Dpiw – Surface Water Models Mersey River Catchment

DPIW – SURFACE WATER MODELS MERSEY RIVER CATCHMENT

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

DOCUMENT INFORMATION

JOB/PROJECT TITLE Surface Water Hydrological Models for DPIW CLIENT ORGANISATION Department of Primary Industries and Water CLIENT CONTACT Bryce Graham

DOCUMENT ID NUMBER WR 2007/027 JOB/PROJECT MANAGER Mark Willis JOB/PROJECT NUMBER E200690/P202167 Document History and Status Revision Prepared Reviewed Approved Date Revision by by by approved type 1.0 J. Bennett Dr Fiona C. Smythe July 2007 Final Ling

1.1 J. Bennett Dr Fiona C. Smythe July 2008 Final Ling

Current Document Approval PREPARED BY James Bennett

Water Resources Mngt Sign Date

REVIEWED BY Dr Fiona Ling

Water Resources Mngt Sign Date

APPROVED FOR Crispin Smythe SUBMISSION Water Resources Mngt Sign Date Current Document Distribution List Organisation Date Issued To DPIW July 2008 Bryce Graham

The concepts and information contained in this document are the property of Hydro Tasmania. This document may only be used for the purposes of assessing our offer of services and for inclusion in documentation for the engagement of Hydro Tasmania. Use or copying of this document in whole or in part for any other purpose without the written permission of Hydro Tasmania constitutes an infringement of copyright.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

EXECUTIVE SUMMARY

This report describes the results of the hydrological model developed for the Mersey River catchment in central-north Tasmania. This report is one of a series of reports that present the methods and results from the development and calibration of surface water hydrological models for 26 Tasmanian catchments under both current and natural flow conditions.

A catchment flow model was developed for the Mersey River, and run under three scenarios:

• Scenario 1 – No Entitlements (Natural Flow);

• Scenario 2 - With Entitlements (with water entitlements extracted);

• Scenario 3 - Environmental Flows and Entitlements (Water entitlements extracted, however low priority entitlements are limited by an environmental flow threshold).

The results of these model runs allowed the calculation of indices of hydrological disturbance. These indices were:

• Hydrological Disturbance Index

• Index of Mean Annual Flow

• Index of Flow Duration Curve Difference

• Index of Seasonal Amplitude

• Index of Seasonal Periodicity

The indices were calculated using formulas developed for the Natural Resource Management (NRM) Monitoring and Evaluation Framework developed by SKM for the Murray-Darling Basin (MDBC 08/04).

A user interface is provided that allows the model to be run under varying scenarios. For information on the use of the user interface refer to the Operating Manual for the NAP Region Hydrological Models (Hydro Tasmania 2004a). This allows the user to see what effect additional extractions can have on the availability of water in the Mersey catchment. The interface provides summaries of flow statistics, duration curves, hydrological indices and water entitlements data for each subarea of the catchment. For information on the use of the user interface refer to the Operating Manual for the NAP Region Hydrological Models (Hydro Tasmania 2004b).

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

CONTENTS

EXECUTIVE SUMMARY ii 1. INTRODUCTION 1 2. CATCHMENT CHARACTERISTICS 2 3. DATA COMPILATION 4 3.1 Climate data (Rainfall & Evaporation) 4 3.2 Advantages of using climate DRILL data 4 3.3 Transposition of climate DRILL data to local catchment 5 3.4 Comparison of Data Drill rainfall and site gauges 7 3.5 Streamflow data 9 3.6 Irrigation and water use 9 3.7 Estimation of unlicensed dams 17 3.8 Environmental flows 19 4. MODEL DEVELOPMENT 22 4.1 Catchment Subarea Delineation 22 4.2 Hydstra Model 22 4.2.1 Accounting for Mersey Hydro Electric Scheme 23 4.3 AWBM Model 26 4.3.1 Channel Routing 29 4.4 Model Calibration 29 4.4.1 Accounting for flow diversions on the Mersey River 29 4.4.2 Calibration Method 30 4.4.3 Adopted Model Parameters 31 4.4.4 Model accuracy: Qualitative description 33 4.4.5 Factors affecting the reliability of the model calibration 37 4.4.6 Model Accuracy – Model fit statistics 39 4.4.7 Model Accuracy throughout the Mersey catchment 44 4.5 Model results 45 4.5.1 Indices of hydrological disturbance 46 4.6 Flood frequency analysis 49 5. REFERENCES 50 5.1 Personal Communications 51 6. GLOSSARY 52

iii

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

LIST OF FIGURES

Figure 2-1 Mersey catchment subarea boundaries 3

Figure 3-1 Climate DRILL Site Locations 6

Figure 3-2 Rainfall and Data DRILL Comparisons 8

Figure 3-3 WIMS (Dec 2006) Water Allocations in the Mersey Catchment 16

Figure 4-1 Hydstra Model Schematic 25

Figure 4-2 Two-tap Australian Water Balance Model Schematic 28

Figure 4-3 Monthly Variation of CapAve Parameter 33

Figure 4-4 Daily time series - typical year (ML/d). Good fit 34

Figure 4-5 Daily time series – low inflow year (ML/d). Good fit. 34

Figure 4-6 Daily time series comparison – high inflow year (ML/d). Good fit. 35

Figure 4-7 Monthly Time Series comparison – Volume (ML) 35

Figure 4-8 Long term average monthly, seasonal and annual flows 36

Figure 4-9 Duration curve – MCF Daily flow proportional difference 41

Figure 4-10 Duration curve - MCF Monthly volume proportional difference 42

Figure 4-11 Duration curve – UIM Daily flow proportional difference 43

Figure 4-12 Duration curve – UIM Monthly volume proportional difference 44

Figure 4-13 Time series of Monthly Volumes – Arm above Mersey (TSM 624.1) (SC9) 45

Figure 4-14 Time series of Monthly Volumes – Don River upstream of Bass Highway (TSM 16200.1) (SC30) 45

Figure 4-15 Daily Duration Curve for Modelled flows 01/01/1900 – 01/01/2006 at the Calibration Site (SC6) 46

Figure B-1 Forth catchment – monthly volumes at secondary site. 59

Figure B-2 George catchment – monthly volumes at secondary site. 59

Figure B-3 Leven catchment – monthly volumes at secondary site. 60

Figure B-4 Swan catchment – monthly volumes at secondary site. 60

Figure B-5 Montagu catchment – monthly volumes at secondary site. 61

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

LIST OF TABLES

Table 3-1 Data DRILL Site Locations 7

Table 3-2 Calibration Site 9

Table 3-3 Assumed Surety of Unassigned Records 10

Table 3-4 Estimated Unlicenced Direct Water Extractions from Mersey and Don River Catchments 11

Table 3-5 Monthly water extractions (ML) from Mersey River by subarea 12

Table 3-6 Average capacity for dams less than 20 ML from Neal et al. (2002) 18

Table 3-7 Environmental Flows 20

Table 4-1 Boughton & Chiew, AWBM surface storage parameters 26

Table 4-2 Hydstra/TSM Modelling Parameter Bounds 29

Table 4-3 Adopted Calibration Parameters 32

Table 4-4 Long term average monthly, seasonal and annual comparisons 37

Table 4-5 Model Fit Statistics – Mersey River 39

Table 4-6 Coefficient of Determination (R 2) Fit Categories 40

Table 4-7 Hydrological Disturbance Indices at the Catchment Outflow measuring disturbance between Scenario 1 and Scenario 3 at 3 sites in the Mersey Catchment 48

Table A-1 Hydro Tasmania Sites used for long term mean model inputs (ML/day)55

Table B-2 Tascatch Models’ performance at secondary sites 62

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

1. INTRODUCTION

This report forms part of a larger project commissioned by the Department of Primary Industries and Water (DPIW) to provide hydrological models for 26 regional catchments.

The main objectives for the individual catchments are:

• To compile relevant data needed to develop and calibrate an Australian Water Balance Model (AWBM) hydrological model for the Mersey River catchment;

• To compile more than 100 years of daily time-step rainfall and streamflow data for input to the hydrologic model;

• To develop and calibrate the hydrologic model under both natural and current catchment conditions;

• To develop a User Interface for running the model under varying scenarios;

• To prepare a report that summarises the methods and assumptions used to develop the model. This report discusses the results of calibration and validation as well as material relating to the use of the hydrologic model (and associated software).

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

2. CATCHMENT CHARACTERISTICS

The Mersey River is fed by a 1680 km 2 catchment in central-north Tasmania. It flows northward and discharges into Bass Strait at the city of Devonport. The south of the catchment is distinguished by the steep, mountainous terrain and alpine plateaus of the central highlands of Tasmania, and includes Tasmania’s tallest mountain, Mt Ossa, which stands at 1617 m ASL. Further north the catchment becomes less mountainous but remains undulating. Narrow alluvial plains form along the banks of the Mersey in the north of the catchment.

Land-use in the Mersey catchment is divided between agricultural areas and large tracts of protected native forest and alpine tundra. The alluvial plains - about one tenth of the catchment area - are largely dedicated to agriculture. The hills in the catchment (the remaining catchment area) – particularly the mountains of the south - are covered in native forests. There are several towns in the Mersey catchment area, but the quantity of land covered by urban areas is negligible compared to forested and agricultural land.

Annual rainfall varies significantly across the catchment owing to the catchment’s large size and diverse topography. Average annual rainfall ranges from 2000 mm in the southern mountains to 800 mm in the catchment’s north.

This model also simulates flows for the adjoining Don River catchment, which is located to the west of the Mersey and also discharges into Bass Strait. The Don River is fed by a 130 km 2 catchment.

