Remote control

Remote sensing and web 2.0 tools make water use more efficient in Mali

Sander J. Zwart
Lucie M. C. Leclert
Wim Bastiaanssen

Managers of the Office du Niger irrigation scheme in Mali are using remote sensing data to analyse the efficiency of the system without having to physically check the infrastructure. The information will help them prepare for future expansion.

In Mali, the Office du Niger irrigation scheme delivers water from the Niger River to a large rice-farming area. The scheme’s managers wanted to improve the efficiency of the system to cope with future expansion, and to allow for competition for water resources from the development of nearby commercial sugar cane estates.

In 2009, WaterWatch, an advisory firm based in the Netherlands, tested a low-cost remote sensing-based methodology to analyse the efficiency of the Office du Niger irrigation system. While the remote sensing techniques used in the study had been thoroughly tested by others researchers. the methods had rarely been applied to assess the performance of an operating irrigation system. Previous studies had also combined data from field inspections and used only low-resolution remote sensing maps. This was the first study to assess irrigation performance using only remote sensing technology, and with maps providing a high level of detail.

The Office du Niger, situated in the Ségou region of Mali, is one of the oldest and largest irrigation schemes in West Africa. It was developed in the 1920s, and has been in operation ever since.Today, the 80,000 hectares of rice fields provide approximately 465,000 tons of rice each year – around 40% of the national production – which makes this farmland extremely important for Mali’s food security.

Rice grows mainly during the rainy season (approximately June to December). Water is abundant at this time, and the entire irrigation network is continuously filled with water. Farmers can irrigate as long as they have paid a seasonal contribution (a redevance). In recent years, the government has promoted a second rice season (March to June) by lowering the redevance.

The entire system is divided into five administrative zones, each responsible for land and water management in their part of the irrigation scheme. The zones are subdivided into casiers, which are further sub-divided into blocks representing, on average, ten farmers. Each farmer is responsible for the irrigation and drainage of an individual rice field.

Even out

In an effort to keep assessment costs low, the researchers used images from the Landsat7 satellite (among others). The Landsat programme has been gathering high-resolution satellite images since 1984, and all archived material can be accessed for free from the United States Geological Survey global visualization (Glovis) website. The researchers also made use of information from the nearby Ségou meteorological station, including air temperature, humidity and wind speed data.

The WaterWatch team used the data to look at four factors in their assessment of the Office du Niger scheme: cropping intensity – a measure of the ‘greenness’ of an area; productivity of water – the rice yield compared to the amount of water used; uniformity – whether the water is evenly distributed throughout the system, and; head-tail performance – a comparison of the amount of water consumed at the start of the irrigated area (head) and at the end (tail).

Analysis of the satellite images showed that there was vegetation in much of the Office du Niger area during the main growing season, but also during the second growing season. Very little rain falls at this time, so the farmers must have been using the irrigation system. This suggests that reducing the redevance has allowed farmers to grow a second rice crop.

The study showed that rice yield varies between the five administrative zones. One explanation is that the water is unevenly distributed throughout the system. An ideal irrigation system distributes water equally to all users, but under practical field conditions several factors – such as soil type, water quality and fertilisation – affect water consumption in different fields and zones.

Low yields can also be an indicator of problems with the irrigation system. Poor maintenance of canals and structures, inadequate management and illegal water use can, for example, mean that farmers with crops at the tail of the schemes have lower yields than those at the head. The study revealed that estimated yields at the tail end of the Office du Niger system were 18% lower than at the head end.

In one casier, the study found high yields at the head and tail of that system, but lower yields in the middle. According to the zone manager, the casier receives sufficient water for everyone, but due to the bad design of a weir located halfway along, less water reaches the fields beyond that point. The increase in water consumption at the lower end of the casier is caused by excess drainage water gathering there.

The research suggests that the lower yields at the tail ends can, at least in part, be attributed to irrigation water management, although factors such as deterioration of water quality may also play a role. The analysis shows that there is significant scope for improving the productivity of the water resources in the Office du Niger system. The construction of a well-functioning drainage network, for example, could lower salinity levels, reduce flooding and improve crop yields.

Working knowledge

The WaterWatch team worked closely with the irrigation scheme managers to gather feedback, discuss the limitations and advantages, and find ways to refine the assessment methodology in the future. In the case of the casier with the faulty weir, for example, the management knew about the problem, but they found that the remote sensing analysis helped to show the physical impacts on the crops.

The managers were able to bring their in-depth knowledge of the day-to-day workings of the system to help the research team interpret their results. The study confirmed many of the managers’ perceptions of the problems affecting system performance. The water reaching crops during the rainy season, for example, is higher than
the actual demand, causing low efficiency of irrigation water application, water logging and water losses to drainage areas. By scheduling water supplies to match demand, managers can improve efficiency and water productivity.

However, incorporating remote sensing data into the routine monitoring and evaluation work of managers would require changes to the current system and an investment to train staff to interpret the images. The methodology used by the researchers would also need some improvements to provide more accurate results.

For example, it was not possible to distinguish between rice crops and other plant types – weeds, for instance – in the analysis of the images used in this study. Future research could focus on improved methods to discriminate rice from non-rice vegetation, and on the accurate estimation of yield from rice-only producing areas. The results of the study could also be linked to data from soil maps, surface water flows and water quality analysis to determine the causes of low yields.

During the discussion of the study, the Office du Niger and donor agencies showed great interest in the results and possibilities that remote sensing offers. The managers saw a chance to identify and define priority areas for rehabilitation, and also to map crops that make use of the irrigation system but which are not part of the official casier areas.

WaterWatch has applied the methodology to agricultural areas all over the world, working with different crop types and environments. This gives the team the chance to continually review and adapt their techniques to give farmers a more accurate view of the strengths and weaknesses of their irrigation systems. And Water Watch now uses web 2.0 tools to make study results more easily available to irrigation managers.

For recent projects in Egypt and South Africa, researchers have integrated soil quality and remote sensing data to be viewed in Google Earth. Irrigation engineers can use the application to get a high quality, detailed picture of the land, and identify areas where water is wasted and where more irrigation is needed. By keeping the cost as low as possible and making the data readily available, WaterWatch hopes their methodology can be applied to more irrigation schemes, and that eventually farmers start using remote sensing data as a routine part of their work.

---------------------------------------------------------------------------------------------------------------------------------------------

Sander J. Zwart is a remote sensing and GIS specialist with the Africa Rice Cente r ( www.warda.org). Lucie M. C. Leclert is an irrigation specialist/social scientist at Euroconsult MottMcDonald, and Wim Bastiaanssen is director of WaterWatch (www.waterwatch.nl).

---------------------------------------------------------------------------------------------------------------------------------------------

Related Links

Water use in the Western Cape vineyards
WaterWatch has produced Google Earth maps showing water use in wine grape farms in the Western Cape, South Africa.
www.waterwatch.nl/grapes

Office du Niger
www.office-du-niger.org.ml

United States Geological Survey Glovis image archive
http://glovis.usgs.gov

30 August 2010

Copyright © 2014, CTA. Technical Centre for Agricultural and Rural Cooperation (ACP-EU)