Abstract

Five categories of data are analyzed to enhance understanding of river-aquifer exchanges—the processes by which water moves between stream channels and the adjacent groundwater system—in the Yakima River basin. The five datasets include (1) results of chemical analyses of water for tritium (³H, a radioactive isotope of hydrogen) and the ratios of the stable isotopes of hydrogen (²H/¹H) and oxygen (¹⁸O/¹⁶O), (2) series of stream discharge measurements within specified reaches (seepage investigations or “runs”), (3) vertical hydraulic gradients (between stream stage and hydraulic heads the underlying aquifer) measured using mini-piezometers, (4) groundwater levels and water temperature in shallow wells near stream channels, and (5) thermal profiles (continuous records of water temperature along river reaches). Exchanges are described in terms of streamflow, vertical hydraulic gradients, groundwater temperature and levels, and streamflow temperature, and where appropriate, the exchanges are discussed in terms of their relevance to and influence on salmonid habitat.

The isotope data shows that the ultimate source of surface and groundwater is meteoric water derived from atmospheric precipitation. Water from deep wells has a different isotopic composition than either shallow groundwater or surface water, indicating that the deep groundwater system contributes, at most, only a small component of the surface-water discharge. The isotope data confirms that river-aquifer exchanges involve primarily modern streamflow and modern, shallow groundwater.

Net exchanges of water for 46 stream sections investigated with seepage runs ranged from nearly zero to 1,071 ft³/s for 28 gaining sections, and -3 to -242 ft³/s for 18 losing sections. The magnitude of the upper 50 percent of the net gains is an order of magnitude larger than those for net losses. The sections have a normalized net exchange (as absolute value) that fully ranged from near 0 to 65.6 (ft³/s)/mi. Gaining-section values ranged from about 0.1 to 65.6 (ft³/s)/mi, and losing section values ranged from about -0.1 to -35.4 (ft³/s)/mi. Gains are much more vigorous than the losses with 55 percent being larger than 3.0 (ft³/s)/ mi, whereas, only 6 percent of the negative net exchanges were larger than 3.0 (ft³/s)/mi. Gains and losses for 167 measured reaches within the 46 sections ranged from about 70 to -75 (ft³/s)/mi, and ranged more than 5 orders of magnitude. The median values for the gains and losses were 5.1 and -4.4 (ft³/s)/mi, respectively. The magnitude of the gains was larger than the losses; more than 40 percent of the gains were greater than 10 (ft³/s)/mi, and only about 25 percent of the losses were greater than 10 (ft³/s)/mi. Reaches with large gains are identified and these reaches represent potentially important areas for various life stages of salmonids and possibly for preservation or restoration of that habitat.

Ninety-nine measurements of vertical hydraulic gradients (VHGs) were made using mini-piezometers. The median for the measurements was -0.35 ft/ft (negative values indicate downward flow), and in terms of absolute values, the median was 0.05 ft/ft. The VHGs tended to be small. Seventy VHG values were negative (indicating streamflow losses), and 29 were positive (indicating streamflow gains). VHGs vary more than 4 orders of magnitude, and in terms of magnitudes, 65 percent were less than 0.1 ft/ft. The negative VHG values are not only more prevalent but are larger than the positive values. The magnitudes of almost 50 percent of the negative VHGs are greater than 0.05 ft/ft and only 33 percent of the positive VHGs are greater than 0.05 ft/ft. The percentile distribution of the VHG data, which is similar to the shape of the seepage data distribution, shows that beyond the 80th percentile, the positive values become much larger, indicating that the largest VHGs have a different controlling mechanism. The VHGs were formulated in terms of fluxes per unit area and the negative VHGs ranged from 0.005 to 24 in/d and 96 percent are less than 3 in/d. These fluxes are determined to be “reasonable,” and river losses could support such values. Fluxes per unit area for the 29 positive VHGs ranged from 0.01 to 19.3 in/d, and 86 percent are less than 2.3 in/d. Formulating the values in terms of normalized discharge [(ft³/s)/mi] and comparing these values to the seepage data shows that the very large positive VHGs are not the controlling factor for exchanges and that other mechanisms, such as lateral inflow (groundwater discharge is not vertical), dominate the hydrologic exchange process.

Data from the near bank and flood plain monitoring sites display highly variable characteristics that reflect complex relations between groundwater levels and temperature, and water quality of the shallow system and streamflow, surface-water bodies, the flow in alluvial aquifers, and irrigation. In many cases, groundwater levels mimic river stage at both gaining and losing sites and show the effects of river-stage pressure on the shallow groundwater flow system. These effects may raise groundwater levels to the extent they intercept the land surface in depressions and sloughs. Groundwater temperature thermographs can be clearly delineated by the magnitude of the annual amplitude as to whether they are surface-water or groundwater dominated. Amplitudes are as large as 16°C and as small as 1°C, and depending on the physical setting and type of climatic year, the annual maximum temperature of groundwater lagged that of streamflow by less than a month to more than two months. At sites with streamflow losses, temperature effects in shallow groundwater are attenuated in as little as 50 ft from the river. The temperature data show that bank storage is not as important as wetting-up side channel and sloughs for supplying cool water to the shallow groundwater system. The magnitude of streamflow is an important control on exchanges, with rain-on-snow runoff events being more important than the seasonal spring runoff because the former can produce higher discharges than the latter. Groundwater levels and temperatures differ distinctively between wet and dry years, and the differences show the importance of the type of year on exchanges throughout the Yakima River basin. The increased releases from the Naches River arm reservoirs beginning in September resulted in detectable changes in both groundwater temperature and levels downstream of the reservoirs. Vertical variations of water levels, temperature, and water quality in the shallow system occur over distances as small as 10 ft.

The longitudinal temperature gradient of the water in 16 reaches within some 160 river miles were recorded in thermal profiles. Reaches ranged in length from about 5 to 14 mi, and stream gradient ranged from 0.0002 to 0.0055 ft/ft. The profiles exhibit inter- and intra-profile variations that integrate the factors controlling the temperature of a parcel of water as it moves downstream. Such longitudinal variations previously had not been documented in a riverine system. Thermal gradients range from as small as 0.00002 to as large as 0.004°C per mile per minute, and unexpectedly, the smaller gradients are not confined to the upper parts of the basin. Effects of river-aquifer exchanges and surface-water inflows are clearly displayed in the profiles. The profiles document the riverine systems’ temperature templet or longitudinal (environmental) gradient that defines a physical habitat templet, which provides for the overall biological community templet, including the different life stages and life history patterns of salmonids. The templet leads to a logical progression of the longitudinal gradient of fish assemblages, and invertebrate and algal community structure. The longitudinal gradient, overlaid with the distribution of temperature patches, compose a continuum from the headwaters to the mouth, along which habitat, and thus, species, are arranged.

First posted April 7, 2011