Data gaps and opportunities to reduce predictive uncertainty

The qualitative uncertainty analysis in Section highlighted several model choices and assumptions that have a high potential impact on the predictions, such as the connection status of the river stream network, the representation of the Cenozoic cover and Quaternary alluvial sediments, and the implementation of the coal resource development pathway. These assumptions are all driven by limited data availability.

A comprehensive assessment of the connection status of the stream network in the Galilee subregion will allow for the nuanced assumption that all but the Belyando River are maximally losing streams, and the implementation in numerical models. These assumptions imply that there is potential for a change in streamflow anywhere where a non-negative drawdown is simulated in the unconfined aquifer. Whether or not this drawdown will manifest itself as a change in streamflow will depend on local conditions, including the connection status of the stream.

The geological model developed and presented in companion product 2.1-2.2 for the Galilee subregion (Evans et al., 2018) provides a solid basis for the representation of the regional geology. Adding local detail on the Cenozoic cover sediments and the position and extent of coal seams in the upper Permian coal measures will not only allow for the making of more robust and accurate predictions of the hydrological change in the Cenozoic cover, but it will allow independent estimates of mine pumping rates, as for example has been done in Turvey et al. (2015).

Further investment in the development of the Galilee Basin hydrogeological (GBH) numerical model presented in Turvey et al. (2015), such as improving numerical stability, and integration of such a model in a probabilistic framework, such as is outlined, for example, in companion submethodology M09 (as listed in Table 1) for propagating uncertainty through models (Peeters et al., 2016), will allow for the making of more robust predictions, and formally test the effect of the technical limitations in the analytic element model, such as the layer geometry and spatially varying properties.

The sensitivity analysis indicated that the hydraulic properties of the upper Permian coal measures and the Cenozoic cover are the most influential parameters to estimate maximum drawdown and year of maximum change. Turvey et al. (2015) indicated that the current observation dataset is not well suited to constrain these parameters. The predictive uncertainty has the most potential to be reduced by gathering additional information on both the upper Permian coal measures and the Cenozoic cover. This includes both observations of the hydraulic parameters, the conductivity and storage, and observations of the state variables such as fluxes and groundwater levels.

Further discussion of some other gaps, limitations and opportunities identified from the wider body of work undertaken for this BA is provided in Section 3.7.4 of companion product 3-4 (Lewis et al., 2018) for the Galilee subregion.

Last updated:
6 December 2018
Thumbnail of the Galilee subregion

Product Finalisation date