Text and photo: Marko Kallio
Our entire society is based on the sufficient availability of clean water. Still, clean water cannot be taken for granted. Natural examples of this include arid deserts and steppes and, in particular, dry conflict areas in Syria and Yemen, among others. In these areas, sufficient amounts of water may not be physically available. However, clean water can also be a socioeconomic or administrative question.
Water poverty means a situation where a sufficient amount of clean water is not available in order to fulfil basic needs. It is a multifaceted problem which is reflected in the entire society and may result from social structures, any lack thereof or a permanent or temporary drought.
Southeast Asia has a tropical monsoon climate which divides the year into two extreme seasons: drought and heavy rainfall. In Laos, more than 90% of the annual rainfall (1,000–3,500 mm) takes place during the rainy season between May and November. The heavy rainfall causes floods, destroys roads and damages infrastructure. During the dry season, the problem is that sufficient amounts of water may not be available for all villages.
This problem is emphasised at the end of the dry period when the last rains have passed months ago. At this time, temperatures also soar high, rising close to 40 °C. However, the availability of water, the level of the infrastructure and the socioeconomic position of the population are not static or random factors; they vary according to the time and place following a well-known principle, the Tobler’s First Law of Geography.
The water poverty index is an indicator of water and its management
In my thesis work (and later as part of my postgraduate studies), I examined the division of water poverty in Laos between geographic locations and seasons. Laos is one of the poorest and least developed countries in Asia. Its government has set a goal of exiting the list of the least developed countries in the world by 2020. The key reason why the country still remains on this list is malnutrition, which afflicts as much as 30% of the non-adult population. Understanding the reasons for water poverty is particularly important for a country that depends on the cultivation of rice, i.e. the sufficient availability of water.
In my study, I calculated a multivariate water poverty index (WPI) for 8,215 villages located in different parts of Laos using openly available data. The WPI consists of five components: the amount of water, access to clean water, the capacity of water resources management, the use of water and the state of the (hydrological) environment. In principle, my data sources included census data and hydrological modelling results regarding the availability of water. I applied spatiotemporal data mining methods to the calculated index and its components.
The availability of water is not the main reason for the lack of safe water in Laos
In a geographic sense, the distribution of water poverty was as expected. The best situation is near Vientiane, the country’s capital, and along the Mekong river on the border between Thailand and Laos. The situation near the capitals of each province is better than in more rural areas, and the most difficult in mountainous and sparsely populated regions next to Vietnam and China. This geographic distribution is fairly similar during both the dry and rainy season.
What was surprising was that the availability of water does not seem to be the determining factor in terms of water poverty in Laos. Instead, differences between regions can be explained by socioeconomic and infrastructure-based factors, such as poverty, the level of education, sources of tap water, irrigation systems and the distance between a village and the closest administrative capital. Therefore, the WPI of villages in the poorest position does not increase at all during the rainy season, whereas that of villages in a better position increases by as many as dozens of points. As the WPI is measured using a range from 0 to 100, this is a major improvement.
When reasons for water poverty are examined as processes, agricultural elements are emphasised. This is logical, considering that 70% of the population of Laos live in rural areas where agriculture is the primary source of income. In a geographically weighted regression analysis, eight out of ten key variables are related to agriculture. However, their importance varies so that reasons related to agriculture are more significant in sparsely populated areas and less significant near larger cities. In particular, the dependence of the income of village populations on agriculture is also a good indicator of water poverty. It seems that the agricultural population of Laos is in the weakest position in terms of water.
Spatial data in key role in understanding reasons for water poverty
The geographically weighted principle component analysis I have developed further during my postgraduate studies confirms the same conclusions: Socioeconomic factors are in a determining position across the entire country when examining water poverty. However, subsequent factors vary according to the time and place. Environmental factors and income from agriculture are emphasised in northern parts of Laos.
In southern parts, it was not possible to sufficiently distinguish different processes from one another. This may mean that the variables used to calculate the WPI are not fit to examine the processes in the South. Therefore, an analysis based on spatial data is necessary in order to understand local differences related to water poverty. This analysis also showed that a single set of indicators cannot be applied to all parts of the country. This is not apparent, unless the analysis is location-specific.
My study shows that spatial data is closely linked with water-related problems, also aside the physical avail- ability of water. This was the first study in which the WPI was calculated for a large number of villages. This was also the first time when the WPI and its components were studied by means of data mining. Spatial data turned out to be invaluable in examining water poverty, but also in understanding its underlying processes.
It should also be noted that, even though my study focused on one of the least developed countries in the world, it was fully conducted using openly sourced material. The tools used were also open source spatial data programs: QGIS and R. The extended study based on the thesis work was published in Social Indicators Research in December 2017. The data used in the study is available here.