Previous Page Table of Contents Next Page

Essential water delivery policies for modern on-farm irrigation management

CM. Bun, Irrigation Training and Research Center, Department of Agricultural Engineering, California Polytechnic State University, San Luis Obispo, USA


Successful on-farm irrigation scheduling requires that water delivery projects adopt a service concept, and that water deliveries be made with a high degree of reliability. An arranged delivery schedule and seasonal volumetric allocations are additional requirements. Until these conditions are met, attempts to implement modem on-farm irrigation scheduling techniques to maximize water use efficiency and yields will be only partially successful.

The agronomic concept of irrigation scheduling is to apply water to the crop in the correct amounts and at the proper times to maximize crop production and/or profit, while maintaining a reasonably high irrigation efficiency. A typical irrigation project authority concept of irrigation scheduling is to develop and implement a schedule of water deliveries which is compatible with the water delivery system capabilities and constraints, with the least amount of inconvenience. It is desirable for an irrigation project to combine the two concepts by developing an irrigation scheduling policy which extracts the most flexible and reliable performance available from the water delivery system to meet the agronomic goals.

Various theoretical computer models have been developed which examine the complex hydraulics of water delivery systems together with the evapotranspiration/soil-water balance concepts in the field. This author believes that the most beneficial uses of such models are associated with training or with examining general patterns of water use. They generally have little or no practical use in terms of implementing a desirable schedule of water deliveries on a real-time basis if the goal is to maximize personal farmer initiative and crop yields. Farmer initiative is very important, as it will impact many on-farm decisions, the willingness to pay for water, and the desire to maintain (or not destroy) project facilities.

Successful irrigation scheduling must be simple to implement and easy to understand, from both the farmer and project management personnel standpoints. Practical irrigation scheduling policies must also recognize that hydraulic, weather, crop, and soil conditions are variable and dynamic. Irrigation scheduling programmes must be robust enough to withstand physical and human uncertainties.

This paper points out a number of factors which the author feels are very important for implementing a viable irrigation water management scheme which includes irrigation scheduling as one component. The author has seen the importance of these factors on projects throughout the world. Examples from California are used to illustrate how they impact projects which have already implemented various levels of modern on-farm irrigation management.


A discussion of irrigation scheduling must consider the degree of flexibility which is to be offered to the farmer (or to a group of farmers). Flexibility has three components: (i) frequency; (ii) flow rate; and (iii) duration. There are numerous variations of these three components, and many authors have developed terminology to describe various types of delivery schedules, such as 'rotation', 'arranged', and 'limited rate demand'.

In theory, it would be nice if farmers had water available on complete demand. In practice, areas with very advanced on-farm irrigation systems, such as California, do well with an arranged schedule. The following conditions are generally required in California for an arranged schedule if optimum on-farm irrigation efficiency and crop yields are to be obtained:

· The flow rate must be large enough to minimize labour requirements.

· The farmer should be able to open and close his own turnout (offtake) without needing a project employee present.

· The water can be ordered within 24 hours' advance notice. If a farmer has a known weekly schedule, the water supplier can provide that schedule without being re-notified.

· The water is available on a volumetric basis, which allows both the farmer and district to understand that there is an upper limit to the amount of water available for a season/year. Furthermore, the flow rate and volume delivered per irrigation is controllable and is known to the farmer in advance.

· The water delivery can be shut off at any time by the farmer with minimal or no advance notice.

· The flow rate and duration of irrigation can be changed from one irrigation event to another.

Additional water management flexibility is sometimes obtained downstream of the offtake with reservoirs, or with conjunctive use of well water.

It must be acknowledged that in many countries, this degree of flexibility cannot be given to individual farmers because the fields are too small. However, this degree of flexibility may be given to a larger farmer unit area of 50 ha.

Certainly, not all farms in California receive water with as much flexibility as described above. However, as water management becomes more important, most California water districts are rapidly moving toward providing such flexibility.


A key idea abides within the flexibility criteria outlined above: the irrigation project does not dictate the schedule of irrigations on a farm. Irrigation project authorities rarely have the expertise, manpower, or ability to determine the optimum schedule of irrigations on individual fields, much less implement that schedule. How many engineers have seen the well-defined rotation schedules displayed prominently on boards in irrigation projects, and then realize that they exist only on the board? We should recognize the failure of such attempts and move on to other options.

