The Development of an Aquifer Vulnerability Assessment Methodology with Application to the Manawatu Area of New Zealand

Abstract

In Manawatu, New Zealand four factors predominantly control vulnerability: the soil, the unsaturated zone, rainfall recharge, and the aquifer medium. This study models these factors, and the uncertainty of their delineation regionally. The vulnerability assessment is based on available meteorological, soil, and groundwater data and the modelling is computerised using Geographical Information Systems (GIS). The effect of soil on groundwater vulnerability is modelled by mapping two soil attributes: organic carbon content and cation exchange capacity. The spatial distribution of these two parameters was determined from correlation of data from the New Zealand Land Resource Inventory and the Manawatu Wanganui Regional Legend. Recharge modelling is based on a simple 'leaking bucket' soil moisture model; recharge to groundwater occurring only if the soil moisture content is above field capacity. The model calculates mean annual recharge for each rainfall station using the Monte-Carlo technique. Randomised soil moisture parameters were used to approximate the inherent variability of soils. The resulting spatially-modelled surface shows the expected high level of recharge, above 200 mm, over high rainfall areas and a good agreement between modelled and actual groundwater levels. This indicates high recharge vulnerability in the study area. Uncertainty of recharge estimation is expressed as the standard deviation of the estimates of mean annual recharge. The impact of the unsaturated zone is based on the total sorption capacity at each bore. This assessment uses welldriller's logs which provide the only available database covering the entire unsaturated zone of the study area. These "point estimates" of total sorption capacity are treated as a spatially distributed parameter which can be modelled as a continuous surface. Since the base of the unsaturated zone is the top of the phreatic surface, this was modelled to define the lower limit of the layers of interest. Welldriller's descriptions of the unsaturated zone, between this surface and the base of the ‘soil’, were encoded. An unsaturated zone sorption index was calculated for each bore: using welldriller's descriptions of the various lithologic units, their properties, and their thicknesses. These indices were then modelled regionally. Although the quality of driller's logs vary considerably, 'high' and ‘low’ sorption areas can be readily distinguished and the final map of unsaturated zone sorption shows clear regional trends. The sorption index generally decreases, and therefore unsaturated zone vulnerability increases, with increasing distance from the recharge areas. Aquifer medium was assessed on the basis of welldriller's logs in a similar manner to the modelling of the unsaturated zone. Aquifer media has been classified, in the order of increasing vulnerability as gravel, gravel and sand, sand, and fine material. Results of the individual vulnerability factor modelling were integrated into a 'vulnerability sheet’. A new colour-coding scheme was developed: colours of the combined map represent the main control on vulnerability, or the most vulnerable factor. This colour coding scheme outlines the different vulnerability settings within the study area. The methodology developed can be applied, with modifications when necessary, to other hydrogeological settings.