| Abstract | The effects of ground water and precipitation are investigated with the aim of reducing an uncertainty of several microgals which remains in both superconducting and absolute gravity measurements at the
Canadian Absolute Gravity Site in the days-to-seasons frequency band after corrections have been made for tides, polar motion, local atmospheric pressure and presumed gravimeter drift (in the case of the superconducting gravimeter). The study is
based on superconducting gravimeter data (GWR12) generated during the period 1990-1993. Daily values of cumulative precipitation, maximum potential evapotranspiration, and snowmelt potential are estimated from precipitation data, snow-cover data,
temperature and energy considerations and then regressed on year-long residual gravity data. This accounts for about 90% of the variance in each year-long record of gravity residuals. From the annual regression coefficients, a single set of four
coefficients (trend is also fitted) can be selected which, with a single reservation, yields a satisfactory fit to the full 3 1/2 years of residual data. The single reservation concerns the estimate of snowmelt effect for one of the four spring
periods and needs further attention. Higher-order regression analysis of precipitation on gravity and precipitation on well-level reveals interesting dynamic effects due to ground water movement. |