 |
|
Executive Summary The Kansas Water Office (KWO) contracted with the Kansas Geological Survey (KGS) in the spring of 2007 to develop a numerical ground-water model as a component of a larger, comprehensive review of the regional water supply in the Smoky Hill River basin. The objective of the model is to better understand the stream-aquifer interactions by simulating streamflow in the Smoky Hill River and ground-water levels in the surrounding alluvial deposits downstream of Kanopolis Reservoir. In addition, the model will be used to simulate climatic, streamflow, and pumping conditions and their effects on the surface- and ground-water supplies. The model was developed with input from a stakeholder advisory committee consisting of individuals from the KWO, the Kansas Department of Agriculture’s Division of Water Resources, the City of Salina, water users in the Smoky Hill River valley, and other interest groups. Since 1948, when construction on Kanopolis Reservoir was completed, the downstream reach of the Smoky Hill River and the hydrologically connected alluvial aquifer have seen extensive surface- and ground-water-right development, primarily for irrigation and municipal demands. The City of Salina, located just west of the confluence of the Smoky Hill and Saline rivers, owns the largest and some of the most senior water rights downstream of the reservoir. The regional water supply is sensitive to periods of extensive drought and extreme flooding conditions, both of which have taken place in recent years. As recently as July 2006, extended periods of below-normal precipitation and resultant low streamflow in the Smoky Hill River prompted the City of Salina to seek protection of their water rights. The climatic conditions also affected reservoir levels in Kanopolis. Operating levels were far below normal in 2006, only to fill to flood-pool elevations in a matter of months during the spring of 2007. The KGS model is an adaptation of MODFLOW, a widely-used software program for modeling ground-water flow and stream-aquifer interactions developed by the U.S. Geological Survey (USGS). The study area includes the Smoky Hill River and the hydrologically connected alluvial aquifer from the Kanopolis outlet tubes to the confluence of the Smoky Hill River with the Saline River just east of Salina. The model is subdivided into 11,484 rectangular grid cells, each 0.25-by-0.25-mile in size. The model was calibrated to represent both a steady-state condition (predevelopment period of 1944 to 1947) and a transient condition (predevelopment period to the end of 2006). The transient portion models flow conditions that change over time and uses six-month time steps centered on the “growing” season (April to September) and “winter” season (October to March). The ancestral channel of the Smoky Hill River is very different from its present channel location. In geologic history, the river used to flow south between present-day Marquette and Lindsborg into the Arkansas River. Geologic evidence indicates that headward erosion from a tributary to the Kansas River captured or cut off the Smoky Hill River, resulting in its present northward flow towards Salina. Geologic review of well logs indicates that the Arkansas River paleochannel hydrologically connects the Smoky Hill alluvium and the Equus Beds portion of the High Plains aquifer to the south. However, the connection appears to be small and insignificant in terms of water movement between the two aquifer units. The lithologic review also shows much of the area contains a confining layer of low-permeability material (silts and clays) sitting on top of the permeable alluvial aquifer sediments (sands and gravels). This indicates that stream-aquifer interactions, although present and significant in practical relevance to this model, may not be as strong as in other typical alluvial systems. In addition, the confining layer causes the aquifer to behave as a confined aquifer in some places. The model was calibrated to match area-wide water levels, selected well hydrographs, and mean six-month streamflows at the Lindsborg and Mentor gaging stations on the Smoky Hill River. Steady-state or predevelopment results show that the river was primarily a gaining stream throughout the model area. The transient phase of the model indicates on average from 1962 to 2006 the river generally gained water from the aquifer except in the Salina area where the river begins to lose water to the aquifer. The river resumes its gaining state just east of the city extent to its exit from the model area. No wells in the model area contain long-term depth-to-water records over the entire transient period. To calibrate the model to water-levels, well groups (wells located within proximity to each other and whose combined measurement histories provide an extended timeline) were selected throughout the valley and in the Equus Beds aquifer. The model-simulated water levels compare very favorably with observed water-levels within the core areas of the alluvial aquifer (near the river and relatively higher saturated thicknesses) and less so in areas of little ground-water development near the alluvial aquifer edges or in the upland terrace deposits where the permeable sediments are thin. Long-term recharge to the aquifer system is estimated to be in the 1 to 1.5 inch range. During the transient phase of the model, recharge is computed based on the total precipitation in each stress period. The transient portion of the model indicates there is an overall slight decline in net storage of the aquifer system from the predevelopment period (1944 to 1947) to 2006 and an increasing trend in ground-water development. Two scenarios were requested by the KWO and simulated with the calibrated transient model, both of which were based on a non-traditional but innovative management approach. The scenarios were based on a “back calculation” approach of repeating climatic conditions from 1948 to 2006 with present day water demands. Under the first scenario, primary surface-water inflows to the model (e.g., releases from Kanopolis) were systematically adjusted each year until a target low-flow rate (20 ft 3/sec, provided by the KWO) for the Smoky Hill River was achieved near Salina. Out of the 118 six-month time steps in the model, 63 required some amount of flow to be released from Kanopolis Reservoir in order to achieve the target flow near Salina while maintaining the 2006 level of water right development. The second scenario simulated the flow rate near Salina again by repeating climatic conditions from 1948 to 2006 with no surface-water inflows (e.g., no water being released from Kanopolis) and no ground- or surface-water pumping. In the second scenario, the target flow of 20 ft 3/sec near Salina was not achieved in 30 of the 118 model time steps. |
|