Data gaps and opportunities to reduce predictive uncertainty

In companion products 1.1 (McVicar et al., 2014), 2.1-2.2 (Frery et al., 2018) and 2.3 (Dawes et al., 2018) for the Gloucester subregion, data and knowledge gaps are highlighted. The sensitivity analysis presented in Section and the discussion on the qualitative uncertainty analysis presented in Section, however, indicated that not all of these data gaps have the same effect on predictions.

The overall, high level conceptualisation of the Gloucester subregion outlined in companion product 2.3 (Dawes et al., 2018) is well-established. There is, however, still considerable discussion and uncertainty on the geometry of the stratigraphic units, including the position and number of coal seams and the presence, position and hydrogeological functioning of faults.

A more detailed geological model that covers the entire subregion will allow for more nuanced, less conservative numerical modelling. Such a geological model will, for example, allow to address the main source of predictive uncertainty, the mine pumping rates. Local information on the position and extent of coal seams will allow to independently estimate the mine dewatering rates.

Additional information on the presence, position and nature of large and small-scale faults will allow for a more robust stochastic generation of the fault network. This is especially needed for the stochastic generation of subseismic faults, which to date is largely based on international literature. The current parameterisation of the hydrogeological functioning of faults is conservative, that is, biased to overestimating drawdowns. More detailed research on the hydrogeological behaviour of faults, as presented in Parsons Brinckerhoff (2015), will enable this parameterisation to be nuanced. Note that in the current modelling, despite the conservative approach, faults do not appear to influence the predictions much.

The sensitivity analysis in Section did highlight that the drawdown predictions are very sensitive to hydraulic properties of the deeper sedimentary basin, especially those of the surface weathered and fractured rock layer. To better constrain the predictions there is a need to improve the knowledge of the hydraulic properties of the surface weathered and fractured rock layer, especially the storage. In addition to that, more depth-specific information on hydraulic conductivity and storage is needed to more robustly establish the variation of hydraulic properties with depth and lithology.

The dataset of groundwater level observations is limited, especially with regards to long time series of groundwater level observations. As such, the groundwater level observations, especially in the alluvium, have limited potential to directly constrain the most sensitive regional-scale parameters (see Section Such measurements, however, are invaluable to establish local flow conditions and local hydrogeological properties. These estimates can subsequently be upscaled to regional scale. Any observations that integrate spatial and temporal scales, such as water balance estimates or environmental tracers, also have great potential to constrain the regional-scale properties of the surface weathered and fractured rock layer.

Last updated:
5 November 2018
Thumbnail of the Gloucester subregion

Product Finalisation date