Irrigation project authorities should be cognizant of the fact that their responsibility is to provide service, not to dictate schedules. The only way this concept of 'service' can possibly work is if the water delivery system is broken into 'levels of service'. The main canals provide service to the secondary canals, which in turn serve the distributaries, which serve the offtakes, etc. Obviously, all canals and levels cannot provide service of the same flexibility or reliability. For example, run-of-the-river schemes have inherent limitations that reservoir-supplied projects do not. Nevertheless, when the concept of service is wholeheartedly adopted by project staff, amazing agronomic and social improvements can be made.

Of course, providing service means a shift in power and attitudes. It requires a shift from pride in power to pride in service. The prospect of relinquishing power may also bring up the following question: How can project managers prevent anarchy if the project management is not able to dictate water delivery schedules?

This question often ignores reality. In many projects one should ask: Can the present state of water delivery anarchy be alleviated with more realistic water delivery schedules (and operational guidelines) at all levels in the delivery system?


We must remember that farming is risky business, and the farmer is the one and only person who truly takes a risk. Therefore, the farmers should be given water with as much flexibility as possible so that they can manage their water well and minimize their risk. The irrigation project, however, only has a certain amount of water available, and there are system design constraints (which can often be overcome) that put limitations on flexibility. The irrigation project authorities, once they buy into the service concept, will do their best to provide the maximum flexibility possible.

How is water wastage minimized if flexibility is provided? Quite simply, farmers must be allocated a known and reasonable volume of water. When that volume is used up, there is no more water forthcoming unless the farmer can purchase water from other farmers. Does this mean that modern on-farm irrigation scheduling is impractical on most existing projects until volumetric deliveries are established? This author believes that the answer is yes.

There is another point to consider: there is no justification for project authorities to concern themselves with the details of developing on-farm irrigation schedules. Those details are too complicated for central authorities to tackle. Instead, the burden should be placed on the farmer, who will willingly pick up that burden, if the volume of water is known and guaranteed.

This is not to say that project authorities should not assist farmers with demonstrations and technical assistance regarding irrigation scheduling. If the demonstrations and technical assistance are simple and worthwhile, the farmers will adopt the recommended practices. If 'modern' irrigation scheduling schemes have not been adopted by farmers, these schemes must not have provided substantial, identifiable benefits.

This discussion highlights several key points:

· Good water management and irrigation scheduling require volumetric allocations and volumetric deliveries. How can farmers begin to think in terms of proper depths of irrigation (i.e., irrigation scheduling) unless this is so?

· There must be an ability to withhold water from farmers once they use up their allotment.

· The water volume must be provided in a fashion that it is usable and beneficial. If a volume is allocated, but that volume is only available in an unreliable or untimely manner, the programme will fail.

Such a volumetric delivery programme requires extensive training programmes to make farmers aware of the concepts of water depths and volumes, basic crop water use, and soil water holding capacity. These programmes do not depend upon sophisticated soil moisture monitoring equipment, remote sensing, or other such elaborate devices/tactics.


Irrigation scheduling discussions often concentrate upon crop water requirement computations and the degree of flexibility needed for water deliveries, just as was stated in the previous paragraphs.

We must not lose sight of the fact that unless the water deliveries are reliable, confidence in a new programme will be immediately lost and irrigation scheduling programmes will be disasters. Given a choice between a programme having high flexibility but unreliable deliveries (due to operator errors, technical difficulties, politics, etc.) versus a less flexible but very reliable system, the latter is preferred.


FAO has been a leader in publishing information about weather-based evapotranspiration predictions. Numerous conferences, including this one, provide additional insight into the mathematics and details of computerized on-farm irrigation scheduling. Irrigation scheduling computer programs have been available through government programmes and commercial irrigation scheduling companies in the United States since the early 1970s. Noting that there are always exceptions to any rule, the author believes that the two major benefits of these computer programs have been:

· They have provided education for many farmers. If, however, the major benefit has been education rather than real-time decision support, one must ask if there is an easier way to transfer the information to farmers.

· They provide a data base, or accounting of irrigation for farmers with multiple crops and fields. Although this may be important for many farmers in the United States with diversified crops, it is not a typical need of small farmers in most countries.

