                                            Nicholson, S. E. (1979). The methodology of historical climate reconstructions and its application to Africa. Journal of African History, 20(1), 31–49.Find this resource:

                                                                              Nicholson, S. E. (1980). The historical climatology of Africa. In T. M. L. Wigley, M. J. Ingram, & G. Farmer (Eds.), Climate and history (pp. 249–270). Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                                                Nicholson, S. E. (2005). On the question of the “recovery” of the rains in the West African Sahel. Journal of Arid Environments, 63(3), 615–641.Find this resource:

                                                                                  Nicholson, S. E. (2008). The intensity, location, and structure of the tropical rainbelt over West Africa as factors in interannual variability. International Journal of Climatology, 28(13), 1775–1785.Find this resource:

                                                                                    Nicholson, S. E. (2009). A revised picture of the structure of the “monsoon” and land ITCZ over West Africa. Climate Dynamics, 37(7), 1155–1171.Find this resource:

                                                                                      Nicholson, S. E. (2011). Dryland climatology. Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                                                        Nicholson, S. E., Dezfuli, A. K., & Klotter, D. (2012b). A two-century precipitation dataset for the continent of Africa. Bulletin of American Meteorological Society, 93(8), 1219–1231.Find this resource:

                                                                                          Nicholson, S. E., Klotter, D., & Dezfuli, A. K. (2012a). Spatial reconstruction of semi-quantitative precipitation fields over Africa during the nineteenth century from documentary evidence and gauge data. Quaternary research, 78(1), 13–23.Find this resource:

                                                                                            Nicholson, S. E., & Norrgård, S. (forthcoming). Rainfall in Africa in the 1780s. In S. White, C. Pfister, & F. Mauelshagen (Eds.), The Palgrave handbook of climate history. London: Palgrave Macmillan.Find this resource:

                                                                                              Nicholson, S. E., Some, B., & Kone, B. (2000). An analysis of rainfall conditions in West Africa, including the rainy season of the 1997 El Niño and 1998 La Niña Years. Journal of Climate, 13(14), 2628–2640.Find this resource:

                                                                                                Nicholson, S. E., Tucker, C. J., & Ba, M. B. (1998). Desertification, drought, and surface vegetation: An example from the West Africa Sahel. Bulletin of American Meteorological Society, 79(5), 815–829.Find this resource:

                                                                                                  Nicolini, G., Tarchiani, V., Saurer, M., & Cherubini, P. (2010). Wood-growth zones in Acacia seyal Delile in the Keita Valley, Niger: Is there any climatic signal? Journal of Arid Environments, 74(3), 355–359.Find this resource:

                                                                                                    Norrgård, S. (2013). A reconstruction of the climate during the slave trade era, including an analysis of the climatically facilitated trans-Atlantic slave trade. (Dissertation), Åbo Akademi University.Find this resource:

                                                                                                      Norrgård, S. (2015). Practising historical climatology in West Africa: A climatic periodisation 1750–1800. Climatic Change, 129(1), 131–143.Find this resource:

                                                                                                        Owusu, K., & Waylen, P. (2009). Trends in spatio-temporal variability in annual rainfall in Ghana (1951–2000). Weather, 64(5), 115–120.Find this resource:

                                                                                                          Parker, J., Rathbone, R. (2007). African history: A very short introduction. Oxford, U.K.: Oxford University Press.Find this resource:

                                                                                                            Pilcher, J. R., Baillie, M. G. L., Schmidt, B., & Becker, B. (1984). A 7,272-year tree-ring chronology for Western Europe. Nature, 312(5990), 150–152.Find this resource:

                                                                                                              Seifert, W. W., & Kamrany, N. M. (1974). A framework for evaluating strategies for the development of the Sahel-Sudan region (vol. 1). Summary report: Project objectives, methodologies, & major findings. Center for policy alternatives. Cambridge. MA: Massachusetts Institute of Technology.Find this resource:

                                                                                                                Shanahan, T. M., Overpeck, J. T., Anchukaitis, K. J., Beck, J. W., Cole, J. E., Dettman, D. L., et al. (2009). Atlantic forcing of persistent drought in West Africa. Science, 324(5925), 377–380.Find this resource:

                                                                                                                  Schöngart J., Orthmann, B., Hennenberg, K. J., Porembski, S., & Worbes, M. (2006). Climate-growth relationships of tropical tree species in West Africa and their potential for climate reconstruction. Global Change Biology, 12(7), 1139–1150.Find this resource:

                                                                                                                    Swift, J. (1977). Sahelian pastoralists: Underdevelopment, desertification, and famine. Annual Review of Anthropology, 6, 457–478.Find this resource:

                                                                                                                      Talbot, M., & Delibrias, G. (1977). Holocene variations in the level of Lake Bosumtwi, Ghana. Nature, 268, 722–724.Find this resource:

                                                                                                                        Thorncroft, C. D., Nguyen, H., Zhang, Z., & Peyrillé, P. (2011). Annual cycle of the West African monsoon: Regional circulations and associated water vapour transport. Quarterly Journal of the Royal Meteorological Society, 137(654), 129–147.Find this resource:

                                                                                                                          van der Geest, K., & Dietz, T. (2004). A literature survey about risk and vulnerability in drylands, with a focus on the Sahel. In T. Dietz, R. Ruben, & A. Verhagen (Eds.), The impact of climate change on drylands (pp. 117–146). New York: Kluwer.Find this resource:

                                                                                                                            Vershuren, D. (2004). Decadal and century-scale climate variability in tropical Africa during the past 2000 years. In R. W. Battarbee, F. Gasse, & C. E. Stickley (Eds.), Past climate variability through Europe and Africa (vol. 6; pp. 139–158). Dordrecht, Netherlands: Springer.Find this resource:

                                                                                                                              Wagner, R. G., & Da Silva, A. (1994). Surface conditions associated with anomalous rainfall in the Guinea coastal region. International Journal of Climatology, 14(2), 179–199.Find this resource:

                                                                                                                                Webb, L. A., Jr. (1995). Desert frontier ecological and economic change along the Western Sahel. Madison: University of Wisconsin Press.Find this resource:

                                                                                                                                  Weiss, H. (2008). Västafrikanska krönikor som källor för historisk ekologisk-klimatologisk forskning om 1700-talets samhällskris. Möjligheter och begränsningar. Historisk Tidskrift för Finland, 93(1), 38–55. (In Swedish).Find this resource:

                                                                                                                                    Winterbottom, T. (1803). An account of the native Africans in the neighbourhood of Sierra Leone. Cambridge, U.K.: Cambridge University Press.Find this resource:

                                                                                                                                      Worbes, M. (2002). One hundred years of tree-ring research in the tropics: A brief history and outlook for future challenges. Dendrochronologia, 20(1–2), 217–231.Find this resource: