Hydrology [H]

H13B MCC:Level 2 Monday

Stream Temperature, Low Flows, and Surface-Subsurface Interactions I Posters

Presiding: C Luce, USDA Forest Service; D Moore, University of British Columbia


Data Collection Methodology For Dynamic Temperature Modeling, Testing, and Corroboration

* Neilson, B (bethany.neilson@usu.edu) , Utah Water Research Laboratory, Utah State University, Logan, UT 84322
Bandaragoda, C (cmay@cc.usu.edu) , Utah Water Research Laboratory, Utah State University, Logan, UT 84322

Model testing and corroboration is becoming more of a concern as both hydrologic and water quality models have become critical in managing water quality and ensuring that instream water quality standards are met. During calibration many modelers use available data and change model parameters and inputs without an understanding of whether calibrated parameters or adjusted inputs are realistic or representative of the system. In a preliminary temperature model application on the Virgin River in southwestern Utah, USA, for a low flow scenario, a decrease in the solar radiation inputs by 15-20% during calibration was required to match observations. This suggested that important components of the energy balance may be missing when modeling the system with a typical surface flux temperature model. Possible missing energy flux components were identified as bed conduction, substrate warming due to solar radiation penetration, and transient storage effects. The identification and relative importance of each of these missing energy balance components required a well-designed and comprehensive data collection effort. This presentation will detail the data collection methodology used in the Virgin River, the reasoning behind each type of data collected, and the relative importance of each data type in further model development, testing, and parameterization.


Application of a process-based, basin-scale stream temperature model to cumulative watershed effects issues: limitations of Forest Practice Rules

* Allen, D M (allen@eps.berkeley.edu) , Department of Earth and Planetary Science, U.C. Berkeley, McCone Hall 3rd Floor, Berkeley, CA 94720 United States
* Allen, D M (allen@eps.berkeley.edu) , Stillwater Sciences, 2855 Telegraph Avenue, Ste. 400, Berkeley, CA 94705 United States
Dietrich, W E (bill@eps.berkeley.edu) , Department of Earth and Planetary Science, U.C. Berkeley, McCone Hall 3rd Floor, Berkeley, CA 94720 United States

Elevated stream temperatures increase the metabolic demands on fish, and at sufficiently high levels, can be lethal. Timber harvesting increases stream temperature by removing dense riparian canopy and exposing the channel to direct shortwave radiation. We have developed a simple, process-based stream temperature model, {\it BasinTemp}, which predicts water temperatures for very short reaches for entire channel networks. Spatially explicit radiation predictions are generated in a GIS, while heat and mass transfer processes are modeled using a simple 1D steady-state energy balance model. The model assumes that direct shortwave radiation is the primary control on water temperatures during summer months and that topography and most importantly, riparian vegetation, moderate the amount of direct solar radiation received at the stream surface. An optimization routine uses measured thermograph data from the basin of interest to improve model predictions. This feature eliminates the need for substantial site-specific field data to calibrate the model. The model has been tested on several basins in Northern California and Southern Oregon. Model performance has generally be excellent. For example, an application to Bull Creek, a 110km2 basin in the South Fork Eel River basin in Northern California, generated an RMSE of 0.25 degrees C and an R2 of 0.99 for observed versus predicted temperatures. An important feature incorporated into the model is the capability to adjust riparian tree heights and to examine the resulting local and downstream cumulative temperature effects. We show results where we alter riparian vegetation through a range of plausible shade scenarios on low order, headwater channels. These channels, which comprise most of the stream network, typically receive little if any protection from state Forest Practice Rules. Using a channel slope-based classification system appropriate for freshwater coho salmon lifestages, we demonstrate that available habitat downstream in the basin is strongly dependent on the amount of riparian shade protection provided to upstream channels.


