2.6.1.6 Prediction


Summary

This section presents the predictive posterior distribution for each hydrological response variable. To guide the interpretation of the results, Table 7 summarises the location and catchment area of each model node.

As described in the uncertainty analysis (refer to Section 2.6.1.5), the maximum raw change (amax), maximum percent change (pmax) and year of maximum change (tmax) reported here are for the hydrological response variables for streamflow: annual flow (AF), interquartile range (IQR), daily streamflow at the 99th percentile (P99), number of flood (high-flow) days (FD), number of low-flow days (LFD), number of low-flow spells (LFS), longest low-flow spell (LLFS) and zero-flow days (ZFD).

The change in surface hydrology predicted due to the additional coal resource development in absolute terms is predicted to have a median decrease of less than 0.01 GL/y, which corresponds to a change of about 0.01%. These changes are several orders of magnitude smaller than the observed mean streamflow (Table 26, Section 2.1.4.1 of companion product 2.1-2.2 for the Clarence-Moreton bioregion (Raiber et al., 2016)). Their effect on mean and high-streamflow hydrological response variables will therefore be minimal. Even the effect on low-streamflow hydrological response variables will be very small, especially in the perennial streams.

The maximum change in surface water – groundwater flux simulated by the groundwater model is several orders of magnitude less than the observed or simulated historical streamflow. The simulated increases in low-flow metrics are considered to be an erroneous overestimate due to artefacts in the simulation of low flow and the definition of the hydrological response variables. Accurately measuring and simulating low-flow conditions is very challenging and requires further efforts.

This section presents the predictive posterior distribution for each hydrological response variable. To guide the interpretation of the results, Table 7 summarises the location and catchment area of each model node, which are also shown in Figure 5 in Section 2.6.1.3.1 . In Table 7 and the figures in this section, the model nodes are grouped per catchment and ordered from upstream to downstream.

Table 7 Summary of model nodes with their upstream contribution area


Model node

Easting

Northing

River

Area

(km2)

CLM_011

489500

6855400

Findon Creek at Terrace Creek

136

CLM_009

496657

6846680

Richmond River at Wiangaree

712

CLM_014

499528

6833860

Richmond River at Kyogle

903

CLM_001

497683

6814311

Richmond River above Eden confluence

1071

CLM_010

480200

6833566

Ironpot Creek at Toonumbar

97

CLM_013

490300

6829100

Ironpot Creek at Ettrick

185

CLM_012

492425

6818675

Eden Creek at Doubtful

582

CLM_002

497684

6813757

Eden Creek above Richmond confluence

696

CLM_003

499147

6808771

Richmond River downstream of West Casino Gas Project

1816

CLM_008

505285

6806928

Richmond River at Casino

1874

CLM_004

497685

6799908

Shannon Brook at Middle Creek confluence

241

CLM_005

497685

6799354

Middle Creek at Shannon Brook confluence

214

CLM_007

506213

6798245

Shannon Brook at Yorklea

498

CLM_006

514499

6800453

Shannon Brook at tidal limit

543

CLM_015

516017

6821200

Leycester Creek at Rock Valley

178

CLM_016

499946

6779831

Myrtle Creek at Rappville

392

Data: NSW Office of Water (Dataset 1)

The streamflow in the prediction catchments is only likely to change due to coal resource development in this bioregion via a change in the surface watergroundwater flux. As described in the uncertainty section (refer to Section 2.6.1.5), the maximum raw change (amax), maximum percent change (pmax) and year of maximum change (tmax) reported here are for hydrological response variables for streamflow: annual flow (AF), interquartile range (IQR), daily streamflow at the 99th percentile (P99), daily streamflow at the 1st percentile (P01), number of flood (high-flow) days (FD), number of low-flow days (LFD), number of low-flow spells (LFS), the longest low-flow spell (LLFS) and number of zero-flow days (ZFD). Zero streamflow is identified using the minimum detectable flow. For ease of applicability, a threshold of 0.01 ML/day is set for determining the number of ZFD for all surface water nodes (see companion submethodology M06 for surface water modelling (Viney, 2016)).

It is important to reiterate that both the calibration and uncertainty analysis indicate that the Australian Water Resources Assessment landscape model (AWRA-L) predictive capability is adequate for high-flow aspects of the hydrograph, while the predictive capability is not as good for low-flow metrics in the Clarence-Moreton bioregion. In addition to this, such low changes in flow are extremely hard to observe as the largest uncertainties in the rating curves used to transfer measured stage heights to flows are associated with low-flow measurements (Tomkins, 2014).

Last updated:
11 July 2017
Thumbnail images of the Clarence-Moreton bioregion

Product Finalisation date

18 October 2016