Recent developments in instream flow methodology have shifted from minimum flow regulation to the regime-based approach, recognizing that natural streamflow variation is an essential component in sustaining aquatic ecosystem biodiversity. However, when restoring streamflow variability within a water resources management project, it is difficult to precisely connect a specific species or community to its flow variability requirements. One alternate approach is to potentially promote aquatic biodiversity by increasing both the spatial and temporal habitat diversity of a river system. In this paper, a two dimensional numerical model is used to simulate depth and velocity distributions for multiple streamflow conditions at a stream section downstream of a reservoir. The model is based on depth-averaged theory and is specific to an alluvial river with a movable bed. Fuzzy concepts are applied to modeled pools, riffles, and other habitat units - which are defined by depth and velocity - to assign approximately 290,000 mesh areas of the numerical model under different streamflow conditions. Both spatial and temporal habitat variation is evaluated to indicate which set of streamflow conditions maintains the more diverse habitat objective. This allows downstream flow-based habitat diversity to become an ecological objective of reservoir operation.