Groundwater and surface-water interaction within the upper Smith River Watershed, Montana 2006-2010

Abstract

The 125-mile long Smith River, a tributary of the Missouri River, is highly valued as an agricultural resource and for its many recreational uses. During a drought starting in about 1999, streamflow was insufficient to meet all of the irrigation demands, much less maintain streamflow needed for boating and viable fish habitat. In 2006, the U.S. Geological Survey, in cooperation with the Meagher County Conservation District, initiated a multi-year hydrologic investigation of the Smith River watershed. This investigation was designed to increase understanding of the water resources of the upper Smith River watershed and develop a detailed description of groundwater and surface-water interactions. A combination of methods, including miscellaneous and continuous groundwater-level, stream-stage, water-temperature, and streamflow monitoring was used to assess the hydrologic system and the spatial and temporal variability of groundwater and surface-water interactions. Collectively, data are in agreement and show: (1) the hydraulic connectedness of groundwater and surface water, (2) the presence of both losing and gaining stream reaches, (3) dynamic changes in direction and magnitude of water flow between the stream and groundwater with time, (4) the effects of local flood irrigation on groundwater levels and gradients in the watershed, and (5) evidence and timing of irrigation return flows to area streams. Groundwater flow within the alluvium and older (Tertiary) basin-fill sediments generally followed land-surface topography from the uplands to the axis of alluvial valleys of the Smith River and its tributaries. Groundwater levels were typically highest in the monitoring wells located within and adjacent to streams in late spring or early summer, likely affected by recharge from snowmelt and local precipitation, leakage from losing streams and canals, and recharge from local flood irrigation. The effects of flood irrigation resulted in increased hydraulic gradients (increased groundwater levels relative to stream stage) or even reversed gradient direction at several monitoring sites coincident with the onset of nearby flood irrigation. Groundwater-level declines in mid-summer were due to groundwater withdrawals and reduced recharge from decreased precipitation, increased evapotranspiration, and reduced leakage in some area streams during periods of low flow. Groundwater levels typically rebounded in late summer, a result of decreased evapotranspiration, decreased groundwater use for irrigation, increased flow in losing streams, and the onset of late-season flood irrigation at some sites. The effect of groundwater and surface-water interactions is most apparent along the North and South Forks of the Smith River where the magnitude of streamflow losses and gains can be greater than the magnitude of flow within the stream. Net gains consistently occurred over the lower 15 miles of the South Fork Smith River. A monitoring site near the mouth of the South Fork Smith River gained (flow from the groundwater to the stream) during all seasons, with head gradients towards the stream. Two upstream sites on the South Fork Smith River exhibited variable conditions that ranged from gaining during the spring, losing (flowing from the stream to the groundwater) during most of the summer as groundwater levels declined, and then approached or returned to gaining conditions in late summer. Parts of the South Fork Smith River became dry during periods of losing conditions, thus classifying this tributary as intermittent. The North Fork Smith River is highly managed at times through reservoir releases. The North Fork Smith River was perennial throughout the study period although irrigation diversions removed a large percentage of streamflow at times and losing conditions persisted along a lower reach. The lowermost reach of the North Fork Smith River near its mouth transitioned from a losing reach to a gaining reach throughout the study period. Groundwater and surface-water interactions occur downstream from the confluence of the North and South Fork Smith Rivers, but are less discernible compared to the overall magnitude of the main-stem streamflow. The Smith River was perennial throughout the study. Monitoring sites along the Smith River generally displayed small head gradients between the stream and the groundwater, while one site consistently showed strongly gaining conditions. Synoptic streamflow measurements during periods of limited irrigation diversion in 2007 and 2008 consistently showed gains over the upper 41.4 river miles of the main stem Smith River where net gains ranged from 13.0 to 28.9 cubic feet per second. Continuous streamflow data indicated net groundwater discharge and small-scale tributary inflow contributions of around 25 cubic feet per second along the upper 10-mile reach of the Smith River for most of the 2010 record. A period of intense irrigation withdrawal during the last two weeks in May was followed by a period (early June 2010 to mid-July 2010) with the largest net increase (an average of 71.1 cubic feet per second) in streamflow along this reach of the Smith River. This observation is likely due to increased groundwater discharge to the Smith River resulting from irrigation return flow. By late July, the apparent effects of return flows receded, and the net increase in streamflow returned to about 25 cubic feet per second. Two-dimensional heat and solute transport VS2DH models representing selected stream cross sections were used to constrain the hydraulic properties of the Quaternary alluvium and estimate temporal water-flux values through model boundaries. Hydraulic conductivity of the Quaternary alluvium of the modeled sections ranged from 3x10-6 to 4x10-5 feet per second. The models showed reasonable approximations of the streambed and shallow aquifer environment, and the dynamic changes in water flux between the stream and the groundwater through different model boundaries.