For modelling purposes, the Mersey and Don catchments were divided into 41 subareas. The delineation of these areas is shown in Figure 2-1.

There are 691 registered current entitlements for water extraction. These entitlements are spread across 31 subareas, all located in the northern half of the catchment. Most extraction entitlements are for agriculture, but a small number are used for water supply, recreational, aesthetic and commercial purposes.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Figure 2-1 Mersey catchment subarea boundaries

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

3. DATA COMPILATION

3.1 Climate data (Rainfall & Evaporation) Daily time-step climate data were obtained from the Queensland Department of Natural Resources & Mines (QDNRM).

QDNRM provides interpolated evaporation and rainfall data (called ‘climate DRILL data’) at intervals of 0.05 o latitude and 0.05 o longitude (i.e., grid points on a grid of squares approximately 5 by 5 km in size). These interpolated rainfall and evaporation data are based on over 6000 rainfall and evaporation stations in Australia (see www.nrm.qld.gov.au/silo for further details of climate drill data).

3.2 Advantages of using climate DRILL data These data have a number of benefits over other sources of rainfall data including:

• Continuous data back to 1889. (However, for earlier years there are fewer input sites available and therefore quality is reduced. The makers of the data-set state that gauge numbers have been somewhat static since 1957, therefore back to 1957 distribution is considered “good” but prior to 1957 site availability may need to be checked in the study area.)

• Evaporation data (along with a number of other climatic variables) are also included for use in the AWBM model. According to the QDNRM web site, all Data Drill evaporation information combines a mixture of the following data.

1. Observed data from the Commonwealth Bureau of Meteorology (BoM).

2. Interpolated daily climate surfaces from the on-line NR&M climate archive.

3. Observed pre-1957 climate data from the CLIMARC project (LWRRDC QPI- 43). NR&M was a major research collaborator on the CLIMARC project, and these data have been integrated into the on-line NR&M climate archive.

4. Interpolated pre-1957 climate surfaces. This data-set, derived mainly from the CLIMARC project data, is available in the on-line NR&M climate archive.

5. Incorporation of Automatic Weather Station (AWS) datum records. Typically, an AWS is placed at a user's site to provide accurate local weather measurements.

The evaporation data derived for the Mersey catchment were examined. Before 1970 the

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1 evaporation information is based on the long-term daily averages of data collected after 1970. In the absence of any reliable long-term site data this evaporation data-set is considered to be the best available for this catchment.

3.3 Transposition of climate DRILL data to local catchment Ten climate Data Drill sites were selected to give coverage of the Mersey catchment. Because the Mersey catchment is large, data DRILL sites were chosen to reflect the range of annual rainfall: i.e. where rainfall varied greatly across small areas (notably in the mountainous south of the catchment) more DRILL sites were selected.

See Figure 3-1 below for a map of the climate Data Drill sites and Table 3-1 for the location information.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Figure 3-1 Climate DRILL site locations

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 3-1 Data DRILL Site Locations

Site Latitude Longitude Mersey_01 -41:15:00 146:27:00 Mersey_02 -41:18:00 146:21:00 Mersey_03 -41:24:00 146:18:00 Mersey_04 -41:24:00 146:27:00 Mersey_05 -41:33:00 146:15:00 Mersey_06 -41:33:00 146:27:00 Mersey_07 -41:39:00 146:24:00 Mersey_08 -41:42:00 146:15:00 Mersey_09 -41:48:00 146:12:00 Mersey_10 -41:54:00 146:09:00

3.4 Comparison of Data Drill rainfall and site gauges As rainfall data are critical inputs to the model it is important to have confidence that the Data DRILL long-term generated time series reflect what is observed within the catchment. There were a number of Hydro Tasmania sites available within the catchment that provided almost complete daily rainfall records longer than 10 years. These records were compared to the nearest Data DRILL ‘virtual’ rain gauges (Figure 3-2). Visual inspection and R 2 values indicate an acceptable correlation between the DRILL interpolated annual rainfalls and more recent rainfall records (since c. 1975). However, upon visual inspection DRILL data showed less fidelity to rainfalls from an earlier record (1956-1968 - Figure 3-2). As noted earlier, DRILL evaporation data before 1970 are also based on post-1970 means. As neither evaporation nor rainfall data are completely reliable before 1970, hindcast flows for periods preceding 1970 should be treated with caution.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

R2 = 0.94

R2 = 0.89

R2 = 0.87

Figure 3-2 Rainfall and Data DRILL comparisons

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

3.5 Streamflow data The stream flow gauge site at Kimberley was identified as a suitable calibration site. The site is located low in the catchment (see Figure 2-1) and has a long flow record, detailed in Table 3-2.

Table 3-2 Calibration Site

Site Name Site No. Sub- Period of Latitude Longitude catchment Record Location

Mersey at Kimberley 22 SC6 08/03/1921 -41:23:50.2 146:29:40.5 - Present

This flow record was retrieved from the Hydro Tasmania database. A brief investigation of the site rating history on Hydro Tasmania archives revealed that seven ratings applied to the site between 01/01/1986 and 01/01/2006 (the period used for calibration). The site has a naturally controlled water body that is regularly monitored for rating changes, and the record is considered reasonably reliable.

3.6 Irrigation and water use Information on the current water entitlement allocations in the catchment was obtained from DPIW from the Water Information Management System (WIMS) December 2006 data-set. The extractions or licenses in the catchment are of a given Surety (from 1 to 8), with Surety 1-3 representing high priority extractions for modelling purposes and Surety 4-8 representing the lowest priority. The data provided by DPIW include a number of sites that had a surety of 0. DPIW staff advised that in these cases the surety should be determined by the extraction “Purpose” and assigned as shown in Table 3-3, below.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 3-3 Assumed Surety of Unassigned Records

Purpose Surety Aesthetic 6 Aquaculture 6 Commercial 6 Domestic 1 Industrial 6 Irrigation 6 Storage 6 Other 6 Power Generation 6 Recreation 6 Stock and Domestic S & D 1 Stock 1 Water Supply 1

In total there were 3184 ML unassigned entitlements (surety = 0) identified for inclusion in the surface water model, of which 1038 ML were assigned Surety 1 and 2146 ML assigned Surety 6.

DPIW staff also advised that the water extraction information provided should be filtered to remove the following records:

• Extractions relating to fish farms should be omitted as this water is returned to the stream. These are identified in the WIMS database by the purpose labels “acquacult” or “fish farm” . One fish farm was identified in the catchment.

• The extraction data-set includes a “WE_status” field where only “ current” and “existing” should be used for extractions. All other records, for example deleted, deferred, transferred, suspended and proposed, should be omitted.

When modelling Scenario 3 (Environmental Flows with Entitlements), water will only be available for Low Priority entitlements after environmental flow requirements have been met.

DPIW estimated that a total of 8889 ML were extracted in addition to allocations currently recorded in the WIMS database. Unlicensed direct extractions in the Mersey and Don River catchments were assigned to individual streams and tributaries (Table 3-4), and were assigned to subareas in the model user interface accordingly. Where streams flowed through a number of subareas (e.g. Dasher River, Mersey River), unlicensed direct extractions were assigned according to the proportions of licensed direct extractions in these streams, in the absence of data to the contrary. Allowances for unlicensed dam extractions are discussed in Section 3.7.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 3-4 Estimated Unlicensed Direct Water Extractions from Mersey and Don River Catchments

Estimated Extractions Subareas Where Stream (ML) Extractions Apply Bella Macargie Creek & SC30 Tributaries (Don River) 93 SC15 Bonneys Creek & Tributaries 342 SC33 Cockers Creek & Tributaries 112 SC11 Coiler Creek & Tributaries 691 SC24, SC10, SC7 Dasher River & Tributaries 1232 SC7 Dodder Rivulet 304 SC35, SC36, SC37, SC38, SC39 Don River 1 464 SC31 Figure of Eight Creek 72 Knights Creek & Tributaries and SC13 Greens Creek & Tributaries 271 SC29 Latrobe Creek & Tributaries 133 SC21, SC19 Lobster Rivulet & Tributaries 351 SC16 Mersey River 3187 SC23 Minnow Creek & Tributaries 475 SC12, SC14 Mole Creek 540 SC13 Redwater Creek & Tributaries 493 SC10 Smiths Creek 433 SC30 Stave Creek & Tributaries 159 Total 9353

A summary table of monthly water extraction volumes by subarea is shown in Table 3-5 and in the Catchment User Interface. A map of the water extraction allocations in the catchment is shown in Figure 3-3.

