Results and Discussion

Water Freedom System

Survive Global Water Shortages

Get Instant Access

Different levels of water availability, expressed as equivalent millimeters of water, were included in this study to represent a wide variety of situations, ranging from no deficit (Q = 175 mm) to severe water scarcity (Q = 25 mm). As a first approximation EVI was assessed considering a farmer that bases his allocation only on expected mean values of crop potential evapotranspiration and does not measure the real values of the variable in the current and previous periods. In that case no further revision of the allocation strategy is possible. Clearly, the value of information received at the beginning of the growing season has more value for this farmer than for a decision-maker that updates the knowledge of the system with actual measurements. Figure 9.1 presents the results of EVI disaggregated for each phase of El Niño. EVI is present in all phases of El Niño when available water for irrigation has a value between 25 and 175 mm, reflecting that an allocation using ENSO-driven climate forecasts has a potential use. The weighted average of EVI values ranges from 35 to 200 U.S. dollars depending on the magnitude of the constraint. When available water is higher than 175 mm, the total satisfaction of crop requirements is achieved (i.e. an unconstrained maximization is observed) and, irrigation is therefore the simple management of the soil water budget. On the other hand, when available water is below 25 mm, no differences in water allocation strategies are found because the decision-maker is forced to irrigate the crop with maximum marginal productivity.

It is also possible to estimate the EVI for a decision-maker that revises his allocation strategy incorporating the realization of previous and observed values of crop evapotranspiration. For this farmer it is possible to correct the wrong estimates of ETc with real values. In this case the value of information derived from future conditions of El Niño is reduced. However due to the time dependence of previous decisions and their effect on final yield, EVI is never equal to zero. This feature can be observed looking at the evolution of the objective function of the decision-maker that allocates wa-

Fig. 9.1. Expected value of information (U.S. dollars) for different levels of available water and El Niño phases in the growing period

ter using El Niño forecasts and the one that bases his decision on historical ETc values. Figure 9.2 shows this situation for a situation where available water was 55 mm. Because perfect knowledge of El Niño phases is assumed, the objective function using El Niño forecasts does not change with time. During La Niña years (Fig. 9.2a) EVI decreases as a consequence of a revised optimization. As time progresses much of the final outcome is already determined, and the corrections made in water allocation reduce the relevance of climate forecasts. For normal years (Fig. 9.2b) the situation is different, since ETc in these years is higher than the climatological average, the misinterpretation of future ETc values is carried out, and even though some adjustment can be made, EVI remains close to the initial values.

Fig. 9.2. Objective function of a decision-maker that allocates water using El Niño-driven climate forecasts (straight line) and a decision-maker that uses climatological values but updates the objective function considerng observed ETc values (line with triangles); a results for La Niña years; b results for normal years

Fig. 9.2. Objective function of a decision-maker that allocates water using El Niño-driven climate forecasts (straight line) and a decision-maker that uses climatological values but updates the objective function considerng observed ETc values (line with triangles); a results for La Niña years; b results for normal years

Given the results obtained in this work, it is concluded that: (1) In locations where ENSO signal has an effect on water demands (for instance central Chile) there is an economic potential for the use of climate forecasts on water resources allocation at the farm level, (2) The expected value of information is not a monotonic function because when water is very scarce allocation decisions are limited, (3) In some situations wrong estimations of future ETc conditions can be corrected but the time dependence nature of the system makes advisable to incorporate information as a tool for irrigation management.

The results of this work correspond to a preliminary assessment of EVI. The results strongly depend on yield model representations. In this case Jensen's model only represents yield variability as a consequence of a difference between actual and potential evapotranspiration and the same final outcome would be achieved in all El Niño phases if water was not a limiting factor. The use of more sophisticated weather driven crop simulation models, as a tool to refine yield forecasts, will probably produce better estimates of the consequences, and allow the decision-maker to capture a higher proportion of the additional economic benefits associated to climate information.

Was this article helpful?

0 0

Post a comment