|
 |
Predicting future water quality in the Lower Waitaki groundwater aquifer
The Waitaki River delta is an important agricultural area, with increasing amounts of dairy farming. Agricultural and domestic water is supplied by irrigation races from the Waitaki River, and groundwater wells that tap the shallow gravel aquifer immediately below the surface. Because this aquifer is so shallow (2-15 m depth below surface), there is significant potential for pollution from agricultural activities. This study focussed on the aquifer in the delta on the southern side of the Waitaki River.
The Lower Waitaki aquifer has natural recharge mainly from the hills to the south, with some local leakage from the Waitaki River to the north. Groundwater is discharged to the Waitaki River via springs that feed streams flowing towards the river. This study involved modelling of the Lower Waitaki aquifer using Visual MODFLOW® software for water flow and MT3DMS® software for mass-transport of pollutants. The models were calibrated with existing data on piezometric surface and known pollution levels. Modelling consisted of Visual MODFLOW simulations, utilising approximately 60,000 100m by 100m cells, containing lumped aquifer-estimates for hydraulic conductivity, thickness, porosity, dispersion and decay values. Initially physical modelling using MODFLOW was required to examine the physical stresses and boundaries of the Lower Waitaki Aquifer system. MT3D was then used to simulate nitrate and bacterial flux through the aquifer over the period of 1998 – 2002 (5 years) with monthly time steps/stress periods. This calibrated model was then used to predict water quality for 5 years, from 2007 to 2012.
Modelling suggests that the main recharge mechanism for the Lower Waitaki aquifer is the excess irrigation usage for the area. The annual recharge value for irrigation usage in the Waitaki was generally of the order of 900 mm, and from modelling of various zones of irrigation intensity based on land use information and observed heads, the range for average annual recharge was from 968 – 1613 mm inclusive of rainfall. In steady state, irrigation infiltration was about 5300 L/s, or 78% of the total recharge flux. Stream recharge to the aquifer provided 21% of the total water input to the aquifer system.
Hydraulic conductivity in the aquifer varies due to the heterogeneity of the materials present. However, based on historical and recent drill log information, the Waitaki area could be divided into three major hydraulic conductivity zones. These zones closely resemble the terrace formations and soil types for the area, in which hydraulic conductivity varies from 120 m/day in the southern part of the aquifer, through to 400 m/day for intermediate terrace areas, and 750 m/day for the lower terrace areas near the Waitaki River in the north. These values for hydraulic conductivity were generalised for modelling purposes, and large variations may exist within zones and at specific sites. From the transient forward prediction modelling, faecal coliforms showed little increase in the level of contamination from 1998 – 2002 levels. There was also no spatial variation in 2012 levels of faecal coliforms from that discernible from 1998 – 2002 levels. However, for nitrate the prediction was for a 0.232 mg/L/year rise in nitrogen levels. The level of nitrogen in 2012 may reach 6 mg/L in some parts of the aquifer (mainly the higher intensity farming areas) although modelling suggested that the levels may have peaked just before that time. The spatial variation of nitrogen in the aquifer for 2012 from 1998 – 2002 levels showed an expansion of the higher concentration zones already identified.
The amount of irrigation water flux through the aquifer is currently seen as a major dispersive mechanism for nitrate, although it also induces widespread faecal coliform contamination of the aquifer. The limitation of irrigation waters to efficient application practices will reduce the amount of water flux to the aquifer and possibly have a significant effect on faecal coliform levels in groundwater. However, as nitrate is conservative, the flushing with recharge water would not occur to the same degree, and concentrations in the aquifer would increase markedly. Any reduction in available recharge will also have a medium – long term effect on groundwater levels and saturated depth in the aquifer. The management of the surface water irrigation scheme in the area in terms of bywashing and flushing of groundwater-fed springs and wetlands is important from an ecological perspective. Whilst it is not good resource management to promote water wastage, it does allow dilution of nitrate, particularly in Welcome Creek, to levels insignificant from an ecological perspective. The management of the irrigation scheme bywash and irrigation overflow is seen as primarily important from a water quality perspective in Welcome Creek. There are also some surface water takes in Welcome Creek that are reliant on bywash water to the stream. The management of Waitaki River flows is also important from a hydraulic gradient perspective. The nitrate flux predicted from the aquifer is likely to have a more pronounced effect on the associated springs and wetlands bordering the Waitaki River. Local flushing effects of surface irrigation schemes and the Waitaki River are (currently) able to cope with elevated nitrate loads. Changes in boundary parameters of the Waitaki River (level of the river) inevitably induce changes in the hydraulic gradient and groundwater flow vector to local springs and wetlands. Any reduction of Waitaki River mean flow (levels) as modelled, will increase the mobility of the higher nitrate concentration zone toward Welcome Creek and wetland areas adjacent to the river.
|
|
 |
 |