1 Don River Direct extractions extrapolated from Bella Macargie Creek Estimate. It was assumed extractions in each Don River Subarea would be the same as the extraction for Bella Macargie Creek, which is wholly contained in subarea SC30.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 3-5 Monthly water extractions (ML) from Mersey River by subarea

Subcatch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

High Priority Entitlements

SC1 30.74 27.77 30.74 29.75 0.57 0.55 0.57 0.57 0.55 30.74 29.75 30.74 213 SC2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC5 67.10 60.60 67.10 64.93 1.42 1.37 1.42 1.42 1.37 67.10 64.93 67.10 466 SC6 58.68 53.00 58.68 56.79 8.11 7.85 8.11 8.11 7.85 58.68 56.79 58.68 441 SC7 136.44 123.24 136.44 132.04 46.37 44.88 46.37 46.37 44.88 136.44 132.04 136.44 1,162 SC8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC10 132.29 119.48 132.29 128.02 33.50 32.42 33.50 33.50 32.42 132.29 128.02 132.29 1,070 SC11 118.00 106.58 118.00 114.19 46.49 44.99 46.49 46.49 44.99 118.00 114.19 118.00 1,036 SC12 74.84 67.60 74.84 72.43 21.72 21.02 21.72 21.72 21.02 74.84 72.43 74.84 619 SC13 103.06 93.09 103.06 99.74 5.11 4.94 5.11 5.11 4.94 103.06 99.74 103.06 730 SC14 0.00 0.00 0.00 0.00 1.42 1.37 1.42 1.42 1.37 0.00 0.00 0.00 7 SC15 50.07 45.22 50.07 48.45 10.50 10.16 10.50 10.50 10.16 50.07 48.45 50.07 394 SC16 70.30 63.50 70.30 68.03 26.45 25.60 26.45 26.45 25.60 70.66 68.38 70.30 612 SC17 61.94 55.95 61.94 59.94 2.23 2.16 2.23 2.23 2.16 61.94 59.94 61.94 435 SC18 1.06 0.96 1.06 1.03 0.00 0.00 0.00 0.00 0.00 1.06 1.03 1.06 7 SC19 40.07 36.19 40.07 38.77 6.32 6.11 6.32 6.32 6.11 40.07 38.77 40.07 305 SC20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC21 13.42 12.12 13.42 12.98 15.41 14.92 15.41 15.41 14.92 13.42 12.98 13.42 168 SC22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC23 85.06 76.86 85.06 82.39 8.08 7.90 8.08 8.08 7.90 85.06 82.39 85.06 622 SC24 0.38 0.35 0.38 0.37 2.37 2.29 2.37 2.37 2.29 0.38 0.37 0.38 14 SC25 248.58 224.52 248.58 240.56 12.15 11.76 12.15 12.15 11.76 248.58 240.56 248.58 1,760 SC26 0.00 0.00 0.00 0.00 9.36 9.06 9.36 9.36 9.06 0.00 0.00 0.00 46 SC27 0.00 0.00 0.00 0.00 0.57 0.55 0.57 0.57 0.55 0.00 0.00 0.00 3 SC28 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC29 24.27 21.92 24.27 23.49 14.68 14.21 14.68 14.68 14.21 24.27 23.49 24.27 238 SC30 19.54 17.65 19.54 18.91 31.78 30.76 31.78 31.78 30.76 19.54 18.91 19.54 291 SC31 10.63 9.61 10.63 10.29 15.42 14.93 15.42 15.42 14.93 10.63 10.29 10.63 149 SC32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC33 16.38 14.79 16.38 15.85 3.40 3.29 3.40 3.40 3.29 16.38 15.85 16.38 129 SC34 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC35 26.90 24.30 26.90 26.04 35.74 34.59 35.74 35.74 34.59 26.90 26.04 26.90 360 SC36 15.60 14.09 15.60 15.10 19.62 18.98 19.62 19.62 18.98 15.60 15.10 15.60 204 SC37 14.67 13.25 14.67 14.20 8.76 8.48 8.76 8.76 8.48 14.67 14.20 14.67 144

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Subcatch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

SC38 32.47 29.33 32.47 31.42 63.72 61.66 63.72 63.72 61.66 32.47 31.42 32.47 537 SC39 17.83 16.11 17.83 17.26 57.59 55.73 57.59 57.59 55.73 17.83 17.26 17.83 406 SC40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - SC41 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - Total 1,470 1,328 1,470 1,423 509 493 509 509 493 1,471 1,423 1,470 12,568

Low Priority Entitlements

SC1 45.58 41.17 45.58 44.11 1.45 1.40 1.45 1.45 1.40 1.45 1.40 45.58 232

SC2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC5 97.34 109.89 121.66 117.74 13.18 12.76 13.18 13.18 12.76 13.18 12.76 37.88 576

SC6 45.19 40.82 45.19 43.73 36.92 35.73 36.92 36.92 35.73 36.92 35.73 45.19 475

SC7 18.42 16.63 18.42 17.82 67.43 65.26 67.43 67.43 65.26 67.43 65.26 18.42 555

SC8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC10 17.45 15.76 17.45 16.89 46.31 44.82 46.31 46.31 44.82 46.31 40.34 17.45 400

SC11 37.60 33.96 37.60 36.38 47.88 46.33 47.88 47.88 46.33 47.88 46.33 37.60 514

SC12 4.12 3.72 4.12 3.99 170.83 165.32 170.83 170.83 165.32 170.83 45.12 4.12 1,079

SC13 2.31 2.09 2.31 2.24 43.27 41.88 43.27 43.27 41.88 43.27 41.88 2.31 310

SC14 46.33 41.85 46.33 44.84 47.68 46.14 47.68 47.68 46.14 47.68 44.84 46.33 554

SC15 0.08 0.08 0.08 0.08 37.31 36.11 37.31 37.31 36.11 37.31 36.11 0.08 258

SC16 12.88 11.63 12.88 12.46 77.15 74.66 77.15 77.15 74.66 77.15 72.87 12.88 594

SC17 91.46 82.61 91.46 88.51 2.28 2.21 2.28 2.28 2.21 2.28 2.21 91.46 461

SC18 0.65 0.59 0.65 0.63 0.65 0.63 0.65 0.65 0.63 0.65 0.63 0.65 8

SC19 39.39 35.58 39.39 38.12 45.52 44.06 45.52 45.52 44.06 45.52 22.86 29.97 476

SC20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC21 8.58 7.75 8.58 8.30 82.82 80.15 82.82 82.82 80.15 82.82 80.15 8.58 614

SC22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC23 347.10 314.40 354.64 347.52 249.61 241.59 249.61 249.61 241.59 249.61 241.59 337.94 3,425

SC24 0.00 0.00 0.00 0.00 1.30 1.26 1.30 1.30 1.26 1.30 1.26 0.00 9

SC25 364.98 329.66 364.98 353.21 20.99 20.31 20.99 20.99 20.31 20.99 20.31 364.98 1,9 23

SC26 69.05 62.37 69.05 66.82 99.58 96.37 99.58 99.58 96.37 99.58 96.37 69.05 1,024

SC27 0.00 0.00 0.00 0.00 2.02 1.96 2.02 2.02 1.96 2.02 0.00 0.00 12

SC28 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC29 58.87 53.17 58.87 56.97 27.37 26.49 27.37 27.37 26.49 27.37 26.49 58.87 476

SC30 15.80 14.27 15.80 15.29 32.74 31.68 32.74 32.74 31.68 32.74 31.68 15.80 303

SC31 0.21 0.19 0.21 0.21 10.48 10.14 10.48 10.48 10.14 10.48 10.14 0.21 73

SC32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC33 0.44 0.40 0.44 0.43 4.79 4.63 4.79 4.79 4.63 4.79 4.63 0.44 35

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Subcatch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

SC34 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC35 5.73 5.17 5.73 5.54 33.88 32.79 33.88 33.88 32.79 33.88 32.05 5.73 261

SC36 1.78 1.61 1.78 1.73 21.99 21.28 21.99 21.99 21.28 21.99 21.28 1.78 161

SC37 8.49 7.67 8.49 8.22 17.69 17.12 17.69 17.69 17.12 17.69 10.60 8.49 157

SC38 19.75 19.76 21.88 17.97 76.36 73.90 76.36 76.36 73.90 76.36 66.48 16.37 615

SC39 28.48 25.72 33.12 32.92 95.69 92.60 95.69 95.69 92.60 95.69 78.89 28.48 796

SC40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

SC41 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -

Total 1,388 1,279 1,427 1,383 1,415 1,370 1,415 1,415 1,370 1,415 1,190 1,307 16,373

All Entitlements

SC1 76.32 68.93 76.32 73.86 2.02 1.95 2.02 2.02 1.95 32.19 31.15 76.32 445 SC2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC5 164.43 170.49 188.76 182.67 14.60 14.13 14.60 14.60 14.13 80.28 77.69 104.97 1041 SC6 103.87 93.82 103.87 100.52 45.03 43.58 45.03 45.03 43.58 95.60 92.52 103.87 916 SC7 154.86 139.87 154.86 149.86 113.80 110.13 113.80 113.80 110.13 203.87 197.30 154.86 1717 SC8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC10 149.74 135.25 149.74 144.91 79.82 77.24 79.82 79.82 77.24 178.60 168.36 149.74 1470 SC11 155.59 140.54 155.59 150.57 94.37 91.32 94.37 94.37 91.32 165.87 160.52 155.59 1550 SC12 78.96 71.32 78.96 76.42 192.55 186.34 192.55 192.55 186.34 245.68 117.55 78.96 1698 SC13 105.37 95.17 105.37 101.97 48.38 46.82 48.38 48.38 46.82 146.33 141.61 105.37 1040 SC14 46.33 41.85 46.33 44.84 49.10 47.52 49.10 49.10 47.52 47.68 44.84 46.33 561 SC15 50.15 45.30 50.15 48.54 47.81 46.27 47.81 47.81 46.27 87.38 84.56 50.15 652 SC16 83.17 75.13 83.17 80.49 103.60 100.26 103.60 103.60 100.26 147.81 141.25 83.17 1206 SC17 153.40 138.56 153.40 148.45 4.51 4.37 4.51 4.51 4.37 64.22 62.15 153.40 896 SC18 1.71 1.54 1.71 1.65 0.65 0.63 0.65 0.65 0.63 1.71 1.65 1.71 15 SC19 79.45 71.76 79.45 76.89 51.84 50.17 51.84 51.84 50.17 85.59 61.63 70.03 781 SC20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC21 22.00 19.87 22.00 21.29 98.23 95.06 98.23 98.23 95.06 96.23 93.13 22.00 781 SC22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC23 432.15 391.26 439.70 429.91 257.69 249.49 257.69 257.69 249.49 334.66 323.98 422.99 4047 SC24 0.38 0.35 0.38 0.37 3.67 3.55 3.67 3.67 3.55 1.69 1.63 0.38 23 SC25 613.56 554.18 613.56 593.77 33.14 32.07 33.14 33.14 32.07 269.56 260.87 613.56 3683 SC26 69.05 62.37 69.05 66.82 108.94 105.42 108.94 108.94 105.42 99.58 96.37 69.05 1070 SC27 0.00 0.00 0.00 0.00 2.59 2.51 2.59 2.59 2.51 2.02 0.00 0.00 15 SC28 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC29 83.14 75.10 83.14 80.46 42.05 40.70 42.05 42.05 40.70 51.64 49.98 83.14 714

