HR: 0830h
AN: H21D-0874    [PDF]
TI: Estimation of Streambed Seepage using Time-Series Analysis of Heat as a Tracer in Three California Coastal Rivers
AU: * Hatch, C E
EM: chatch@es.ucsc.edu
AF: UCSC, 1156 High Street, Santa Cruz, CA 95064 United States
AU: Fisher, A T
EM: afisher@es.ucsc.edu
AF: UCSC, 1156 High Street, Santa Cruz, CA 95064 United States
AU: Revenaugh, J
EM: justinr@tc.umn.edu
AF: U. of Minnesota, 310 Pillsbury Drive SE, Minneapolis, MN 55455-0219 United States
AU: Constantz, J
EM: jconstan@usgs.gov
AF: USGS, 345 Middlefield Road, Menlo Park, CA 94025 United States
AU: Ruehl, C
EM: cruehl@es.ucsc.edu
AF: UCSC, 1156 High Street, Santa Cruz, CA 95064 United States
AU: Los Huertos, M
EM: marcos@ucsc.edu
AF: UCSC, 1156 High Street, Santa Cruz, CA 95064 United States
AU: Shennan, C
EM: cshennan@ucsc.edu
AF: UCSC, 1156 High Street, Santa Cruz, CA 95064 United States
AB: Heat has been used as a natural tracer for decades, generally using forward models to simulate thermal conditions within the stream and streambed. We are developing a new method for analysis of streambed thermal records from multiple depths to estimate seepage rates. This approach differs from those developed previously in that we use models of heat transport through the stream bed to derive "type curves" of time-series behavior, then interpret long- term records to estimate seepage rates throughout the water year. Daily variations in stream temperature of several degrees or more are common in many river systems. These thermal perturbations propagate downward into the streambed with time. Streambed sediments act as a filter on these perturbations, reducing the amplitude of temperature variations with greater depth, and causing the peaks in temperature to be shifted in time. These behaviors are readily predictable, based on the thermal properties of the streambed sediments, the rate and direction of fluid flow, the magnitude of temperature variations, and the spacing between sensors. Time series data (T vs. t) were collected in three settings in California during 2002-03 to test this method: Russian River, Pajaro River, and Corralitos Creek. We analyzed a subset of these data using a forward model to estimate apparent seepage rates during selected intervals, then applied the time series approach to estimate seepage rates during the same time intervals. Processing of field data required filtering to extract the diurnal signal, and a series of programs was developed to aid picking of temperature peaks and troughs so that records of phase shift and amplitude ratio could be derived. We also show results of modeling studies where we vary seepage rates and generate time-series records, then back-calculate variations in seepage rate to assess the accuracy of this approach. Flow rates derived from model temperature records are similar to those estimated from differential discharge measurements during field seasons, equivalent to about 0.6 to 1.5 m/day near the end of the water year.
DE: 1800 HYDROLOGY
DE: 1829 Groundwater hydrology
DE: 1832 Groundwater transport
DE: 1836 Hydrologic budget (1655)
DE: 1899 General or miscellaneous
SC: Hydrology [H]
MN: 2003 Fall Meeting