2.6.1.4 Calibration

Summary

This section describes the calibration of the two components of surface water modelling in the Hunter subregion. These components are the Australian Water Resources Assessment (AWRA) landscape model (AWRA-L), and the AWRA river model (AWRA-R).

AWRA-L is regionally calibrated at 14 unregulated catchments using two calibration schemes: one biased towards high streamflow, and another biased towards low streamflow. The two parameter sets obtained from the two model calibrations are used to guide the generation of the 3000 parameter sets used for the uncertainty analysis (Section 2.6.1.5) and predictions (Section 2.6.1.6). The high-streamflow and low-streamflow calibrations perform reasonably well for predicting daily runoff for a wide range of streamflow conditions. For the nine hydrological response variables predicted by the model (Section 2.6.1.6), the high-streamflow calibration outperforms the low-streamflow calibration for predicting annual flow (AF) only. The low-streamflow calibration provides better predictions for the other eight hydrological response variables. Both calibrations noticeably over estimate high-flow days (FD) and interquartile range (IQR), but under estimate zero-flow days (ZFD), low-flow days (LFD), low-flow spells (LFS) and length of the longest low-flow spell (LLFS).

AWRA-R is calibrated for 20 streamflow gauging sites of the Hunter River and tributaries. Using runoff from both AWRA-L calibrations, two concurrent AWRA-R calibrations are conducted: a high-streamflow calibration and a low-streamflow calibration. Both variants perform well overall, with the high-streamflow calibration outperforming the low-streamflow calibration. However, the low-streamflow calibration is markedly better than the high-streamflow calibration for six out of nine hydrological response variables (including all low-streamflow metrics); and marginally worse for IQR, FD and daily streamflow at the 99th percentile (P99).

This section also assesses the AWRA-R model components representing river management (including water resources assessment and allocations, dam storage volumes and dam releases) that were calibrated using simulated or modelled outputs (Section 2.6.1.3). Results show that these components of the model capture relevant aspects of river management for a wide range of climate conditions.