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Subcatch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

SC30 35.34 31.92 35.34 34.20 64.52 62.44 64.52 64.52 62.44 52.28 50.59 35.34 593 SC31 10.85 9.80 10.85 10.50 25.90 25.06 25.90 25.90 25.06 21.11 20.43 10.85 222 SC32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC33 16.82 15.19 16.82 16.28 8.19 7.93 8.19 8.19 7.93 21.16 20.48 16.82 164 SC34 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC35 32.63 29.47 32.63 31.58 69.62 67.38 69.62 69.62 67.38 60.78 58.09 32.63 621 SC36 17.38 15.70 17.38 16.82 41.61 40.27 41.61 41.61 40.27 37.59 36.38 17.38 364 SC37 23.17 20.93 23.17 22.42 26.46 25.60 26.46 26.46 25.60 32.37 24.80 23.17 301 SC38 52.21 49.09 54.35 49.39 140.08 135.56 140.08 140.08 135.56 108.83 97.90 48.84 1152 SC39 46.31 41.83 50.96 50.18 153.28 148.33 153.28 153.28 148.33 113.52 96.15 46.31 1,202 SC40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 SC41 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 Total 2,858 2,607 2,897 2,806 1,924 1,862 1,924 1,924 1,862 2,886 2,614 2,777 28,940

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Figure 3-3 WIMS (Dec 2006) Water Allocations in the Mersey catchment

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

3.7 Estimation of unlicensed dams Under current Tasmanian law a dam permit is not required for a dam if it is not on a watercourse and holds less than 1ML of water storage (prior to 2000 it was 2.5 ML), and is used only for stock and domestic purposes. Therefore there are no records for these storages. The storage volume attributed to unlicensed dams was estimated by analysis of aerial and satellite photographs by the following method:

• Aerial and Satellite photographs were analysed. Google Earth was selected as the source for the photographs as other aerial photographs were not readily available. While Google Earth covers the entire catchment, the resolution of all but 4 subareas – SC 6, SC13, SC25 and SC26 (all located in the lower lying north-east of the catchment) - was too poor to be able identify dams. Mersey is a mountainous catchment and as a result is frequently in cloud, which may explain the poor resolution of Google Earth photographs of the catchment. The Google Earth photos covering this catchment were taken between September 2002 and January 2007. The number of dams of any size was counted by eye in the 4 visible subareas. The number of unlicensed dams was determined by subtracting the number of licensed dams. A total of 116 unlicensed dams were identified in the 4 visible subareas. The ratio of unlicensed: licensed dams were calculated to be 0.41. This is similar to the nearby Leven & Gawler (0.49) and Panatana (0.41) catchments. This ratio was used to estimate the number of unlicensed dams in subareas that were obscured by multiplying the number of licensed dams by the ratio. This method led to an estimate of 1431 unlicensed dams over 31 subareas in the catchment;

• It was assumed most of these dams would be legally unlicensed dams (less than 1 ML and not situated on a water course). However, it was assumed that there would be a proportion of illegal unlicensed dams up to 20ML in capacity. A frequency distribution of farm dam sizes presented by Neal et al (2002) for the Marne River Catchment in South Australia showed that the average dam capacity for dams less than 20 ML was 1.4 ML (Table 3-6), and this dam size was adopted after discussion with DPIW;

• Following discussions with DPIW, the unlicensed dam demand was assumed to be 100%. The assumption is that all unlicensed dams will be empty at the start of May and will fill over the winter months, reaching 100% capacity by the end of September;

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

• Difficulties in detecting farm dams from aerial photography by eye are compounded when photography is not of suitably high resolution. Depending on the season and time of day that an aerial photograph is taken, farm dams can appear clearly or blend into the surrounding landscape. Vegetation can obscure the presence of a dam, and isolated stands of vegetation can appear as a farm dam when in fact no such dam exists. On balance, however, it was assumed that the number of false detections is countered by the number of missed detections, and in the absence of another suitably rapid method the approach gives acceptable results;

• Assuming this dam size distribution is similar to the distribution given in Neale et al.’s (2002) study catchment in South Australia, the total volume of unlicensed dams in the Mersey catchment is estimated to be 2003.4 ML (1431 * 1.4ML). This equates to approximately 2 ML/km 2 of unlicensed dams in the 31 subareas where dams are present, or 1.1 ML/km 2 over the entire catchment. The total volume of existing permitted dam extractions in the study catchment is 8322 ML. Therefore the volume of unlicensed dams equates to approximately 24 % of the total dam extractions from the catchment.

Table 3-6 Average capacity for dams less than 20 ML from Neal et al. (2002)

Average Total Size Range Volume Number of Volume (ML) (ML) Dams (ML) 0 - 0.5 0.25 126 31.5 0.5 - 2 1.25 79 98.75 2 - 5 3.5 13 45.5 5 - 10 7.5 7 52.5 10 - 20 15 6 90 27.5 231 318.25 Average Dam Volume: 1.4 ML

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

3.8 Environmental flows Scenario 3 was to account for environmental flows within the catchment. DPIW advised that for most of the Mersey catchment they currently do not have environmental flow requirements defined. The exception is flow from Lake Parangana, which conforms to a mandated environmental flow release: under usual conditions, a minimum 172.8 ML/day flows into the Mersey River at Liena through the mini hydro installation or the riparian valve at the Parangana Dam (this flow may be reduced when natural inflow into Lake Parangana is less than 172.8 ML/day). After discussions with DPIW, it was assumed the 172.8 ML/day environmental flow would be passed on to all subareas downstream of Lake Parangana. To account for environmental flows to those subareas not downstream of Lake Parangana, the calibrated catchment model was run under scenario 1 (Modelled – no entitlements (Natural)) and the environmental flows in these subareas were assumed to be:

• The 20 th percentile of flows for each sub-catchment during the winter period (01 May to 31 st Oct).

• The 30 th percentile of flows for each sub-catchment during the summer period (01 Nov – 30 April).

The Modelled – no entitlements (Natural) scenario was run from 01/01/1900 to 01/01/2006.

A summary table of the monthly environmental flows by sub-catchment is provided below in Table 3-5 and in the Catchment User Interface.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 3-7 Environmental Flows

Sub - Environmental Flow (ML/d) Per Month at each subcatchment area size Subarea (km 2) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ave. SC1 2 11.6 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC2 2 46.2 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC3 62.3 200.0 141.6 126.1 188.1 502.8 705.3 922.4 994.7 919.5 690.6 618.0 393.1 533.5 SC4 78.0 26.7 22.5 18.9 24.3 36.0 65.5 151.5 168.9 131.7 59.3 43.7 35.4 65.4 SC5 2 23.2 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC6 2 20.1 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC7 51.5 36.1 29.6 24.2 29.8 34.1 67.9 157.1 220.5 126.6 66.5 54.7 45.2 74.4 SC8 2 46.5 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC9 86.0 38.4 31.7 27.6 37.9 71.9 133.8 224.9 236.6 202.1 136.9 104.8 59.3 108.8 SC10 45.7 23.2 19.3 16.5 19.7 23.5 48.4 123.8 154.3 90.7 45.8 35.3 30.1 52.6 SC11 42.9 8.4 7.0 5.7 5.8 5.9 9.9 18.5 22.8 20.1 15.4 13.3 10.8 12.0 SC12 40.6 8.2 6.8 5.6 5.7 6.1 10.5 20.1 25.6 20.6 15.1 12.7 10.6 12.3 SC13 22.8 7.3 6.1 5.0 5.5 5.2 8.0 18.2 20.5 18.1 13.8 12.0 9.5 10.8 SC14 49.5 10.5 8.6 7.0 7.3 7.8 14.3 28.4 36.0 25.6 19.6 16.3 13.3 16.2 SC15 12.0 2.0 1.6 1.4 1.5 1.5 2.3 4.9 5.2 4.9 3.8 3.3 2.6 2.9 SC16 82.1 19.1 15.8 13.2 14.2 16.2 31.3 61.8 98.3 57.1 34.6 30.2 25.3 34.8 SC17 2 55.3 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC18 2 43.7 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC19 35.6 33.0 26.9 23.2 30.4 37.7 64.0 158.6 188.8 151.9 60.9 51.2 44.3 72.6 SC20 155.9 94.9 72.8 61.5 84.1 151.5 297.4 578.1 603.9 504.2 271.3 211.2 150.3 256.8 SC21 87.0 24.6 20.7 18.1 23.9 31.1 54.6 135.6 153.2 123.8 47.4 38.1 33.2 58.7 SC22 2 61.4 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC23 118.3 35.5 29.7 24.9 31.0 39.5 76.7 198.4 243.3 164.8 76.1 58.9 45.2 85.4 SC24 40.1 11.7 9.7 8.0 9.4 10.7 21.7 62.5 83.8 50.3 22.9 17.8 14.9 27.0 SC25 2 25.7 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC26 19.9 3.6 3.0 2.4 2.6 2.5 3.6 8.6 9.8 8.4 6.6 5.9 4.6 5.1 SC27 2 35.4 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 172.8 SC28 198.1 117.9 83.9 73.7 121.6 345.1 493.3 624.8 634.1 587.6 471.2 413.8 241.2 350.7 SC29 22.2 3.8 3.0 2.6 2.8 2.7 4.2 8.6 9.5 9.2 7.0 6.4 4.9 5.4 SC30 16.1 30.1 24.9 19.8 21.5 21.8 42.7 90.6 126.0 82.8 54.2 47.8 37.0 49.9 SC31 9.1 2.0 1.7 1.3 1.4 1.5 2.8 5.2 7.7 5.0 3.7 3.2 2.5 3.2 SC32 5.1 1.0 0.8 0.7 0.7 0.7 1.2 2.3 3.1 2.4 1.8 1.6 1.2 1.5 SC33 3.8 0.8 0.6 0.5 0.5 0.5 0.9 1.7 2.3 1.8 1.4 1.2 0.9 1.1 SC34 3.3 0.6 0.5 0.4 0.5 0.4 0.8 1.5 1.9 1.6 1.2 1.0 0.8 0.9 SC35 19.0 4.0 3.3 2.7 2.8 2.7 5.0 9.8 13.9 9.8 7.5 6.4 5.0 6.1 SC36 12.8 22.2 18.5 14.6 16.1 17.4 32.8 72.2 97.7 67.0 39.6 35.3 27.5 38.4