Spatially distributed estimates of riparian stream shading from remote sensing: effects of disturbance and relationship to stream temperature

* Luce, C H (cluce@fs.fed.us) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States
Gutierrez, B (bargute@yahoo.es) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States
Nagel, D (dnagel@fs.fed.us) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States
Dunham, J (jdunham@usgs.gov) , USGS FRESC, 3200 SW Jefferson Way, Corvallis, OR 97331

Solar radiation has long been recognized as a major component of the energy budget of streams, and modeling of stream temperature across stream basins requires estimates of riparian stream shade over extensive areas. A variety of methods are available for measuring shade locally, including hemispherical photography, however these are impractical for distributed stream temperature modeling. Methods for estimating spatially distributed shade are rarer. Algorithms requiring estimates of tree heights have been used, but estimates of tree heights distributed across stream basins are often inaccurate. We explored the use of remote sensing data to more directly estimate stream shade by developing relationships between geographically registered hemispherical photography and satellite imagery. Heterogeneity in vegetation families yielded poor relationships between shade and NDVI. Classification into categories of open/grass, shrub communities, and conifer communities, was more successful in discriminating shade measurements. Shade from riparian vegetation is affected by a variety of natural and anthropogenic disturbances, and the Boise river basin has experienced several wildfires and post-fire debris flows, providing an opportunity to validate the shade estimates against observations of disturbance. We further evaluated the shade estimates against stream temperature data from 10 streams with varying levels of disturbance. The strong correlations suggest the possibility of substantially improving empirical models of stream temperature using remote sensing data. The methods also hold promise for use in distributed physically based stream temperature models.


Wildfire, channel disturbance, and stream temperature: spatio-temporal patterns and associations with the distribution of fish and amphibians in central Idaho

Dunham, J (jdunham@usgs.gov) , USGS FRESC, 3200 SW Jefferson Way, Corvallis, OR 97331 United States
* Luce, C (cluce@fs.fed.us) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States
Rosenberger, A (arosenberger@fs.fed.us) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States
Rieman, B (brieman@fs.fed.us) , USDA Forest Service, Rocky Mountain Research Station, 322 E Front St., Suite 401, Boise, ID 83702 United States

Temperature is a critical factor in stream ecosystems, and one that is altered by wildfire and related channel disturbances. In central Idaho streams, temperatures after wildfires may increase following loss of shade from riparian vegetation, and changes in channel structure that increase exposure to solar radiation and decreased hyporheic exchanges. To examine the spatio-temporal aspects temperature in relation to these influences, we employed three approaches: a long-term pre-post fire comparison of temperatures between a pair of streams, one burned and one unburned; a short-term pre-post fire comparison of a burned and unburned stream with spatially extensive data; and a short-term comparative study of spatial variability in temperatures using a space for time substitutive design. These three approaches provided key insights into the value of each study approach and revealed some expected and some surprising associations between temperature and occurrence of native trout and tailed frogs. The results of this work highlight the importance of spatio-temporal variability in study designs to quantify the effects of wildfire and disturbance on stream temperatures, and the implications of stream temperature for aquatic species in a broad landscape context.


Regional Stream Temperature Response to a Severe Drought in Relation to Catchment Characteristics

* Moore, R ( (rdmoore@geog.ubc.ca) , Department of Geography, The University of British Columbia, 1984 West Mall, Vancouver, BC V6T 1Z2 Canada
* Moore, R ( (rdmoore@geog.ubc.ca) , Department of Forest Resources Management, The University of British Columbia, 2424 Main Mall, Vancouver, BC V6T 1Z4 Canada
Stahl, K (kstahl@geog.ubc.ca) , Department of Geography, The University of British Columbia, 1984 West Mall, Vancouver, BC V6T 1Z2 Canada

Stream temperatures generally reach their summer maxima during extended periods of hot, dry weather in association with low streamflow. However, glacier runoff tends to increase under these conditions and is hypothesized to be an effective regulator of summer stream temperature. This study draws upon synoptic surveys of continuous stream temperature records in tributaries of two mountainous drainages in British Columbia, Canada, during a severe drought period in summer 2004. Regression tree modelling revealed that the maximum weekly average temperature (MWAT) for the 2004 summer was, on average, approximately 4 deg C lower for streams draining catchments with more than 5% glacier cover, supporting the hypothesis that glaciers act as thermal regulators. For the catchments with low or no glacier cover, the model included splits based on estimated bankfull width and percentage lake cover, both of which had positive influences on MWAT. The inclusion of bankfull width is consistent with the decrease in shading by riparian forest with increasing stream width. The inclusion of lake cover could reflect the direct effects of lakes in generating high outlet stream temperatures due to summer stratification, but could also be a surrogate for other aspects of catchment morphology that influence stream warming.


