New Mexico Geological Society Annual Spring Meeting — Abstracts

In October 1992, the Los Alamos National Laboratory began recording hourly water levels in numerous test wells that are completed into the regional aquifer below Pajarito Plateau. Daily water level fluctuations typically range from 0.25 to 0.50 ft or more with an instrument measurement error of ±0.04 ft. These data are electronically available on the web
(http://www.esh.lanl.gov/-esh18/teams/hydrology/Welldata.htm). In the frequency domain, these fluctuations can be directly related to changes in barometric pressure and solar-lunar earth tides using spectral analysis. Here the familiar horizontal time axis on a standard water level plot is replaced by a linear transformation to frequency. The vertical axis represents the Fourier transform of the autocovariance function computed from water level measurements that are uniformly distributed over time. The resulting spectrum represents water level departures from the mean expressed as a function of frequency. If water levels are affected by barometric pressure fluctuations, then strong peaks at and 2 cycles per day (cpd) will be obvious. These peaks mimic those in the atmospheric pressure spectrum and are related to atmospheric solar heating. Other strong peaks that are related to
synoptic-scale weather patterns will also be apparent at frequencies below 0.5 cpd. If water levels are affected by solar-lunar earth tides, then additional strong peaks at 0.93, 1.00, 1.93, and 2.00 cpd will be apparent in water level spectra. These frequencies correspond to the O1 lunar diurnal, the K1 solar diurnal, the M2 lunar semi-diurnal, and S2 solar semi-diurnal frequencies, respectively. These influences arise from distortions in the earth's crust in response to revolutional and rotational periodicities in the earth-moon-sun system. The overlap in solar and atmospheric pressure frequencies at 1 and 2 cpd actually provides the basis for alternati ve models than can be used to estimate hydraulic transmitting parameters.

These water level data are analyzed in the frequency domain using models developed by Hsieh et al. (1987) and Ritzi et al (1991). While type curves can be used to estimate aquifer transmissivity (T) and storage coefficient (S), the complex vector estimation criteria suggested by Ritzi et al. (1991) is much more robust. These analyses demonstrate that field values for T can generally be estimated with the atmospheric pressure variation (APV) model. Depending on water level attenuation characteristics, we are occasionally forced to use the solid earth tide (SET) test configuration. S values are typically insensitive in both models. However, the procedure recommended by Bredehoeft (1967) can also be used to estimate S. Finally, the resulting estimates for T and S are compared to results obtained from traditional pumping tests and some general conclusions are developed. It is interesting to note that not all wells display barometric pressure and earth tide effects. For example, no shallow alluvial wells on Pajarito Plateau show these influences. However, intermediate and deep observation wells show barometric influences, while only about 20% of these deep wells also display tidal effects. This pattern suggests a possible relationship between the regional groundwater flow field and the location of the well screen.