Water scarcity and drought affect over one billion people in arid, semi-arid and dry sub-humid areas, threatening livelihoods and food security. In these areas, the only available water may be surface runoff during rainy seasons, or shallow groundwater. This policy brief describes traditional ways of harvesting such water, outlines what improvements modern science offers, and discusses the implications for decision-makers, identifying both constraints and recommendations.


Water sustains life. Since the dawn of civilization, humankind has had to work hard to survive where water is limited. Dry areas (arid, semi-arid and dry sub-humid zones) cover 52 million square kilometres — more than one third of the globe's land surface. Most of this is rangeland, where annual winter rainfall is less than 200 millimetres, or desert, where it is less than 100 millimetres. Many of the billion people affected by water scarcity live in developing countries.

Their situation is not set to improve in coming decades. Global climate change will alter rainfall patterns, and by 2050 many more countries will probably suffer frequent droughts. Such droughts can have devastating effects at local, regional and national levels. For example, the Moroccan drought from 1979 to 1984 forced 40 to 50 per cent of the population to move. The large rivers that carry water into some dry areas may also falter. In 1997, a drought in the Middle East deprived more than 500 Iranian villages of drinking water and reduced the flow of the Tigris and Euphrates rivers in Iraq by about 80 per cent, seriously constraining irrigated production. In Syria, 329,000 people had to sell their livestock, and required urgent food assistance. [1]

As populations grow, they will increase the pressure on scarce water supplies, potentially escalating conflicts, particularly between countries that share water resources, for example in the Middle East.

To prepare for these changes, researchers must assess and predict their likely impacts and develop appropriate responses. Policy makers will need to work with communities to spread new ideas, using accessible programmes that are compatible with local traditional knowledge.

Indigenous water management

Communities gather their indigenous knowledge through years of interacting with their natural environment. The people of arid regions have a wealth of traditional technologies for extracting and using limited water supplies. They use their knowledge as the basis for community-level decisions about water management.

Building on such indigenous knowledge can be a particularly effective way of helping poor communities, as it uses an asset over which they retain direct control. Traditional knowledge understands and respects the limits set by natural systems. Traditional techniques are usually sustainable because they are self-regulating, in that there is a limit to the maximum amount of water one can withdraw. But where human populations (and their cattle) are increasing, these limits are more frequently breached — particularly where modern technology, like mechanical water pumps, is brought in to help without considering sustainability.

Harvesting rainwater

Many communities harvest rainfall during wet seasons, and this is often a major source of water in drylands. Rainwater-harvesting was first practised extensively in the Middle East. In the ancient city of Petra, in Jordan, rainfall was being channelled to large reservoirs dug in the rock 9,000 years ago. Surface reservoirs were also used in China and India more than 4,000 years ago, to collect rainy season runoff for use in the dry season. And in Yemen, at least 1,000 years BC, people were divert­ing runoff water to irrigate 20,000 hectares of farmland that may have fed up to 300,000 people. [2]

Rainwater-harvesting systems still support many communities in arid regions:

  • Rooftop water harvesting provides low-cost water for drinking and household uses. Above-ground storage systems can hold between 5 - 20 cubic metres per building, but underground stores can hold up to 50 - 60 cubic metres, depending on the amount of rainfall, water demand, and intended use. Many people in Latin America the Middle East and sub-Saharan Africa, regularly use rooftop water-harvesting, and communities in some areas, like China's Gansu Province, have extended the idea, collecting water from other hard surfaces like roads.
  • Cisterns are usually underground reservoirs of rainwater, with capacities ranging from 10 – 500 cubic metres. Runoff water from an adjacent catchment is channelled to the cistern, sometimes via basins that let sediment settle out. Water is usually drawn from the cistern by hand, using a bucket and rope. Cisterns remain the only source of drinking water for people and animals in many drylands. In some areas, such as Jordan and Syria, they are usually small pear-shaped holes dug into the rock, about 3 - 6 metres deep and no more than 3 – 5 metres in diameter. But in northwestern Egypt, Bedouins dig much larger cisterns under the natural rock layer below the soil surface. These can hold 200 – 300 cubic metres of water. The collected water is mainly kept for human and animal consumption, but is sometimes also used to irrigate vegetables and fruit trees in domestic gardens.
  • Hafaers, also called khadens, are usually earthen surface ponds created by herders to water livestock. They can be found across the Middle East, where they store between a few hundred and several thousand cubic metres, and in India, where they can reach tens of thousands of cubic metres.

All these systems are simple to install and sustainable. But people often still need some capital to acquire component parts. Desert people (for example Bedouins and nomads) are generally poor, with low incomes and limited technical know-how. These people rarely attract government subsidies, loans, or incentives, making their struggle for water more difficult.

