Use of atmospheric tides to estimate the hydraulic conductivity of confined and semi-confined aquifers

Characterizing groundwater responses to natural drivers is cost effective and offers great potential in hydrogeological inves-tigations. However, there is a lack of method development and evaluation, for example by comparing results with those derived from using conventional methods. This paper presents a modified method to calculate the hydraulic conductivity (K) of confined aquifers using the well water response to atmospheric tides. The approach separates the Earth and atmospheric tide influences on filtered well water-level records in the time domain. The resulting ill-posed regression deconvolution problem can be overcome by constraining the well water response to atmospheric tides in order to follow a physically real-istic semi-diurnal barometric response function (S-2-BRF), or to follow directly a modified hydraulic model (BE-Hvorslev) similar to a slug test evaluation. An analysis with synthetic data shows that K up to 10(-4) m/s can be estimated when pressure records with short sampling intervals are available. Application to a field dataset from Cambodia and Benin, with 20-minute to 60-minute sampling intervals, respectively, results in K values of 5.82 center dot 10(-7) m/s and 2.9<middle dot>10(-7) m/s. This agrees with results independently derived from pumping tests for both confined sediments and semi-confined hard-rock conditions. This method offers a promising and low-cost approach to derive K solely from monitoring datasets in confined aquifers. This is especially advantageous for low-conductivity formations where hydraulic testing takes time.