River-aquifer interactions, geologic heterogeneity, and low-flow management

River-aquifer interactions are crucial for understanding low-flow dynamics, especially in regions where ecological health depends on seasonal water availability. Prior research has shown that aquifer properties influence seepage rates, but the role of geologic heterogeneity remains unclear. This uncertainty drives the need for detailed modeling approaches. Existing studies often assume homogeneous aquifer conditions, which may not reflect real-world complexity. The Cosumnes River in California exemplifies the ecological consequences of declining flows, particularly for salmon migration. However, no prior work has resolved how aquifer heterogeneity affects localized seepage patterns. This gap motivates the use of numerical simulations to explore the influence of textural heterogeneity on river-aquifer exchange. The study aims to bridge the knowledge gap between aquifer structure and river flow dynamics.

The study aimed to assess how textural heterogeneity in alluvial systems affects river seepage and low flows. Declining fall flows in the Cosumnes River have raised concerns about Chinook salmon migration. The researchers sought to determine whether aquifer heterogeneity influences seepage distribution and river connectivity. A groundwater-surface water model was developed to simulate different heterogeneity scenarios. The test case involved the lower Cosumnes River basin in California. The goal was to compare seepage patterns across six heterogeneous models and one homogeneous model. The study focused on spatial variability in seepage and its impact on seasonal river openness. By analyzing model outputs, the researchers aimed to identify how localized reconnections could reduce seepage losses.

The researchers used numerical simulations to model river-aquifer interactions in the Cosumnes River basin. A groundwater-surface water model was developed for the lower river basin. Detailed geostatistical simulations captured aquifer heterogeneity. Six heterogeneous models and one homogeneous model were run for a 3-year period. Each model incorporated different realizations of textural heterogeneity. The simulations tracked annual seepage, water table configuration, and river-aquifer connectivity. The models varied in how they represented spatial distribution of aquifer materials. The study compared seepage patterns and low-flow outcomes across the different scenarios.

Net annual seepage from the river was similar across all models. However, seepage distribution along the channel varied significantly between heterogeneous and homogeneous models. Water table configurations and local river-aquifer connections showed strong differences among the models. Heterogeneous models suggested that local reconnections could reduce seepage losses. The percentage of river channel responsible for 50% of seepage ranged from 10% to 26% in heterogeneous models. In contrast, the homogeneous model showed 23% of the channel responsible for the same seepage. Differences in seepage led to up to 13 days of variation in river openness during fall months. These findings highlight the importance of aquifer heterogeneity in managing low flows.

The study concludes that aquifer heterogeneity significantly influences river seepage patterns. Heterogeneous models suggest that localized reconnections can reduce seepage losses even with a low regional water table. The spatial distribution of seepage is more variable in heterogeneous models compared to homogeneous ones. The findings imply that aquifer structure affects river openness during critical migration periods. The results support the idea that detailed geostatistical modeling is essential for accurate low-flow predictions. The study does not propose new management strategies but highlights the need for considering heterogeneity in river-aquifer interactions. The authors suggest that localized interventions may improve river flow conditions for salmon. The study emphasizes the importance of incorporating aquifer heterogeneity in water resource models.