2. Defining Water Scarcity

Quantitatively defining water scarcity:

In my previous post I loosely defined water scarcity as a function of supply and demand. However, to actually determine who is and isn’t water scarce we must quantitatively define and measure these. This is no easy task and has led to much contention (Damkjaer & Taylor 2017).

Fig. 1 The same conceptual diagram from my first post showing water scarcity as a function of supply and demand (created by me for this blog). This post investigates how we quantify these concepts.

The Water Stress Index (WSI)

This was the first metric established to quantitatively measure water scarcity. It was developed by Malin Falkenmark in 1989 to investigate the link between water scarcity and famine in Sudan in the 1980s and has since become the most widely used water scarcity metric in the world. According to Falkenmark…

- Water supply is defined as mean annual river runoff (MARR).

- Water demand is defined as 1700m³ of water/person/year (for domestic, agricultural and industrial use).

The WSI refers to the annual national balance between supply and demand (calculated using the equation below). Countries are then categorized according to their WSI value (shown in the table below). 

Fig. 2 The equation used to calculate the WSI

Fig. 3 Water stress categories that are assigned to countries according to their WSI value (Source: Damkjaer & Taylor, 2017)

Is Africa water scarce?

The map below shows national WSI values for countries in Africa, using data from 2014 (Damkjaer & Taylor 2017). The pale pink countries have more than 1700 m³/capita/year and therefore are classified as ‘no stress’ (i.e. water sufficient). This map implies that contrary to common perception... the majority of Africa is actually water sufficient! However, we are constantly bombarded with news stories about how people in Africa don't have enough water to meet their basic daily needs. So what's going on? The WSI is widely used because it is so simple, both to understand and to calculate. However… is it an over-simplification? Is it really a representative metric of water scarcity?

Fig. 4 Map showing national-scale water scarcity (using data from 2014) as defined by the WSI and therefore based entirely on MARR (Source: Damkjaer & Taylor 2017)

 

Problems with the WSI definition of SUPPLY:

Firstly, it is based solely on river runoff. It does not account for other important sources of freshwater such as groundwater (all water flows and stores below the ground surface) and green water (all water stored in plants or in the soil). Despite the fact that 30.1% of the world's freshwater is groundwater and only 1.3% is surface water (Shiklomanov, 1993).

Fig. 5: Diagram showing the global distribution of freshwater, divided into surface water, groundwater and glaciers & ice caps (Graphics: IX Power Art Department;Data: Shiklomanov, 1993).

Secondly, it is based on mean annual river runoff, therefore does not account for any temporal variation in supply, both within and between years. For example, in countries with seasonal climates, water surpluses during the wet season may mask significant water shortages during the dry season. 

Fig. 6 Conceptual graph to show how seasonal river runoff (due to seasonal rainfall) may cause water stress in the dry season, even if the annual river runoff implies the country is water sufficient, assuming demand remains constant (diagram created by me for this blog)

Thirdly, it does not consider water quality. Just because water is available does not mean it is fit for use (particularly for domestic and agricultural purposes).

Fig. 7 Polluted Fen River, China (Source: National Geographic) 

Finally, it equates water supply with water availability, with little to no regard for water accessibility. In fact, there is no statistically significant relationship between freshwater availability and freshwater access (Damkjaer & Taylor 2017).

Problems with the WSI definition of DEMAND:
The WSI assumes water demand is fixed and universal. Yet in reality, water demand varies hugely across the world, largely due to climate and level of development. Warmer climates and wealthier people tend to ‘require’ more water. Therefore, perhaps a single water requirement value for everyone in the world is unrepresentative. Maybe we should have different water requirement values for different regions... but where do you draw the line between how much water people want and how much water people need? People living in central Africa only use 2% of the amount of water used by those living in North America, but how much of this is due to actual differences in need? Do we need a perfect green lawn, a swimming pool in our backyard or a bath every day?

***

The shortcomings of the WSI have been recognised and there have been numerous efforts to design more representative indicators (Damkjaer & Taylor 2017). However, I shall not be going into detail about these. My next post further explores the WSI misrepresentation of water supply and tries to determine whether or not Africa is actually water scarce.