Groundwater extraction

Dryland communities also use water that has infiltrated the ground. But only 20 to 75 per cent of water extracted from underground aquifers is recovered through natural recharge, causing a continuous decline in available groundwater. [3] Several methods are available for extracting groundwater, depending on local environmental conditions. The most popular include:

  • Open galleries, which are open channels cut vertically down to about a metre below the water table, that collect and channel groundwater. They are widely used in northern Sinai and the northwest coastal zone of Egypt, where the water table is shallow, needing little excavation.
  • The mawasi system is used in sand dune areas where the water table lies less than a metre below the surface, for example in the Rubelkhali desert of Saudi Arabia, where shallow water-harvesting wells, 1–1.5 metres deep, are dug by hand.
  • Qanats are slightly sloping tunnels that run from a ground water source at one end to a downhill surface at the other. Gravity draws water out, without any need for mechanical devices. Qanats can be tens of kilometres long, in which case vertical shafts may be sunk for inspection and maintenance . This indigenous technique is particularly used in Algeria, Iran, Syria and other countries in the Middle East.
  • Sand-filled reservoirs have the advantage of reduced evaporation and are used in Botswana, Chad and Namibia. Sand can store considerable amounts of water between particles, and also acts as a filter, so the water supply is clean enough for livestock, domestic purposes or small-scale irrigation

Fog harvesting

Fog harvesting is a common practice in many arid countries, including Chile, Peru and Yemen, where rainfall is rare but dense fogs settle over the land for some period of the year. Water is harvested by creating the right conditions for the fog to condense, for example by spreading out sheets of metal or polyfibre, and then channelling the water into underground tanks. In Yemen, tribal people on the south-east coast historically channelled the condensed water so they could collect it from rock faces. But now there are many organisations, like FogQuest, looking at fog harvesting in arid lands and improving upon the basic technology.

Modern science can help manage water resources

Although many traditional techniques are still used, modern science can often improve them. For example, modern pumps, either manual hand-pumps or mechanical devices driven by renewable energy, can efficiently extract water from cisterns, ponds and aquifers.

Modern science can help increase cistern efficiency. Improvements in pumping, catchment materials and hydrological studies help refill cisterns more efficiently during the rainy season. New, low-cost materials, like paraffin wax or watertight surface covers like plastic sheets, can increase runoff by making the ground less permeable. A project in Iraq has shown that applying paraffin wax to the soil surface can triple the runoff from small plots. [4]

Rooftop water harvesting has also improved. Water collected from modern tiled roofs, or corrugated iron ones, is clean and suitable for many domestic purposes. Many regions of the world, including industrialised countries, collect rooftop water for domestic consumption, irrigation and landscape gardening. For example, the Frankfurt International Airport Terminal II building in Germany harvests rainwater from greenhouse rooftops and is used for sanitation and irrigation of plants.

In addition, scientific advances can help use water more efficiently. For example, drip irrigation has been successfully introduced, at low cost, in the Rift Valley Province of Kenya. [4]

Water rights can hinder development policies

But water access for dryland communities will not be improved simply by introducing modern technologies. These areas need effective policies that take other considerations into account. Land tenure in dry rangelands, for example, can severely constrain development. For instance in Syria, rangelands are largely public, so water resources remain undeveloped as farmers do not want to invest in lands they don't own. New policies, sensitive to national and global socio-economic developments, are needed to encourage sustainable investment in dry lands' water resources. A good example comes from Jordan, where allocating tribal drylands to individual Bedouins has increased property values and provided the incentive to harvest water.

Water rights in deserts and rangelands are often poorly defined. But tapping water in the upstream reaches of a wadi, or valley with an ephemeral stream, may adversely affect people living downstream. Avoiding conflict requires clear policies on water rights. Groundwater resources are also common property in most dry areas and, with the help of modern pumps, are often over-exploited, decreasing the quality and quantity of the available water. Policies that regulate water pumping can help re-establish ground water resources, although few have been implemented. But one successful example is the Azraq oasis in Jordan where, despite years of ground-water decline, the oasis was restored after a government and Global Environment Facility project started regulating pumping in 1993.

Recommendations for policymakers

Policymakers should address the constraints outlined above, but there are further actions they can take to improve water access for dryland communities:

  • Offering payment for the 'ecosystem services' drylands provide, such as storing carbon or maintaining biodiversity, is one way of raising finance to improve sustainable water access for local communities. In Columbia, Costa Rica and Nicaragua communities are paid to conserve biodiversity. [5] And in Central America, payment schemes reward good watershed management. [6] Investing public money in dry areas can prove profitable. A study by the International Food Policy Research has shown that in semi-arid India, public investment in high yielding crop varieties, road infrastructure, electricity and education gave positive rates of return, often with higher returns than irrigated districts. [7]
  • Empowering dryland communities is crucial for effective development. Policies should encourage local involvement in decision-making processes, and use indigenous knowledge. Where necessary, community institutions should be strengthened, for example through awareness and training programmes.
  • Eco-tourism can bring additional income to desert areas, but it also increases the demands for water from tourists and also from locals, as living standards improve. Eco-tourism can encourage investment in accessible water supplies, but this investment risks damaging fragile ecosystems and depleting water resources unless it is sustainable. Policies must balance development needs with environmental protection.