Author(s): Douglas N. Mzila; John Male
Keywords: Rnold River diversion; Ground water; Perched aquifer; Glacial outwash; MODFLOW; Water supply wells
Abstract: A three dimensional groundwater model was developed to assess the effects of the river diversion on groundwater resources downstream of the proposed river diversion of the Arnold River, New Zealand. The purpose of the modelling was to enhance knowledge of the aquifer system below the diversion and provide a predictive tool for assessing the effects of reduced river flows and levels on groundwater resources in this area. The investigations were structured to assess the groundwater regime under existing river flow conditions and then to impose river levels due to diversion and determine the resultant changes in the groundwater regime. The groundwater flow model was developed and implemented in the finite difference system using Visual MODFLOW V3.1. 0.86. The model was selected because of suitability for simulating groundwater flow in complex three-dimensional environments containing both largescale and local scale hydrological features. The model is widely accepted as standard in industry. Boundary conditions for the model included pre and post diversion water levels in the Arnold River. The changes in river water levels due to reduced discharge were modelled in a separate model. Modelled river levels relate to the river levels at current mean flow 42 m 3 /s and river levels at 12m3/s assumed to be residual flow in the river after the diversion. It is estimated that the diversion will result in reduction of water levels in the main channel by up to 2.5m in some sections. However, the water level reduction is mostly less than 1m throughout the main river channel. Notable changes in groundwater levels due to the diversion are experienced within a few hundred meters from the river channel and the relationship weakens as the distance increases from the river channel. The relationship is not evident at distances of more than one and two kilometres from the left and right banks of the channel respectively. Temporal and steady state groundwater levels at greater distances are predominantly controlled by rainfall stresses and seem not to respond to Arnold River channel water level changes. The modelling results also show that the diversion scheme will only result in between 0.2 to 0.6m reduction in water levels at the currently identified private water supply wells indicating limited effects on the groundwater resources of the areas downstream of the diversion. The simulated reduction in groundwater levels at these supply wells represent worst-case scenario cases where there is a strong hydraulic relationship between the river channel and the shallow aquifer system. There is a paucity of data on the total depth of these private wells. The shallow private wells are mostly advanced to the perched aquifer and currently periodically run dry. It is anticipated that any significant reduction in groundwater levels at the private wells will only be resultant from leakage from the perched aquifer zone. However, supply wells such as CMP Kokiri well that are located closer to the main Arnold water flow channel are expected to respond to the reduction in channel water levels. Affected wells could require deepening to tap underlying more productive hydrostratigraphic horizons.