In this study, a new methodology is presented for simultaneous agricultural water and return flow (waste load) allocation in rivers. In this methodology, an objective function based on Conditional Value at Risk (CVaR) and a Nonlinear Interval Number Programming (NINP) technique are utilized. The CVaR can handle uncertainties in the form of probability distributions, while NINP incorporates uncertain inputs which are only available as intervals. This CVaR-NINP framework is used for agricultural water and return flow allocation planning under uncertainty. In this paper, to reduce the amount of saline return flow discharged into the river, a part of return flow of each agricultural network is diverted to an evaporation pond. Some meta-models based on Artificial Neural Network (ANN) are trained and validated using the results of Soil, Water, Atmosphere and Plant (SWAP) simulation model to reliably approximate the quantity and Total Dissolved Solids (TDS) load of agricultural return flows in a critical 7-day period. The effectiveness of the proposed methodology is examined through applying it to a part of Karkheh River catchment in the southwestern part of Iran. The results confirm the applicability of the model in incorporating the main uncertainties and generating water and waste load allocation policies in the form of interval numbers.
ASJC Scopus subject areas