OPTIMAL PLANNING AND MANAGEMENT OF IRRIGATION WATER IN A COMMAND AREA

Abstract

Irrigated agriculture is by far the largest consumptive user of water, which is also one of the most essential and costly inputs for crop production. As more water needs to be diverted from irrigation sector for human and industrial consumption with the passage of time, the water in irrigated agriculture has to be used wisely and efficiently employing scientific water and land management strategies. The scientific management of irrigation water includes planned conveyance, application, distribution and use of available irrigation water in the command area of a canal such that maximum irrigation efficiency is achieved and maximum return is obtained per unit volume of water per unit area of cropped land per unit time. Thus, there exists a pressing need to design the surface irrigation systems such that only the required amount of water is allowed to infiltrate into the soil thereby minimizing runoff and deep percolation losses and maximizing water application efficiency. Among others, the first and most important step in design of any surface irrigation system includes the calculation of advance time of water front which is used to test whether or not the maximum flow will complete the advance phase within a prescribed time. The most dominant variables that influence the surface irrigation design are the infiltration properties of the soil, which are often difficult to evaluate because of the spatial and temporal variability of soil parameters. Despite the availability of a large number of infiltration models, owing to their simplicity and yielding reasonably satisfactory results in most applications, some of the available empirical models have been quite popular and frequently used in various water resource applications world over. The wide-spread application of Kostiakov (KT) and modified Kostiakov (MKT) models in irrigation engineering is an example worth citing. Among the multitude of water resources applications of these infiltration models, their application in predicting the water front advance in border irrigation is quite important which in turn is used to design an efficient surface irrigation system. Another important aspect in the planning and management ofa water resources system includes application of optimum amount of water at the most appropriate time so that crop stress is avoided and crop yield is maximized. Furthermore, an optimal crop area allocation model can be formulated using optimization techniques for optimal utilization of land and irrigation water in a command area yielding maximum net return under the constraints of minimum water requirements and minimum food, protein and calorie demand fulfillments. From above view-point, following items were considered for analysis for optimal planning and management of irrigation water and results were detailed, which include: (i) investigation of popular KT and MKT infiltration models for possible improvement, since a robust infiltration model predicting the actual infiltration more accurately would be more effective in planning and design of water resources systems, (ii) testing the suitability and feasibility ofvarious existing and proposed improved infiltration models to provide numerical solution of Philip and Farrel (1964) equation for the derivation of water front advance-time relationship, an important aspect in design of surface irrigation systems, (iii) optimal design of a border irrigation system that optimizes irrigation application efficiency using derived waterfront advance-time relations (iv) development of an optimal irrigation scheduling criteria using site specific crop coefficient, reference evapotranspiration and allowable depletion limits for various stages of growth, incorporating daily rainfall and soil dryness coefficient that show the soil moisture depletion at the end ofeach day and compare crop water requirement thus computed with those estimated using crop coefficient approach ofAllen et al. (1998) and (v) determination ofan optimal cropping pattern that gives maximum net return satisfying certain basic needs of the population of area, requiring minimum amount of water. The techniques proposed were tested using field data of Ramial irrigation command in Orissa (India). On investigation of the popular KT and MKT infiltration models it was observed that the assumption of a constant ultimate infiltration capacity (fc) in all the available empirical model formulations is unwarranted and needs improvement. Physically, f within time to ponding (tp) equals zero. When t approaches t, f starts increasing from zero and rises in power form (cc21 p2), till the soil attains complete saturation, beyond which it approximates a constant value, i.e. a21 p2 =f. The proposed model is designated as revised modified Kostiakov (RMKT) model, which reduces to MKT model when soil is fully saturated. Thus, the former is a more generalized version of the latter. The proposed RMKT model is further generalized incorporating time to ponding, designated as revised modified Kostiakov model incorporating time to ponding (RMKTP). The proposed, along