干旱区生态环境知识资源中心

Uncertainty assessment of groundwater modeling is important for sustainable groundwater management and planning. The purpose of this study is to assess parameter uncertainty of groundwater modeling in the Birjand plain, Iran. This arid aquifer was modeled using MATLAB-based MODFLOW to avoid propagating uncertainty associated with hydraulic conductivity and recharge parameters. So, the aquifer was divided into 17 hydraulic conductivity homogenous zones; besides, 9 recharge zones were considered separately. Parameter uncertainty was evaluated using the Monte Carlo (MC) sampling technique, namely, the generalized likelihood uncertainty estimation (GLUE). The results indicated that the performance of the GLUE based on the inverse error variance likelihood function was satisfied, because it gave the higher bracketing of observations equal to 86 %. Parameter uncertainty is well defined in the zones where they are not influenced directly by an inflow or outflow stream while hydraulic conductivity parameters of these zones follow approximately a normal distribution. In addition, groundwater modeling leads to a uniform exponential distribution in the zones with inflow or outflow streams.