Deterministic and Fuzzy Evaluation of Human and Climate Contributions in Changing Hydrologic Regime: A Case Study of the Gorganrood Watershed at Tamar River Hydrometric Station

Human and climate are two major scio-hydrologic drivers that determine hydrological regimes and patterns. In this regards, Land Use and Land Cover (LULC) changes, agricultural development, etc., on global and regional scales, hydro-climatological components have influenced these regimes. The effects of each driver on the variation of hydrological components have been assessed in different studies, but these approaches are not accurate enough at watershed-scales that experience the simultaneous impacts of climate dynamics and LULC changes. Different studies have considered both climate and human altertions in the hydrological cycle, and quantified their contributions in such basin. Results of these researches can help decision makers in water management of the pros and cons of water and land use policies. The Gorganrood watershed is an important basin in the northern part of Iran, especially from the agricultural point of view, which has considerably experienced hydrological and extreme events changes. While the consequence of each climate change and LULC changes have been assessed in the watershed, there is no study, which considers the complicated interactions of these drivers. In this paper, the authors firstly evaluated the contributions of LULC and climate change on the variation of streamflow. Secondly, the modified fuzzy arithmetic method has been used to achieve their fuzzy contributions. To this purpose, the computational period was firstly divided into two different temporal spans known as the reference and affected periods. The reference period is the first temporal span in which climate controls the hydrological responses. Then, the statistical behavior of the time-serries changes due to human activities, and the affected periods. Two hydrological models, Soil and Water Assessment Tool (SWAT) and a black box Artificial Neural Networks (ANNs), were used to simulate the streamflow in the watershed. However, the results of the hydrological models showed their general acceptable performance to simulate the recorded streamflow at Tamar hydrometric station, but the results of the conceptual model (SWAT) showed that the performance of the model in the dry season is not as good as in the wet season. In the next step, the contributions of human and climate activities were assessed via two different methods. The first method is simple differential method, which compares the projection of the calibrated model in the second period with observations in both periods. The second set of contribution rates was calculated using the climate elasticity method via recorded monthly data and implemented derivation rules. In the first method, the contribution rate of human activities is significantly higher than the rate of climate change, and the result of the second method is a reverse. Because of differences in the methods’ concepts, the calculated contributions rates are different. To assess the uncertainty grouped with the estimations, a novel approach was developed using fuzzy mathematics. The uncertain version of the contribution rates showed that in each α-cut (fuzzy uncertainty level), the contribution of human alternation (LULC change) as the most important human interventions is more significant than climate direvers. In other words, during the simulation period, the effect of LULC change on the flow was very noteworthy, while climate change had relatively less effect on the behavioral change of the flow.