The Influence of Hyporheic Flow on the Temperature of a Riffle-Step-Pool Stream

Rothwell, E (jmcnamar@boisestate.edu) , Department of Geosciences Boise State University, 1910 University Dr., Boise, ID 83725
* McNamara, J (jmcnamar@boisestate.edu) , Department of Geosciences Boise State University, 1910 University Dr., Boise, ID 83725
Luce, C (cluce@fs.fed.us) , Boise Aquatic Sciences Lab Rocky Mountain Research Station, 316 East Myrtle Street, Boise, ID 83702

The ability to predict stream temperature is important due to the many direct and indirect water use and watershed management effects on streams. One approach to evaluate stream temperature change is to construct a heat-energy budget. Failures of the energy budget approach to modeling stream temperature are often due to inadequate representation of all the processes that control stream temperature. Hyporheic flow influences stream temperature by temporarily removing water from the heating impact of solar radiation and other positive heat fluxes at the surface of the stream, heat conduction to the substrate, and mixing with groundwater, all resulting in flow returning from the substrate out of phase with stream temperature. The influence of hyporheic flow is not often included in energy budget models to predict temperature change. In this study stream temperature change is successfully modeled for a 400-meter long reach in a small headwater stream by using an energy budget temperature model that includes the often-Neglected effects of advected water through the hyporheic zone as a major component of the energy budget. Hyporheic flow was evaluated using instream tracer techniques and darcian flux calculations. The dominant energy fluxes into the stream during summer months are net radiation and sensible heat exchange with the atmosphere (61 and 36 percent respectively). The dominant heat energy flux out of the stream was from hyporheic flow, providing 36 to 75 percent of the energy sink.


Wintertime Warming of Hyporeic Water in an Exposed Gravel Bar, American River

* Silver, M H (msilver@csus.edu) , Department of Geology, California State University, Sacramento, 6000 J Street, Sacramento, CA 95819-6043 United States
Horner, T C (hornertc@csus.edu) , Department of Geology, California State University, Sacramento, 6000 J Street, Sacramento, CA 95819-6043 United States

Measurements of intergravel temperature near an exposed gravel bar along the American River near Sacramento, California show warmer intergravel water on the downstream end of the gravel bar than on the upstream end. Temperature loggers were installed at depths of 1', 2', and 4' below the gravel surface and used to collect data during the fall and winter of 2004. Surface water flows around the bar, but measurements of hydraulic head indicate a downward hydraulic gradient at the upstream site and an upward hydraulic gradient at the downstream site, suggesting that water also flows underneath the gravel bar. Temperature profiles from August 2004 suggest the same interpretation: at the upstream site, a large (~4 C) diurnal temperature fluctuation was present at 1' depth and some diurnal fluctuation was present at 4' depth, while at the downstream site, the 4' temperature did not fluctuate, and diurnal temperature fluctuation at the 1' level was dampened. In the following winter months, a similar pattern is observed at the upstream site, although with less diurnal fluctuation in water temperature. However, at the downstream site, discharging water was warmer than recorded upstream, and the thermal profile was inverted: the 4' temperature was the highest and 1' temperature the lowest. Inversion began on October 20, 2004 and continued until March 5, 2005. During this period, 1' downstream temperatures were up to 1 C warmer than those recorded at the same depth upstream and 4' temperatures were up to 3.5 C warmer at the downstream end of the gravel bar. During the period of inversion, the gravel bar appears to contribute heat to the stream, as water flowing out from underneath the bar is warmer than when it entered. This may be due to a seasonal reversal in thermal gradient, where a constant temperature at some depth beneath the stream results in heat flow from the surface downward in the summer and from the sub-surface upward in the winter. In September 2005, we installed instruments to measure temperature at two additional sites in the American River where similar temperature inversion and warming of water that has flowed under a gravel bar may occur. Data will be collected in the fall of 2005 and will be used to determine if the observed warming and temperature inversion occurs as water flows under other gravel bars along the American River.


River-Aquifer Interactions, Geologic Heterogeneity, and River Management

* Fleckenstein, J H (jan.fleckenstein@uni-bayreuth.de) , Department of Hydrology University of Bayreuth,Germany, Universitaetsstrasse 30, Bayreuth, 95447 Germany
Niswonger, R G (rniswon@usgs.gov) , US Geological Survey, Carson City, Nevada, 333 West Nye Lane, Carson City, NV 89706 United States
Fogg, G E (gefogg@ucdavis.edu) , Department of Land, Air and Water Resources, University of California, Davis, 1 Shields Avenue, Davis, CA 95616 United States

