Paper Review: Rainfall and Water Resources Variability in Sub-Saharan Africa during the Twentieth Century

As I have pointed out in previous posts, rainfall and water resources have a key impact on the ability of farmers in Africa to irrigate their crops. This is where I felt it would be interesting to review a paper focusing on this issue and how the access to water has high variability. The paper, written by D. Conway, A. Persechino, S. Ardoin-Bardin, H. Hamandawana, C. Dieulin and G. Mahé, is titled ‘Rainfall and Water Resources Variability in Sub-Saharan Africa during the Twentieth Century'. It will be interesting to look into one area where there is much variability, sub-Saharan Africa that has a serious problem with water shortages. In this post I will study this paper in detail, attempt to draw some conclusions from it and find places where the research may be lacking or need extending.

The aim of this paper from the Journal of Hydrometeorology is to study the spatial and temporal covariability of rainfall and river flows in the sub-Saharan region of Africa. The research is categorised into four key regions, central, south, east and west. The data used in this research come from a variety of sources, mainly the Institute de Recherche pour le Développement (IRD) and other national and international sources. The focus of the paper is on the changing rainfall and river flow levels on a temporal scale from 1901 – 2002, the entire 20^th century, to see if the levels were variable throughout this century. The research studies these trends at a basin scale rather than a nation state level as this highlights that the focus is on the hydrological issues as hand rather than the political issues. Data on rainfall and river flows in sub-Saharan Africa is not perfect, there is often a lack of data for certain regions and the chronology may be incomplete or lack temporal resolution (not be recorded frequently enough). In my opinion, this paper is attempting to come to some more general conclusions at a regional level for rainfall and river flow rather than focusing on a single river basin to carry out this research. Although this provides more conclusions, it can lack detail and miss certain local differences that may also have an influence in rainfall and river variability such as water that is diverted for irrigation at a smaller-scale. There is also a concern with some of the records dating back to the early 20^th century. The data collection techniques from this time are likely to be far less sophisticated and this means that some of the conclusions may be lacking in certainty.

So what did this research actually find? One of the most interesting points that I picked up from reading this paper was that through time there has actually been a shift rather than a trend in the West African rivers. This shift occurs between the two time periods mentioned, 1931-1960 and 1961-1990. The authors go on to state that, ‘Proportionally, the shift is much greater in river flows (from -13% to -51%) than rainfall (from -7% to -14%).’ This means that river flows during the second period are a step below those in the early 20^th century. This is even more interesting when we know that the three other regions in this research do not show much variation between these two time periods, therefore West Africa is far more hydrologically unstable (see graphs in Figure 1). Something occurred during the 20^th century in West Africa that has not in the other regions. The big shift in river flows but not rainfall suggests that something or someone is altering the basin environment and changing how the river flows. Although nothing is mentioned in the article as to what has caused this, I personally feel this may have something to do with large scale dam and irrigation projects that are used in West Africa in many basins to boost food production through irrigation. This would account for the large variation in river flows but less variation in rainfall. 

Figure 1: Annual rainfall (black line) and river flow (gray line) for rivers in West Africa

Another of the figures in the paper (Figure 2) shows the strength of the relation between rainfall and runoff for different rivers studied. Many of the rivers show very strong relationships between rainfall and runoff over the thirty year periods. It is then intriguing to look at which regions have strong and weak relationships and question why this is the case. The overall relationships suggest that variations in rainfall account for the majority of runoff variation. In the instance of East Africa, the geology and unique Rift Valley system accounts for the unusual patterns observed there. Lake Victoria outflows show a very weak relationship to basin rainfall.

Figure 2: Strength of regression between rainfall and runoff for the 30 year periods

To conclude, this paper gives a very clear overview of rainfall and runoff trends through the 20^th century. It is important to remain critical of data in all instances. Data in this research is from a variety of sources that may be inaccurate or incomplete which can leave us drawing conclusions where there are none. I feel that this paper has achieved the aim of looking at these two hydrological characteristics both spatially and temporally and put forward some interesting data and conclusions to better understand the background to water resources in the sub-Saharan Africa region. Personally, I find it very interesting to look at some of the reasons behind the results that this piece of research has put forward despite the possible uncertainty with the results.