2 This subarea is downstream of Parangana Dam. The environmental flow requirement in this subarea was assumed to equal releases mandated for Parangana Dam.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Sub - Environmental Flow (ML/d) Per Month at each subcatchment area size Subarea (km 2) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ave. SC37 11.6 2.5 2.0 1.7 1.7 1.8 3.3 6.0 8.7 6.1 4.7 4.0 3.2 3.8 SC38 36.5 16.6 13.8 11.0 12.3 13.9 25.8 60.7 77.3 52.8 29.6 26.2 21.0 30.1 SC39 33.3 8.5 7.1 5.8 6.7 7.9 14.7 37.5 51.1 31.1 15.6 13.1 10.8 17.5 SC40 37.0 17.4 14.3 12.6 17.0 42.0 68.5 103.6 102.2 97.9 67.5 56.5 27.9 52.3 SC41 44.9 19.8 16.2 14.2 19.1 31.8 76.7 116.8 124.9 105.6 68.2 53.9 27.3 56.2

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

4. MODEL DEVELOPMENT

4.1 Catchment Subarea Delineation Subarea delineation was performed using CatchmentSIM GIS software.

CatchmentSIM is a freely available 3D-GIS topographic parameterisation and hydrologic analysis model. The model automatically delineates watershed and subarea boundaries, generalises geophysical parameters and provides in-depth analysis tools to examine and compare the hydrologic properties of subareas. The model also includes a flexible result export macro language to allow users to couple CatchmentSIM with any hydrologic modelling package that is based on subarea networks.

For the purpose of this project, CatchmentSIM was used to delineate the catchment, break it up into numerous subareas, determine their sizes and provide routing lengths between them.

These outputs were visually checked to ensure they accurately represented the catchment. For Mersey catchment several modifications were required. CatchmentSIM treats lakes differently when calculating routing lengths, and for the three subareas (SC4, SC3, SC2) containing the major lakes in this catchment (Lake Mackenzie, Lake Rowallan, Lake Parangana), the routing lengths were not sufficiently accurate. Routing lengths for these three subareas were estimated by manually tracing routing lengths in the GIS software package ArcMap, and then entered into the model.

For more detailed information on CatchmentSIM see the CatchmentSIM Homepage www.toolkit.net.au/catchsim/

4.2 Hydstra Model A computer simulation model was developed using Hydstra Modelling. The Mersey River subareas, described in Figure 2-1, were represented by model “nodes” and connected together by “links”. A schematic of this model is displayed in Figure 4-1. The flow is routed between each sub-area and through the catchment via a channel routing function.

Rainfall and evaporation were calculated for each subarea using inverse-distance gauge weighting. The gauge weights were automatically calculated at the start of each model run. The weighting is computed for the centroid of the subarea. Subareas were divided into quadrants separated by four radial lines emanating from the centroid (i.e., like slicing a pie into four pieces). A weight for the closest gauge in each quadrant was

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

computed as the inverse of the square of the distance between the gauge and centroid. For each time step and each node, the gauge weights are applied to the incoming rainfall and evaporation data.

The AWBM Two Tap rainfall/runoff model was used to calculate the runoff for each subarea separately. This means the Mersey model is a series of AWBM two tap models, all using a single set of calibration parameters: one AWBM model for each subarea. The output of a given subarea AWBM acts as the input for the subarea immediately downstream, and so on. This series of models was preferred over the usual method of using a single AWBM model for the whole catchment as it more accurately distributes runoff and base flow spatially over the catchment. The AWBM two-tap model was chosen over the more common single tap AWBM model for the whole catchment as it allows better simulation of base flow recessions.

The flow is routed between each subarea through the catchment via a channel routing function.

4.2.1 Accounting for Mersey Hydro Electric Scheme The upper Mersey catchment has undergone extensive modification for the hydro- electric power generation scheme, which needed to be accounted for in the model. There are three major dams in the Mersey catchment: Lake Mackenzie Dam (SC4), Lake Rowallan Dam (SC3), and Lake Parangana Dam (SC2), all built between 1967 and 1972 (see Figure 4-1). Each of the storages significantly alters flow to the subareas downstream. Of particular note is the Parangana dam, which supplies a tunnel that diverts substantial flow from the Mersey to the Forth River; flow volumes in the Mersey River are not conserved at the Parangana subcatchment boundary.

Following discussion with DPIW, these alterations to natural stream flow were accounted for by effectively splitting the Model at each dam. Mean daily inflows into subareas downstream of storages were calculated for each month from Hydro Tasmania records for the period 1997-2007 for Mackenzie and Lake Rowallan dams and for 2002-2007 for Lake Parangana (Table A-1). A shorter period was used for Parangana, as mandatory environmental flows have been released downstream since 2002. Before 2002 the Lake Parangana spilled only during large flood events. Environmental flows from Lake Parangana are mandated for the foreseeable future, so the shorter record better reflects future flows into the Mersey from Parangana, despite its brevity. When the model is run under scenarios 2 or 3 (i.e. not under the “natural” scenario) mean inflows are used as inputs into the subareas directly downstream of the

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

dams. Customised code was entered into the outflow node of the relevant subcatchments. The basic rules associated with this code are:

- Scenario 1, “No Entitlements (Defines ‘Natural’ flows)” will model the catchment with no dam or lake present for all of record.

- Both Scenario 2 “with Entitlements (extraction not limited by Env.Flows)” and Scenario 3 “Environmental Flows & Entitlements (‘Low Priority Ents. Limited by Env Flows’)” will model the catchment with:

1. No dam or lake present in the model prior to its construction completion date.

2. For all years following the completion date, flows downstream of the dam will be a total of the average long term (10 year) monthly flows which will include spill, power station discharge (if applicable) and any known environmental releases.

Because long-term monthly average flows were only calculated from records up to 10 years old, care should be taken if using the model for hindcasting flows before 1997.

The remainder of the model operates as usual, with the modelled outflows from any given subarea acting as inflows to the subarea immediately downstream. The details of individual lakes in the scheme and how they are treated in the model is described in APPENDIX A.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Figure 4-1 Hydstra Model schematic

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

4.3 AWBM Model The AWBM Two Tap model (Parkyn & Wilson 1997) is a relatively simple water balance model with the following characteristics:

• it has few parameters to fit;

• the model representation is easily understood in terms of the outflow hydrograph;

• the parameters of the model can largely be determined by analysis of the outflow hydrograph;

• the model accounts for partial area rainfall-run-off effects;

• runoff volume is relatively insensitive to the model parameters.

For these reasons parameters can more easily be transferred to ungauged catchments.

The AWBM routine used in this study is the Boughton Revised AWBM model (Boughton & Chiew 2003). Boughton & Chiew (2003) showed that when using the AWBM model the total amount of runoff is mainly affected by the average surface storage capacity and much less by how that average is spread among the three surface capacities and their partial areas. Given an average surface storage capacity (CapAve), the three partial areas and the three surface storage capacities are defined in Table 4-1.

Table 4-1 Boughton & Chiew, AWBM surface storage parameters

Partial area of S1 A1=0.134

Partial area of S2 A2=0.433

Partial area of S3 A3=0.433

Capacity of S 1 C1=(0.01*Ave/A 1)=0.075*Ave

Capacity of S 2 C2=(0.33*Ave/ A 2)=0.762*Ave

Capacity of S 3 C3=(066*Ave/ A 3)=1.524*Ave

The AWBM routine produces two outputs: direct run-off and base-flow. Direct run-off is produced after the content of any of the soil stores is exceeded; it can be applied to the stream network directly or by catchment routing across each subarea. Base-flow is

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

usually supplied unrouted directly to the stream network, at a rate proportional to the water depth in the ground water store. The ground water store is recharged from a proportion of excess rainfall from the three surface soil storages.

Although the AWBM accounts for base-flow, it is not intended that the AWBM be used to predict base-flow contribution within catchments. Base-flow in the AWBM routine is used as a fit parameter to obtain a good recession of surface water hydrographs. The AWBM does not specifically account for attributes that affect baseflow such as geology and inter- catchment groundwater transfers.

