Water allocation to different users implies a trade-off between the benefits perceived by different sectors and environmental demands. This conflict, considering the income of different sectors and environmental issues, is very crucial, especially in reservoir operation. In this study, a multi-objective optimization algorithm combined with two simulation models (SWAT and WEAP) has been developed to consider the tradeoff between agricultural benefit and environmental water demand through parallel processing. In this model, decision variables are monthly agricultural demands which are considered as inputs into the Water Evaluation And Planning (WEAP) model. Then, the production of different crops and discharge to the downstream lake are obtained based on the allocated water to each sector using the Soil and Water Assessment Tool (SWAT) model. After the simulation, the objective functions are calculated based on agricultural income and environmental flow discharged to the downstream lake. To evaluate the performance of the proposed model, Mahabad reservoir, in Mahabad river basin (located at the south of the Lake Urmia and northwestern of Iran), was studied during a drought period. This reservoir supplies domestic, industry, agriculture water of the Mahabad plain and downstream environmental demand which is discharged to the Lake Urmia. Non-dominated Sorting Differential Evolution (NSDE) algorithm is used to find the optimal water allocation to different sectors and efficient agricultural irrigation patterns of different crops (winter wheat, alfalfa, apple, and sugar beet). Results showed the optimal operation of the reservoir through irrigation management could improve water stress indicator from 80% in traditional irrigation pattern to 60%, only by 6.5% decrease in agricultural income. Also, the inflow to Urmia Lake was increased from 173 MCM to 354 MCM during the drought period (5 years).