Optimal irrigation system selection: A multiperiod quadratic programming approach

Abstract

This study reports on the optimal temporal selection of irrigation technologies (graded and level furrows, level basin and drip) that maximize net revenues of a farm unit with different soil types. Unique to this study was the inclusion of irrigation water management strategies that allowed unbiased comparison of the different technologies; for each surface and pressurized system of the study simulation models were used to control the governing parameters that determine the distribution and effectiveness of the applied water, infiltrate equal seasonal water depth in the root zone and adequately irrigate a similar fraction of total cropland area. The ultimate effects on crop yield were assessed by linking the areal water distribution depths with specific crop-water production functions. A generalized mean variance model was constructed to simulate the selected production activities of the farm; to integrate the water delivery irrigation systems, crop mix and soil endowments, and to include the alternative irrigation water application strategies. Adjustments to stochastic water supplies and revenues were also included into the model as well as regulatory constraints associated with the quantity of water allocated over the growing season and risk preference levels. Given the model assumptions, the results indicate that: (1) the generalized mean variance model was successful in establishing an intertemporal path for irrigation systems adoption; (2) high uniformity level basins were the technology of choice in all soils, substituting the low uniformity graded furrows; (3) with yield increases realized, drip systems are likely to be first adopted in soils of high intake rates; (4) increasing levels of risk aversion changed the area-technology mix and added to reductions in the areas under production; area reductions under level basin took place in high intake soils, suggesting that adoption of level basin in low intake rate soils provide higher level of farm security and are more likely to be given priority in the farm enterprise planning by growers defined by the model as risk averse; (5) the farm structure is considerably dependent upon cotton for the production of net revenue; wheat contributions to net revenue were relatively small due to the economic inability of the crop to command scarce resources and generate profits.