The AWBM processes are shown below in Figure 4-2.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Figure 4-2 Two-tap Australian Water Balance Model schematic

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

4.3.1 Channel Routing A common method employed in nonlinear routing models is a power function storage relation.

S = K.Q n

K is a dimensional empirical coefficient, the reach lag (time). In the case of Hydstra/TSM Modelling:

K = α.L i

where

Li = Channel length (km)

α = Channel Lag Parameter

n = Non-linearity Parameter

Q = Outflow from Channel Reach (ML/day)

A reach length factor may be used in the declaration of α to account for varying reach lag for individual channel reaches. e.g. α.fl where fl is a length factor.

Parameters required by Hydstra/TSM Modelling and their recommended bounds are given in Table 4-2.

Table 4-2 Hydstra/TSM Modelling Parameter Bounds

α Channel Lag Parameter Between 0.0 and 5.0

L Channel Length (km) Greater than 0.0 (km)

N Non-linearity Parameter Between 0.0 and 1.0

4.4 Model Calibration

4.4.1 Accounting for flow diversions on the Mersey River The choice of calibration period was dictated by the reliability of rainfall and evaporation input data. As noted, these data may not be reliable before 1970 (see Section 3.4). Thus a more recent calibration period was appropriate for this model. This was

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

complicated by the extensive alterations to the Mersey flow regime by the development of the Hydro electric power scheme for the Mersey, the first part of which commenced in June 1967 (see Figure 4-1). Of particular note is the construction of the Parangana Dam and tunnel, which diverts a large amount of flow from the Mersey River to the Forth River, meaning flow volume of the Mersey is not conserved after this dam.

In order to calibrate the model, the model was split at the node located at the settlement of Liena, which is located downstream of Parangana dam. The Mersey at Liena site (TSM 60.1/100.00/1) was chosen as it is the nearest site downstream from the Parangana Dam with a continuous and reliable flow record. For calibration purposes the Mersey at Liena flow record was used as the inflow to the Mersey at Liena node rather than the inflows generated by the model. There are no major flow diversions on the Mersey downstream of Liena. Using the Mersey at Liena flow record as an input into the Model circumvents the need to account for altered flows due to hydropower infrastructure. The site chosen for calibration was Mersey at Kimberley, which offered a reliable and lengthy flow record. The pickup between Mersey at Liena and the calibration site is 676 km 2, which is significantly large for the model to simulate stream routing accurately. Note that this modification to the model was only used during the calibration process. The final model uses long-term monthly downstream discharges from Parangana Dam, as detailed in APPENDIX A.

4.4.2 Calibration Method Calibration was achieved by adjusting catchment parameters so that the modelled flows best replicated the flows observed at the Mersey at Kimberley gauging station (site 22) over 20 years of available flow data (01/01/1986 to 01/01/2006). This period was chosen as there were few interruptions to the flow record in this time, and the rainfall DRILL data matched observed rainfall records for this period (see Section 3.4).

The fit of parameters were chosen by first comparing the volumes of monthly and annual flows over the entire calibration period, and second by comparing annual hydrographs. Regression statistics and engineering judgment were employed when observing daily and monthly time series comparisons.

The calibration process can best be understood as attempting to match the modeled calibration flow (MCF) to the observed flow record. The MCF can be described as:

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

MCF = MNEM - (WE x TPRF)

Where: MCF = Modeled Calibration Flow MNEM = Modeled - No Entitlements (Modified) WE = Water Entitlements TPRF = Time Period Reduction Factor

Water entitlements were included in the calibration model and adjusted to the time period of calibration by applied a Time Period Reduction Factor (TPRF). The TPRF was calculated by a method developed in the Tasmanian State of the Environment report: water demand has increased by an average of 6% annually over the last 4 decades. A 6% annual reduction from 2006 to the middle year of the calibration period, 1996, resulted in a TPRF of 53.9% of the current extractions was applied to all years in the calibration period. For the Mersey River the water entitlement extractions at the calibration site are negligible in relation to the observed flow, thus the model calibration would be unchanged regardless of the TPRF applied.

The model was calibrated to the observed flow as stated in the formula MCF = MNEM - (WE x TPRF). Other options of calibration were considered, including adding the water entitlements to the observed flow. However, the chosen method is considered to be the better option as it preserves the observed flow and unknown quantities are not added to the observed record. The chosen method also preserves the low flow end of the calibration, as it does not assume that all water entitlements can be met at any time.

In the absence of information on daily patterns of extraction, the model assumes that water entitlements are extracted at a constant daily flow for each month. For each daily time step of the model if water entitlements cannot be met, the modelled outflows are restricted to a minimum value of zero and the remaining water required to meet the entitlement is lost. Therefore the MCF takes account of very low flow periods where the water entitlements demand can not be met by the flow in the catchment. Table 4-4 shows the total catchment monthly water entitlements (demand) used in the calibration.

4.4.3 Adopted Model Parameters The adopted calibrated model parameters are shown in Table 4-3. These calibration parameters are adopted for all three scenarios in the user interface. Although it is probable that some catchment characteristics such as land use and vegetation will have changed over time, it is assumed that the rainfall run-off response defined by the calibration parameters has not changed significantly over time. Therefore it is appropriate to apply these parameters to all three scenarios.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

To achieve a better fit of seasonal volumes, the normally constant store parameter CapAve has been made variable for each month. In order to avoid rapid changes in catchment characteristics between months, CapAves of consecutive months were smoothed. A CapAve of a given month was assumed to occur on the middle day of that month. It was assumed that daily CapAves occurring between consecutive monthly CapAves would fit to a straight line, and a CapAve for each day was calculated on this basis. The annual profile of CapAves for the Mersey catchment is shown in Figure 4-3.

Monthly CapAve values followed a generally smooth curve moving from lower values in summer to higher values in winter. The exception was a dip in CapAve in March. This may be explained by the effects of consecutive dry summer months, which could cause the catchment properties to change slightly before the onset of the wetter autumn and winter period.

Table 4-3 Adopted Calibration Parameters

PARAMETER VALUE INFBase 0.8 K1 0.99 K2 0.92 GWstoreSat 120 GWstoreMax 130 H_GW 50 EvapScaleF 1 Alpha 3 n 0.8 CapAve Jan 48 CapAve Feb 67 CapAve Mar 42 CapAve Apr 80 CapAve May 80 CapAve June 102 CapAve July 118 CapAve Aug 143 CapAve Sep 152 CapAve Oct 147 CapAve Nov 120 CapAve Dec 95

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

160

140 120

100

80 60 CapAve 40

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

Figure 4-3 Monthly Variation of CapAve Parameter

4.4.4 Model accuracy: Qualitative description In all plots the “Modelled Calibration Flow” series - representing the modelled output under scenario 2 – is compared to the observed flow at the calibration site.

Time series plots have been displayed for three representative years of the observed and modelled flow: a typical year (Figure 4-4), a low-inflow year (Figure 4-5) and a high-inflow year (Figure 4-6). The quality of fit for each annual plot is described in the caption text. The quality of fit was determined initially by visual inspection, and verified by calculating the coefficient of determination (R 2) for the observed and modelled flows shown in each figure. The catchment’s average precipitation data used in the model are also plotted to show the relative magnitudes of precipitation through the year. Note that the precipitation trace is plotted on an independent scale (marked on the right of each graph). The water extraction entitlements for the subcatchment upstream of the calibration site are small relative to the Mersey’s flow (Table 4-4).

The time series plots show consistent response to rainfall events and good fidelity between modelled and observed hydrographs (Figure 4-4 and Figure 4-5). The calibration method focused on matching flow volumes while matching hydrograph shapes was given lesser importance, and hence a highly accurate match between observed and modelled time series plots was not expected. Despite this hydrograph response was good.

The good fit of the annual time series plots is supported by the excellent match between observed and modelled average flow volumes. The monthly, seasonal and

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

annual volume balances for the whole period of calibration record are presented in Table 4-4, Figure 4-7 and Figure 4-8. The demand values shown represent the estimated monthly demands for the middle year of the calibration period (see section 4.4) for the subareas upstream of the calibration site. The calibration procedure generally ensures that the seasonal and annual volumes are preserved. In this instance average demand is much lower than the volume of flow in the river.

Precipitation Modelled Calibration Flow Observed

18000 60 2 16000 R = 0.84 50

14000 40

12000 30 20 10000 10 8000 0 6000 -10 Precipitation (mm) Daily Flow (ML/day) 4000 -20 2000 -30 0 -40 01/1993 04/1993 07/1993 10/1993 01/1994

Figure 4-4 Daily time series - typical year (ML/d). Good fit

Precipitation Modelled Calibration Flow Observed 9000 60 2 8000 R = 0.81 50

7000 40

6000 30 20 5000 10 4000 0 3000 -10 Precipitation(mm) DailyFlow (ML/day) 2000 -20 1000 -30 0 -40 01/1999 04/1999 07/1999 10/1999 01/2000

Figure 4-5 Daily time series – low inflow year (ML/d). Good fit.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Precipitation Modelled Calibration Flow Observed 25000 60 R2 = 0.88 50 20000 40 30 15000 20 10 10000 0 -10 Precipitation (mm) Daily Flow (ML/day) 5000 -20 -30 0 -40 01/2003 04/2003 07/2003 10/2003 01/2004

Figure 4-6 Daily time series comparison – high inflow year (ML/d). Good fit.