Managing rivers and their underlying aquifers for minimum flows, riparian habitat or aquifer recharge requires an understanding of the spatial patterns and temporal dynamics of river-aquifer exchange. Results are presented from investigations of the effects of geologic heterogeneity on river-aquifer exchange, minimum river flows and water availability in the riparian corridor for a typical alluvial fan system in the western USA. River-aquifer interactions were simulated on a regional-scale (~50km) and for a river reach (~2000m) using numerical codes for saturated and variably saturated flow. Geologic heterogeneity of the alluvial fan system was characterized with a geostatistical approach, based on transition probabilities and Markov Chains. Different hydrofacies models for a 50 km segment and a 2 km reach were created from sequential indicator simulations. Variably saturated flow between the river and the deep regional water table were explicitly simulated in both models. Groundwater levels, river flows, sediment saturation and temperature data were used to calibrate the models. The regional simulations showed that different spatial arrangements of hydrofacies have significant effects on minimum river flows with implications for salmon migration. Although total annual seepage volumes were relatively insensitive to geologic heterogeneity spatial and temporal variability of seepage was large between the different heterogeneous models. Local reconnections developed seasonally in some models and the period with sufficient flows for salmon fall-migration varied by up to 13 days between the models. The reach-scale simulations demonstrated that perched zones, which form between the river and the regional water table, can be important in supporting river base flows and riparian vegetation. Connected pathways between the river channel and the riparian corridor, which could be characterized with the temperature data, may sustain phreatophytes even when the regional water table is far below the river channel. These results elucidate some important effects of geologic heterogeneity on river-aquifer interactions, which could be crucial for the management of alluvial river systems.


Evaluation Of The Hydrologic Connection Between Mammoth Creek And Mammoth Community Water District Water Supply Wells, Mono County, California

* Burak, S (sburak@psln.com) , Snow Survey Associates, P.O. Box 8544, Mammoth Lakes, CA 93546 United States
Farrar, C (cdfarrar@usgs.gov) , United States Geological Survey, P.O. Box 1360, Carnelian Bay, CA 96140 United States

Groundwater-surface water interactions are examined at several locations in the Mammoth Lakes Basin in the eastern Sierra Nevada. The study sites include three streamflow gauging stations and five water supply and monitor wells located near Mammoth Creek. Major ion chemistry and stable isotope composition of groundwater and surface water were determined and analysis provides evidence that surface and groundwater at the study locations share similar isotopic signatures and seasonal differences due to snowmelt runoff. Streamflow records, well water levels and production well pumpage data were used to qualitatively assess stream-aquifer relationships and suggest that pumpage from production wells located close to Mammoth Creek influences water levels in monitor wells and Mammoth Creek discharge primarily in the fall and early winter. Previous and widely held assumptions that groundwater pumping from deep volcanic aquifers had no effect on stream flows must be revised with the results of the stable isotope analysis. Results suggest there is a connection and this information can be used manage timing and magnitude of groundwater extractions.


Coupling Between Periodic Fluctuations in Stream Water Temperature and Groundwater Elevation, Central New Mexico

* Jakubowski, R T (ryanj@nmt.edu) , New Mexico Institute of Mining and Technology, Department of Earth and Environmental Science, 801 Leroy Place, Socorro, NM 87801
Bowman, R S (bowman@nmt.edu) , New Mexico Institute of Mining and Technology, Department of Earth and Environmental Science, 801 Leroy Place, Socorro, NM 87801

Diurnal (24-hour) fluctuations in groundwater levels are often observed in riparian areas. They are generally attributed to periodic changes in barometric pressure, evapotranspirative demand, and recharge events. For losing streams located along semi-arid riparian corridors infiltration of surface water and advection of heat can strongly influence the subsurface hydrogeology. In fact, hourly head and temperature measurements in wells adjacent to the Rio Grande, New Mexico, have revealed diurnal groundwater fluctuations that correlate with diurnal changes in river temperature. We hypothesize that a periodic change in the streambed hydraulic conductivity modulated by variations in temperature may produce a transient flux (pressure wave) into the underlying shallow aquifer. The presence of a streambed restricting layer, diurnal changes in river temperature, limited riparian vegetation, and patterns in head during no-flow conditions in the Rio Grande support a scenario in which variable groundwater recharge from the river contributes to the diurnal head change in the aquifer. We model coupled heat and mass transport to evaluate the potential significance of the hypothesized hydrodynamic interactions.