Westerhoff, R.S.; White, P.A. 2014 Application of equilibrium water table estimates using satellite measurements to the Canterbury Region, New Zealand. Lower Hutt, N.Z.: GNS Science GNS Science report 2013/43 25 p.
Abstract: The ‘Equilibrium Water Table’ (EWT) model calculates a long-term average of the water table surface at the global scale. This model is based on a simplified groundwater flow model and satellite-derived data including terrain elevation estimates and climate time series. The EWT model correctly calculates that the water table is generally shallow in New Zealand’s alluvial aquifer systems. Water table estimates provided by the model could be useful for filling gaps in our understanding of the water table surface where observations are sparse in New Zealand. Water table estimates provided by the EWT model were compared with water table observations in the Canterbury Region. The EWT model represents the general pattern of observed water table in the region but with three important differences: the model commonly estimates very shallow water tables of less than 1 m but observations of very shallow water table depths are rare; modelled water tables are rare in the range 2 to 30 m but are commonly observed; and large differences between water tables calculated by the model and observations (i.e., differences in the range tens of meters to 200 m) typically occur in the upper Canterbury Plains between the Ashburton and Waimakariri rivers. The differences between model calculations and observations at very shallow water tables and in the range 2 to 30 m are consistent with the global data set (Fan et al., 2013a). The differences can have multiple causes, which are outlined in the discussion. EWT model input data sets are generally a poor approximation to the characteristics of New Zealand’s environment, including its groundwater systems, which may explain the difference between modelled and observed water tables in the Canterbury Region. Therefore, we recommend further development of a New Zealand EWT model aiming to improve model results through the use of refined input data sets including: topography, soils, surface geology and recharge estimates at the national scale, where available. This requires model scripting to increase the model efficiency as the identified input data resolution will be greater than that previously used. It is also recommended that the EWT model is used to characterise the distribution of aquifer hydraulic conductivity at national and sub-regional scales. This is because the distribution of hydraulic conductivity is crucial to our understanding of the behaviour of groundwater systems but is generally poorly understood. (auth)