300000 Observed 2 Modelled Calibration Flow R = 0.95 250000

200000

150000

100000 Monthly Volume (ML)

50000

0 1986 1987 1988 1988 1990 1991 1992 1992 1994 1995 1996 1996 1998 1999 2000 2000 2002 2003 2004 2004 2006

Figure 4-7 Monthly Time Series comparison – Volume (ML)

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

3000 Observed

2500 Modelled Calibration Flow

Modelled - No Entitlements 2000 (natural) Demand x 10

1500

1000 Average Flow (ML/Day) 500

0 Jul Oct Apr Jun Jan Mar Nov May Feb Aug Sep Dec WINTER ANNUAL SUMMER

Figure 4-8 Long term average monthly, seasonal and annual flows

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 4-4 Long term average monthly, seasonal and annual comparisons

Long term Averages (ML/Day)

MONTH Observed Demand 4

January 461.06 466.55 487.05 20.53 February 350.23 359.27 379.47 20.55 March 267.06 258.14 278.04 20.66 April 421.60 425.85 446.62 20.74 May 748.65 735.61 749.30 13.24 June 1308.36 1309.79 1322.58 13.24 July 2343.74 2346.37 2358.92 13.24 August 2719.99 2705.29 2718.50 13.24 September 2632.28 2632.12 2645.25 13.24 October 2092.18 2098.69 2119.14 21.07 November 1096.43 1083.38 1103.80 20.55 December 563.62 552.03 572.18 20.21 WINTER 1974.20 1971.31 1985.62 14.55 SUMMER 526.66 524.20 544.53 20.54 ANNUAL 1250.43 1247.76 1265.07 17.54 WINTER from May to Oct, SUMMER from Nov - Apr.

4.4.5 Factors affecting the reliability of the model calibration Regardless of the effort undertaken to prepare and calibrate a model, there are always factors that will limit the accuracy of the output. Significant limitations inherent in the method of calibration were:

• The assumption that water entitlements are taken as a constant rate for each month is unlikely to be correct. Historical monthly extraction rates are probably far more variable. Factors affecting water extraction rates and quantities are too complex to be accurately represented in the model. Further, the Time Period

3 Refer to Section 4.4.2 for explanation of this modelling scenario.

4 A TPRF was calculated for the middle year of calibration period (1996) to be 53.9% of WIMS (Dec 2006) demand. Demand is defined as total extractions upstream of the calibration site.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Reduction Factor applied to extraction entitlements is merely an estimate, and its accuracy is unknown;

• The quantity of water currently extracted from the catchment is not accurately known. Although DPIW has provided water extraction licence information (WIMS Dec 2006) and estimates of extractions in excess of these licenses, these may not represent the true quantity of water extracted. The method of estimating the volume of unlicensed dams, while the best available to this project, is crude. No comprehensive continuous water use data are currently available;

• Catchment precipitation and especially evaporation data used in the models may not always be accurate. This is due to insufficient rainfall gauge information in and around the catchment. Despite the Data DRILL’s good coverage of grid locations, the development of this grid information still relies considerably on the availability of measured rainfall information in the region. This is also the case with evaporation data, which will have a smaller impact on the calibration;

• Catchment freezing and snowmelt affect the upper Mersey catchment, especially during winter months. This may affect the flow regime. Snowmelt has not been specifically accounted for within this model;

• The daily timestep on which the model operates is likely to smooth out rainfall temporal patterns and affect peak flows. For example, an intense rainfall event in which significant rain falls in an hour is treated as if the same quantity of rain had fallen over 24 hours. Such intense rains can generate substantially more runoff than events where the same quantity of rain falls evenly over 24 hours. Thus extreme caution is advised in interpreting modelled flood peaks. The model is designed to predict longer term flow volumes, and not to accurately predict peak flows;

• The model does not explicitly account for changes in vegetation and terrain within individual subareas. Hydrologic effects due to vegetation and terrain are averaged across the whole catchment using the global AWBM fit parameters. Runoff in individual subareas may not be accurately represented by this model. To account for such variation a more complex model is required, the development of which is outside the scope of this project;

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

• The simple operating rules and assumptions used to model the catchment modifications (the Hydro-Electric Power generation scheme) cannot capture the complexities of operation that occur in reality.

4.4.6 Model Accuracy – Model fit statistics

Coefficients of Determination (R 2)

One of the most common measures of comparison between two sets of data is the coefficient of determination (R 2). If two data sets are defined as x and y, R 2 is the variance in y attributable to the variance in x. A high R 2 value indicates that x and y vary together – that is, as one data set changes, the other changes too. In this case x and y are observed flows and modelled flows. So for the Mersey catchment model, R 2 indicates how much modelled flows change as observed flows change. Table 4-5 shows the R 2 values between observed and modelled daily and monthly flows, as well as the proportional difference (%) between long-term (20 year) observed and modelled volumes. The high daily and monthly R2 values returned in conjunction with the small proportional difference between flow volumes show that the Mersey model is effective at simulating observed flows.

Table 4-5 Model Fit Statistics – Mersey River

Measure of Fit Value

Daily coefficient of determination (R 2 value) 0.82

Monthly coefficient of determination (R 2 value) 0.95

Difference in observed and estimated long term - 0.21 % annual average flows

As noted, the focus of the calibration process was to accurately simulate monthly flow volumes over a long period (20 years). Matching daily flows was given less priority. However, without a reasonable simulation of daily flows, a good monthly match would be difficult to achieve. A target of R 2 ≥ 0.7 was set for daily flows, while a target of R2 ≥ 0.85 was set for monthly flows. The lower target for daily flows was deemed acceptable due to model limitations and potential sources of error (see Section 4.4.7).

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Qualitative descriptions of fit were more formally defined by R 2 values as shown in Table 4-6.

Table 4-6 Coefficient of Determination (R 2) Fit Categories

Qualitative Fit Description Daily R 2 Monthly R 2

Poor R2 < 0.65 R2 < 0.8

Fair 0.65 ≥ R 2 > 0.70 0.8 ≥ R 2 > 0.85

Good R2 ≥ 0.70 R2 ≥ 0.85

It should be noted that although R 2 is a useful objective indicator of fit, it has limitations. One of the major limitations in using R 2 is that minor differences in the timing of hydrograph events can significantly affect the R 2 value. That is, even though the good visual fit achieved with the Mersey model was reflected in high R 2 values, this will not always be the case. Thus R 2 values are an aid to the more subjective practice of visually calibrating models, but not a substitute.

Proportional difference (%)

An alternative indicator of the reliability of a calibration is the proportional difference between observed data and the modelled flow measured by percent (%). Undertaking this analysis for the Mersey at Kimberley calibration site and producing meaningful proportional difference results was problematic. The reasons are discussed below:

• The modelled calibration flow (MCF) data includes observed data from the Mersey at Liena site (refer to section 4.4.1). Therefore the MCF will contain a percentage of “actual” observed flow, thus influencing the proportional difference results.

• The model used in the user interface utilises monthly long term averages (refer to section 4.2.1) as an input of flow downstream of lake Parangana, so again this will influence the proportional difference results.

In the absence of a viable alternative methodology for comparing the proportional difference the results from both models have been included below for comparison.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Modelled Calibration Flow (MCF) Results.

The proportional differences for the daily flows and monthly volumes were calculated for the calibration period and are presented as duration curves in Figure 4-9 and Figure 4-10. The graphs show the proportion of time for which the difference between observed and MCF flows is less than a given value. For example, the All Record trace in Figure 4-9 shows that for 50 % of the calibration period the difference between observed and MCF daily flows is 21 % or less. Similarly the All Record trace in Figure 4-10 shows that the difference between observed and MCF monthly flows is less than 17.5 % for 60 % of the 20 year calibration period. The duration curves show three traces, Summer 5, Winter 6 and All of Record . The lower values of the Summer trace are an artefact of the proportionally higher influence of the Mersey at Liena “observed” flows in the modelled data. In winter the influence of the Mersey at Liena “observed” data in the MCF is reduced as there is a higher contribution of modelled flow from the catchments downstream of Lake Parangana.

All record Winter Summer 200 180 160 140 120 100 80

(%) - Observed vs Modelled vs (%) - Observed 60 40

Difference 20 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Calibration period

Figure 4-9 Duration curve – MCF Daily flow proportional difference

5 Summer period = Nov – April, inclusive

6 Winter period = May – Oct, inclusive

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

100 All record Winter Summer 90

40 (%)Modelled - Observedvs 30

20 Difference

0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Calibration Period

Figure 4-10 Duration curve - MCF Monthly volume proportional difference

Overall the MCF model results show a very good result and this is because approximately half of the model is being replaced with observed “actual” flow from the Mersey at Liena site.

User Interface Model (UIM) Results.

The proportional differences for the daily flows and monthly volumes were calculated for the calibration period and are presented as duration curves in Figure 4-9 and Figure 4-10. The graphs show the proportion of time for which the difference between observed and User Interface Model (UIM) flow is less than a given value. For example, the All Record trace in Figure 4-11 shows that for 50 % of the calibration period the difference between observed and UIM daily flows is 40 % or less. Similarly the All Record trace in Figure 4-12 shows that the difference between observed and UIM monthly flows is less than 40 % for 60 % of the 20 year calibration period. The duration curves show three traces, Summer, Winter and All of Record . The higher values of the Winter trace are an artefact of monthly long-term means used as inputs to account for hydro-electric power generation infrastructure (see APPENDIX A). Essentially, the flows downstream of hydro-power infrastructure (Lake Parangana) are treated by the model as constant for each month, while actual flows would be far more variable.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Because the long-term mean flows are proportionally much higher in winter, the differences between actual flows and modelled flows will be greater, causing the higher differences in Winter.

