==Field experiences==
In the last few decades extensive civil engineering investments have been made, mostly in large spate irrigation systems in Yemen, and to a lesser degree in Pakistan, Eritrea and Tunisia. The track record of all these large civil engineering investments is at best patchy. Investments in flow division and regulation in Pakistan (for instance on the Gaj Nai in Sindh) have performed reasonably well, but the same cannot be said for modern flow diversion structures. An evaluation of 47 relatively minor spate systems built with national funding in Balochistan between 1960 and 1990 established that only 16 were still operational.
In the Tihama plains of Yemen the designs of the modernised systems became more sophisticated over time. The first scheme to be modernised, Wadi Zabid, suffered from serious water diversion and sedimentation problems, and is now operated rather like a traditional system with diversion essentially controlled by bunds constructed by bulldozers. Later schemes included effective diversion and sediment handling arrangements, albeit at a high investment cost. A profound change in some large modernised systems is that they ceased to be managed by farmers. The most extreme manifestation has been in Yemen, where the Tihama Development Authority (TDA) assumed full responsibility for operation and maintenance after the civil works on the various systems were completed. TDA has struggled to find the funds needed to carry out the maintenance, particularly removing sediments from canals, required to keep the schemes functioning. In other cases where new structures have been provided responsibility for maintenance is ambiguous, particularly when local capacity to manage complex civil engineering projects is limited.
The main lessons learned from these experiences are:
* The planning and design of the rehabilitation and improvement works have mostly been carried out without effective partnership with farmers and land users. Farmers’ valuable knowledge of spate irrigation, and their preferences regarding the scope and type of works and changes in the layout of their irrigation system were often not properly considered during the design process.
* The investment costs have been very high and it is doubtful if they can be justified in purely economic terms.
* The operation and maintenance of the larger diversion structures and canal systems, is difficult and expensive. In particular sedimentation at intakes and in canals is often not properly controlled in ‘modernised systems’.
* New structures have promoted larger inequity in the distribution of irrigation water due to the collapse of traditional evolving water rights – “modernised” diversion structures give much larger control over spate flows to favoured groups of the upstream farmers.
==Reference manuals, videos, and links==