2.5.1.2 Water balance uncertainty


The components of the water balance are not included in the uncertainty analysis reported in companion product 2.6.1 (Gilfedder et al., 2016) and 2.6.2 (Cui et al., 2016) for the Clarence‑Moreton bioregion. The water balance analysis was carried out based on 30 model simulations selected from a total of 3877 simulations conducted for the sensitivity analysis reported in companion product 2.6.2 (groundwater numerical modelling) for the Clarence‑Moreton bioregion (Cui et al., 2016). The sample size of 30 was used because it is commonly considered as the threshold sample size separating small-sample statistics from large-sample statistics (Davis, 2002). While the 30 samples will not provide the rigorous uncertainty analysis that is presented in companion product 2.6.1 (Gilfedder et al., 2016) and product 2.6.2 (Cui et al., 2016) for the Clarence‑Moreton bioregion, it is expected to provide a statistically significant estimate of the 10th, 50th and 90th percentiles of the water balance components. Losses are not able to be included in the estimation of water balance uncertainty because they would not form part of the hydrological response variables used should any impact and risk analysis be conducted.

The selection of suitable simulations and subsequent output analysis involved the following steps:

  1. Filter models using the forecast water production. Only models with a median annual water production that is within the range from one-tenth of the forecast volume to ten times the forecast volume were used for further analysis.
  2. Compute the head objective function (sum of squared residuals) using the observed heads described in companion product 2.6.2 for the Clarence‑Moreton bioregion (Cui et al., 2016).
  3. Compute the baseflow objective function (sum of squared residuals) using estimated baseflow at the river stage gauge site near Casino (CLM008) (Figure 3).
  4. Rank the models based on the sum of the two normalised components of the objective function described in step 2 and step 3.
  5. Conduct a statistical summary analysis using the top 30 models derived in step 4.

The computing approaches for various water balance terms are summarised in Table 3 and Table 4.

Table 3 Computing approaches of water balance terms in the surface water model for the Clarence‑Moreton bioregion


Water balance term

Computing approach

Rainfall

Measured rainfall modified using scaling factors from global climate models (GCMs)

Surface water outflow

From AWRA-L model

Licensed extractions

AWRA-L does not model extractions

CSG produced water

Any discharge to streams is subject to stringent control and could not be included in the AWRA-L model since timing and volume is not known.

Losses

Rainfall minus surface water outflow. This includes evapotranspiration, and would also include water extractions since AWRA-L does not model them explicitly

AWRA-L is the Australian Water Resources Assessment landscape model

Table 4 Computing approaches of water balance terms in the groundwater model for the Clarence‑Moreton bioregion


Water balance term

Computing approach

Recharge (diffuse)

From the Recharge Package of MODFLOW

Recharge (streams)

From the River Package of MODFLOW

GHB boundary

From the General-Head Boundary Package of MODFLOW

Discharge (surface water)

The perennial streams were explicitly simulated in the groundwater model using the River Package of MODFLOW. Other surface water features, such as local intermittent streams and swamps, were lumped together and implicitly simulated using the Drain Package of MODFLOW. Thus, discharge to the surface water represents the sum of the discharge components of these two packages.

Non-CSG bores

From the MNW2 Package of MODFLOW that was implemented in accordance with the licensed extractions.

CSG wells

The CSG water pumping process was simulated by the Drain Package of MODFLOW. Thus, the water balance term is from the component representing CSG wells in the Drain Package.

Storage change

Directly from the storage component in MODFLOW water balance output

CSG = coal seam gas

Last updated:
11 July 2017