with the existing, versions were tested on 40 datasets of infiltration derived from five different soils in India and U.S.A. The former indicated an improved performance over their respective existing versions. The test results showed that on the data of all the soils other than Yolo Light Clay (YLC) and Narsinghpur Clay (NC), RMKT performed better than KT and MKT models. On YLC and NC data, RMKT performed as well as did MKT. However, RMKTP performed significantly better than others on all the soils. The sensitivity analysis showed that at 5 minutes, 013 was the most sensitive parameter, followed by 03, a2 and p2. However, at 105 minutes, P3 was the most sensitive parameter, followed by p2, a2 and 013. At 5 minutes, a 10% error in computation of a2, p2, 013 and P3 produced, respectively, errors of 0.3, 0.12, 9.69 and 3.9 % in estimation of infiltration rate (y). Similarly, at 105 minutes, a 10% error in computation of a2, p2, a.3 and p3 produced, respectively, errors of 5, 5.7, 4.9 and 5.77 % in estimation of infiltration rate. To test the suitability of KT, MKT and proposed RMKT infiltration models to provide numerical solution of Philip and Farrel (1964) equation for the derivation of water front 111 advance-time relationship, three alternate solutions of Philip and Farrel equation were suggested, which reduces to two equations on employment of KT infiltration model. On testing the solutions with field data it was observed that on one data set, RMKT infiltration model showing enhanced performance in predicting accumulated infiltration-time relation also showed enhanced performance in prediction ofadvance-time relations over those due to employment of MKT and KT infiltration models. On two data sets, MKT infiltration model yielded enhanced performance in prediction ofaccumulated infiltration-time relation and advance-time relations over those due to employment of KT model. On three datasets, all three models performed identically and showed enhanced performance over those due to Michael (1968). Using the proposed water front advance-time relationship, Michael's (2002) design procedure for border irrigation was simplified and elaborated, and it was applied to wheat, maize, pulses and groundnut crops of Ramiala irrigation command (Orissa, India), for which border irrigation system is most suited. Using the field data ofsandy loam soil ofstudy area, the optimum values ofstream size (q), border length (L) and time ofapplication (T) for those crops for maximum irrigation water application efficiency were determined. The application efficiencies for wheat, maize, pulses and groundnut of study area were 83.39, 83.07, 82.66 and 82.66, respectively. To ensure that soil moisture is never allowed to deplete below management allowable depletion (MAD) level and crop stress is completely avoided, maximizing yield, scheduling of irrigation employing water balance approach was considered, which is another important aspect of irrigation water management. Therefore, an optimal irrigation scheduling criteria using site specific soil, climatic, meteorological and crop growth conditions, incorporating daily rainfall data and soil dryness coefficient that show the soil moisture depletion atthe end ofeach day was developed using MS Excel Spread Sheet, with the provision for modification of each input IV variable according to varying site and management requirements. Using the derived optimum water application efficiency the gross irrigation requirement (GIR) for all important crops were estimated employing water balance approach and crop coefficient approach. The GIR estimated using water balance approach appeared to be more realistic than those estimated using conventional crop coefficient approach as it adjusted the irrigation requirement incorporating more input parameters. To extract maximum return per unit area of cropped land per unit time per unit volume of water, an optimal cropping pattern was developed. An area allocation model was formulated using linear programming (LP) approach to allocate the land area among various crops to maximize net return from the command area. It was subject to the availability of canal water as well as groundwater, land area limitations under different crops and in different seasons of the year, protein and calorie demand of the population of the area and minimum and maximum crop area constraints. For determining optimal cropping pattern, the GIR estimated using water balance approach was considered. To evaluate feasibility of alternate management plans, five plans were considered. Plan-1 maximized net benefit from crops meeting food, protein and calorie requirements, Plan-2 did it meeting minimum food production requirements (MPR), Plan-3 considered food protein and calorie requirement with maximum allowance of 40% excess production over MPR, Plan-4 allowed 20% deviation from existing cropping pattern and Plan-5 allowed the existing cropping pattern to continue. The crop plan 1 yielded the maximum net return (120% more beneficial economically with 10% less amount of water consumption than that due to the existing cropping pattern) and was considered to be the best plan.