Harvest water, not just crops. New mapping methods help planners see where water is available dry areas

Run-off strips (the bare strips between crops) harvest rainwater at ICARDA. This arrangement can double water-use efficiency with no loss in yield.

The majority of the world’s farmers live in low-rainfall areas. These farmers could substantially increase their harvests if rainfall were more plentiful. But in fact, water is not as scarce as we think. Even limited supplies can be sufficient, if they are used wisely. That’s where the science of rainwater harvesting comes in.

Water harvesting simply means collecting rainwater which would otherwise flow away, and storing it for later use, for drinking, washing, or irrigating crops. Rainwater harvesting specifically for agriculture, is slightly different. The principle is simple. All lands receive rain, but not all lands are suited for agriculture. The solution: deprive some fields of their share of rain. Allow the water to flow away as ‘runoff’, and redistribute it to more productive land. The collected runoff is stored in the soil, behind dams or terraces, in reservoirs and cisterns, or seeps into the soil to recharge natural aquifers. A vital but scarce resource — water — is thus targeted more effectively, resulting in higher yields and bigger profits from the same amount of rainfall.


Old tradition, new science

Types of water harvesting systems Old ideas still work. This 2000-year old rainwater harvesting system still sustains agriculture in Yemen’s mountain zone. There are many different types of water harvesting systems, but all have three major components: a catchment or source area that provides the water, a storage facility, and a target area where the stored water will be used. Water harvesting systems can be broadly classified as micro-catchments or macro-catchments. Micro-catchment systems are small-scale, designed for a single farm. The source and target areas are close together, often in the same field. The water is stored in the soil’s root zone, where plants can absorb it immediately, or in a small reservoir for later use. Macro-catchment systems are much bigger in scale, with several small target areas sharing a large catchment. Runoff water from the catchment is stored within each farm in small reservoirs, underground cisterns, or in the soil.

ICARDA has helped design and promote rainwater harvesting methods in more than 30 countries, using modern science to improve traditional methods that farmers have used for centuries. One example is the use of Geographic Information Systems (GIS) for planning and building efficient water harvesting systems.

Eddy De Pauw, is the former head of ICARDA’s GIS unit. He’s worked in 22 countries, developing new GIS techniques and new applications for existing techniques; and training researchers and development planners on the use of these tools. He explains how ICARDA is using GIS to design water harvesting projects, and improve water-use efficiency in areas where every drop counts.

 “Location is the key,” he says. “If you build a water harvesting system at the wrong location, it won’t work. Using GIS, we can evaluate thousands of potential locations quickly and accurately. It’s much cheaper and much faster than traditional land survey methods.”

 
Choosing the right spot

Whether a location is suitable for harvesting rainwater depends on several factors: rainfall, slope, soil characteristics (depth, texture, salinity), land use, land cover and geology. ICARDA scientists have researched these factors for 15 years, in a dozen countries. The ICARDA  team is  now using this knowledge, applying sophisticated GIS techniques to analyze site characteristics to identify the best places to build water harvesting systems.

The project was launched following a request from the Syrian government, which was planning a massive countrywide effort to harvest rainwater. The ICARDA team worked with national research centers to collect information from various sources: rainfall records, soil maps, geological surveys, agricultural records, and remote sensing data on vegetation and land cover. To accurately measure slope and topography, they used a ‘digital elevation model’ constructed using radar measurements taken from satellites. Next, they used GIS tools to match this information to the broad requirements of 14 different kinds of water harvesting systems: a generalized all-purpose macro-catchment system and 13 micro-catchment systems, based on different combinations of techniques and crop groups.

“We had to devise a method to analyze data from thousands of potential sites across the country,” Dr De Pauw explains. “The method had to be quick, simple and low-cost, because ours was only a proof-of concept study. The full study would have to be done by national R&D agencies with limited budgets, limited staff, and very little experience in either GIS or water harvesting.

” First, the team had to develop an accurate, consistent method to compare different sites. At each site, each criterion (e.g. slope) was measured and scored, and the individual scores combined, after statistical manipulation, to give a combined score for the site -easier said than done. For example, a 10% slope is suitable, while 5% is not. But what about 9.9%? To avoid such classification traps, their software programs used ‘fuzzy logic’ membership functions programmed to automatically ‘moderate’ scores whenever needed.

Technologies for development

Map 1. Suitability for micro-catchment water harvesting in Syria on a scale of 0-100. A series of maps were created, for different combinations of water harvesting methods and crop groups. This map is for one combination, contour ridges + rangeland shrubs.

Slightly different approaches were required for micro-catchment and macro-catchment systems (see box for definitions). For micro-catchments, each individual factor was scored on a 0-100 scale. The combined score for the site was calculated using the ‘loser-takes-all’ method, where the combined score is simply the lowest individual score. If a site receives high scores for most criteria, but zero for one criterion (perhaps the land is absolutely flat, so no runoff is possible), then it cannot be used, and receives a combined score of zero. After several months of number crunching, the team created a map of the entire country, showing suitability for micro-catchment water harvesting on a scale of 0 to 100 (Map 1).

To identify areas suitable for macro-catchment systems, two separate assessments were undertaken: the first to evaluate suitability to serve as a catchment, and the second to evaluate suitability as a target area, with the additional constraint that both areas should be within a certain distance of each other. Map 2 shows the suitability for macro-catchments. Areas in blue are highly suitable as catchments, areas in red are highly suitable as target areas. The ideal locations are those where blues and reds lie close to each other.

Map 2. Suitability for macro-catchment water harvesting. The whole of Syria was mapped; only a small portion is shown here.

“This fast-track, low-cost method can be used at different levels, from global to province or district level,” says Theib Oweis, Director of ICARDA’s Integrated Water and Land Management Program. Following the study in Syria, the method is now being used in Libya, as part of a five-year US$14 million project funded by the Libyan government. The aim is to map water-harvesting potential across the country; and then build micro- and macro-catchment water harvesting systems to stabilize and improve crop and livestock production. It’s a good example of ICARDA’s approach: not just research, but research-for development, using the best scientific tools to address the most basic needs: food and water.

For more information, contact ICARDA’s Geographic Information Systems Unit.