Over abstraction for agricultural production and droughts in the Bakhtegan basin in southwest Iran has led to decreased lake volumes and has even dried in some years. These problems have occurred in the basin as a result of neglecting the roles that humans have played in lake desiccation. This study developed a socio-hydrological model to understand such dynamic interactions between the economy and community's sensitivity to the environment in the period of 1999 to 2013. The WEAP model was used to simulate the hydrological system that is (bi-directionally) coupled to a society model simulating community sensitivity. A multi-objective optimization algorithm was used to estimate the parameters of the coupled model. The results of the model calibration and validation in estimating the simulated discharges show good performance of the model in simulating observed streamflows (R² and NSE ranging between 0.73 and 0.99 and 0.31 to 0.99 respectively), agriculture water supplied (R² and NSE ranging between 0.00 and 0.69 and −7.08 to 0.58 respectively), migration and population (R² and NSE for migration: 0.29 and 0.63). The results of the integrated model on community response showed that attention to environmental conditions, such as lake levels, was heightened right after the drought events of 2000–2001 and 2008–2010. The effect of the latter event was even accentuated by several human interventions such as operationalization of Sivand and Mollasadra dams and transfer of pristine lands to agriculture. This led to major contraction of the lake. The model interprets that heightened community sensitivity as a result led to public policy dialogue and change. Water was taken away from agriculture and given to the environment leading to temporary lake revival. This was also confirmed by reviewing government decisions in 2011, further validating the value of socio-hydrological models in interpreting feedbacks between community sensitivity and environmental flows.