In California, well under 5% of the farmers personally own and use such computer programs for real-time information, even though they are readily available. Government assistance programs in California have largely moved away from providing real-time scheduling and soil monitoring of individual fields. Rather, the computer programs are often used by water districts or the California Department of Water Resources (the CIMIS program) to provide an estimate of ETo, and perhaps an estimate of daily or weekly ET for the most popular crops in an area.

The author does not want to make light of these programs - but wants to put things in perspective. In California and other locations in the United States, there are many farmers who do hire commercial irrigation scheduling services just as they hire soil fertility specialists. These irrigation scheduling services provide information which, when combined with other information such as the need for tillage and spraying, enables the farmer to make a more informed decision.

A small farmer with only a few hectares of land will not suffer from a lack of computerized irrigation scheduling information. Such a farmer does not require sophisticated soil moisture sensing equipment either. In general, the most limiting factors for this farmer's excellent on-farm water management are the lack of flexible and reliable water deliveries, and poor land grading.


The following steps must be followed in order to take improved water management off the books and into the field:

1. Recognize that 'irrigation scheduling' tools are just that, tools. Irrigation scheduling programmes, whether they are computer programs or one-on-one demonstrations of how to estimate soil moisture 'by feel', are means to an end, not the end. The goal should be improved water management, not the implementation of a specific irrigation scheduling programme.

Perhaps the best way to have proper implementation of the various tools is to require that significant numbers of senior and junior irrigation project personnel receive practical and pragmatic training and experience with on-farm irrigation. This training and experience requires more than taking field trips and tours of farms. Although this may sound somewhat ridiculous to many engineers, the author believes that until people try to live with their own rules, and the consequences of those rules, they will not understand how to improve those rules. This is especially true with on-farm irrigation, with its complexity of changing soil infiltration rates, unpredictable weather patterns, difficulties with agronomic issues, and constraints of water delivery. Rules and formulas which appear to be very logical in the office may be completely useless at the on-farm level or to the operator of a check structure.

The recommendation, then, is that in many projects we can move forward faster by first taking a step backward and making certain that all of the decision-making personnel have the basic knowledge and hands-on experience they need in order to make proper decisions.

2. The service concept must be adopted by the water delivery agency so that water can be delivered with the highest possible degree of reliability and flexibility. This requires that decisions be decentralized, so that the management becomes responsive rather than dictatorial.

This requires several years of training at all levels. In California, the ITRC has assisted irrigation districts in this effort by training all levels of management and field personnel at different times, but on the same subject. Through this intensive training, all employees share a common vision of the purpose of the irrigation project. They also learn more about the responsibilities and constraints of others within the hydraulic network. Management must also have open communications with the field personnel, so that good ideas can filter upward and affect policy implementation.

A responsive system of any type also requires that information on the status of water levels and flows at key points in the irrigation network, as well as water orders, be readily available to canal operators and project management. Therefore, a key first step is to improve communications, transportation, and monitoring points throughout a canal system. For example, real-time knowledge of the flows at the inlets and downstream ends of all major canals can greatly assist with making the correct decisions for canal operation.

3. Water must be allocated and delivered volumetrically, and in a manner (i.e., flow rate, duration) that is compatible with good farming and irrigation practices.

The concept of volumetric allocation is of fundamental importance. The very nature of modern irrigation scheduling involves delivering the correct volume of water at the correct time. Accountability and billing require knowledge of water volumes. However, implementation of volumetric allocations and metering is often difficult and expensive. Water law may need to be developed or changed in a country before volumetric policies are implemented.

From an engineering standpoint, volumetric billing is easiest if the project authority delivers water to individual fields. In practice, a delivery point may supply water to 3-100 users. A first step is to increase the number of delivery points, so that less cooperation is needed between users, and so that it is easier to measure the volume used by individual users.

Volumetric billing also requires either a volumetric flow meter (impractical for most small delivery points due to the high cost and maintenance), or a flow rate meter combined with a controlled and constant delivery flow rate (the volume can be computed by knowing the constant flow rate and delivery time). Most water delivery systems need significant modernization in order to be able to provide controlled and constant flow rate (a flow rate which does not change with time during the delivery).

In short, the shift to flexible deliveries with volumetric billing will require a significant investment in the water delivery infrastructure, including major attention to personnel training, redesign of structures, and long term maintenance.

Only after points 1-3, as listed above, have been adequately addressed will on-farm demonstrations and discussions of on-farm irrigation scheduling be meaningful to most farmers.

Previous Page Top of Page Next Page