All record Winter Summer 200 180 160 140 120 100 80

(%) - Observed vs Modelled vs (%) - Observed 60 40

Difference 20 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Calibration period

Figure 4-11 Duration curve – UIM Daily flow proportional difference

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

200 All record Winter Summer 180

160

140

120

100

80 (%) -Modelled Observed vs 60

40 Difference

0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of Calibration Period

Figure 4-12 Duration curve – UIM Monthly volume proportional difference

Overall the UIM model results show a poorer result than the MCF model and this is because approximately half of the model is being replaced with a coarser long term monthly average.

4.4.7 Model Accuracy throughout the Mersey catchment The model has been calibrated to provide a good simulation of monthly and seasonal flow volumes at the calibration site. Calibration sites are typically selected low in the catchment to encompass as much of the catchment as possible. It is difficult to assess how reliably the model performs throughout the catchment, although it is assumed that the model operates satisfactorily at other sites in the catchments. The ability of five other Tascatch models (developed by the same method as the Mersey Model) to simulate flows throughout these catchments was assessed. These assessments are detailed in APPENDIX B. These analyses suggest that on average the models predict volumes well throughout their catchments (see APPENDIX B).

The fit of the hydrograph shape (which plots daily flows) is expected to vary between sites within the catchment. Therefore it is expected that hydrograph fit will deteriorate as the catchment area decreases.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

In the Mersey Catchment there are two gauging sites that can be used to assess the calibration fit at alternative locations: the Arm River before it flows into the Mersey (Arm above Mersey – TSM 624.1), and the outflow of the Don River (Don River upstream of Bass Highway – TSM 16200.1). Plots of the monthly time series volumes and corresponding R 2 values are shown in Figure 4-13 and Figure 4-14. The model performed very well at these sites, and thus it is assumed the model will perform with reasonable accuracy throughout the Mersey catchment.

30000 Observed Modelled Calibration Flow R2 = 0.95 25000

20000

15000

10000 MonthlyVolume (ML)

5000

0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Figure 4-13 Time series of Monthly Volumes – Arm above Mersey (TSM 624.1) (SC9)

25000 Observed 2 Modelled Calibration Flow R = 0.95

20000

15000

10000 Monthly Volume (ML) 5000

0 1980 1980 1981 1982 1984 1984 1985 1986 1988 1988 1989 1990

Figure 4-14 Time series of Monthly Volumes – Don River upstream of Bass Highway (TSM 16200.1) (SC30)

4.5 Model results The completed model and user interface allows data for three catchment demand

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

scenarios to be generated:

• Scenario 1 – No Entitlements (Natural Flow);

• Scenario 2 - With Entitlements (with water entitlements extracted);

• Scenario 3 - Environmental Flows and Entitlements (Water entitlements extracted, however low priority entitlements are limited by an environmental flow threshold).

For each of the three scenarios, daily flow sequence, daily flow duration curves, and indices of hydrological disturbance can be produced for any subarea.

For information on the use of the user interface refer to the Operating Manual for the NAP Region Hydrological Models (Hydro Tasmania 2004b).

Outputs of daily flow duration curves and indices of hydrological disturbance at the model calibration site are presented in Figure 4-15, Table 4-7 and in Section 4.5.1 below. The outputs are a comparison of scenario 1 with scenario 3 over the period 01/01/1900 to 01/01/2006. Note that this catchment has been extensively modified by the hydro- electric power generation scheme, and the influence of these modifications may overwhelm any effects that water extraction entitlements may have on these results.

100000.00 Natural

10000.00 Entitlements Extracted 1000.00

100.00

Flow (ML/d) 10.00

1.00

0.10 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent Of Time Exceeded

Figure 4-15 Daily Duration Curve for Modelled flows 01/01/1900 – 01/01/2006 at the Calibration Site (SC6)

4.5.1 Indices of hydrological disturbance The calculation of the estimates of natural flow (scenario 1) and flow less water extractions but including environmental flows (scenario 3) were used to calculate indices of hydrological disturbance. These indices include:

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

• Hydrological Disturbance Index

• Index of Mean Annual Flow

• Index of Flow Duration Curve Difference

• Index of Seasonal Periodicity

• Index of Seasonal Amplitude

The indices were calculated using the formulas developed for the Natural Resource Management (NRM) Monitoring and Evaluation Framework by SKM for the Murray- Darling Basin (MDBC 08/04).

Table 4-7 shows the Hydrological Disturbance Indices (HDIs) at the Catchment outflow (SC1), comparing scenario 1 (natural) and scenario 3 (environmental flows with all extractions included) for period 01/01/1900 to 01/01/2006. Two sites in addition to the calibration site have been selected to give an indication of the variability of the indices of hydrological disturbance across the catchment. Note that hydrological disturbance is evident in the hydrological indices presented in Table 4-7. This is due to the difference in flow regime between natural flows and flows affected by the hydro-electric power generation scheme in the upper Mersey catchment.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table 4-7 Hydrological Disturbance Indices at the Catchment Outflow measuring disturbance between Scenario 1 and Scenario 3 at 3 sites in the Mersey Catchment

Disturbance Indices Undisturbed SC3 (High SC6 SC1 in the (Calibration (catchment (natural flow) catchment) site) outflow)

Index of Mean Annual Flow, A 1.00 1.00 0.81 0.81

Index of Flow Duration Curve Difference, M 1.00 1.00 0.78 0.78

Index of Seasonal Amplitude, SA 1.00 1.00 0.82 0.82

Index of Seasonal Periodicity, SP 1.00 1.00 0.92 0.92

Hydrological Disturbance Index, HDI 1.00 1.00 0.82 0.82

Hydrological Disturbance Index (HDI): This provides an indication of the hydrological disturbance to the river’s natural flow regime. A value of 1 represents no hydrological disturbance, while a value approaching 0 represents extreme hydrological disturbance.

Index of Mean Annual Flow: This provides a measure of the difference in total flow volume between current and natural conditions. It is calculated as the ratio of the current and natural mean annual flow volumes and assumes that increases and reductions in mean annual flow have equivalent impacts on habitat condition.

Index of Flow Duration Curve Difference: The difference from 1 of the proportional flow deviation. Annual flow duration curves are derived from monthly data, with the index being calculated over 100 percentile points. A measure of the overall difference between current and natural monthly flow duration curves. All flow diverted would give a score of 0.

Index of Seasonal Amplitude: This index compares the difference in magnitude between the yearly high and low flow events under current and natural conditions. It is

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

defined as the average of two current to natural ratios. Firstly, that of the highest monthly flows, and secondly, that of the lowest monthly flows based on calendar month means.

Index of Seasonal Periodicity: This is a measure of the shift in the maximum flow month and the minimum flow month between natural and current conditions. The flows of the month with the highest mean monthly flow and the flows of the month with the lowest mean monthly flow are calculated for both current and natural conditions. Then the absolute difference between the maximum flow months and the minimum flow months are calculated. The sum of these two values is then divided by the number of months in a year to get a monthly proportion (measured in %). This proportion is then subtracted from 1 to give a value range between 0 and 1. For example a shift of 12 months would have an index of zero, a shift of 6 months would have an index of 0.5 and no shift would have an index of 1.

4.6 Flood frequency analysis No flood frequency plot has been developed for this model as the river is highly regulated by the hydro–electric generation scheme, which affects the calibration site and the lower catchment where the majority of water entitlements are located.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

5. REFERENCES

Boughton, W.C. and Chiew, F., (2003) Calibrations of the AWBM for use on Ungauged Catchments

CatchmentSIM Homepage www.toolkit.net.au/catchsim/ , December 2006

QNRM Silo (Drill Data) Homepage www.nrm.qld.gov.au/silo , January 2005

SKM (2003) Estimating Available Water in Catchments in Catchments Using Sustainable Diversion Limits. Farm Dam Surface Area and Volume relationship, report to DSE, Draft B October 2003

Hydrology Theme Summary of Pilot Audit Technical Report – Sustainable Rivers Audit. MDBC Publication 08/04.

National Land and Water Resources Audit (NLWRA) www.audit.ea.gov.au/anra/water/ ; January 2005.

Hydro Tasmania (2004a) South Esk River Catchment Above Macquarie River, Impact of Water Entitlements on Water and Hydro Power Yield.

Hydro Tasmania (2004b). Operating Manual for the NAP region Hydrological Models. Hydro Report 118783 – Report -015, 17 September 2004.

Hydro Tasmania, (2005), NAP Region Hydrological Model, North Esk Catchment.

Neal B, Nathan RJ, Schreider S, & Jakeman AJ. 2002, Identifying the separate impact of farm dams and land use changes on catchment yield. Aust J of Water Resources, IEAust, 5(2):165-176.

Parkyn R, Wilson D, (1997) Paper: Real-Time Modelling of the Tributary Inflows to ECNZ's Waikato Storages. Published in 24th Hydrology & Water Resources Symposium Proceedings Auckland NZ 1997.

State of the Environment Report, Tasmania, Volume 1 Conditions & Trends 1996. State of Environment Unit, Lands Information Services, DELM.

SKM (2005) Development and Application of a Flow Stress Ranking Procedure, report to Department of Sustainability and Environment, Victoria.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

5.1 Personal Communications DPIW (2007) Bryce Graham, Section Head, Ecohydrology, Water Assessment. March- May 2007.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

6. GLOSSARY

Coefficient of determination (R 2): One of the most common measures of comparison between two sets of data is the coefficient of determination (R 2). If two data sets are defined as x and y, R 2 is the variance in y attributable to the variance in x. A high R 2 value indicates that x and y vary together – that is, the two data sets have a good correlation

High priority entitlements: Water entitlements with an assigned Surety 1 to 3.

Low priority entitlements: Water entitlements with an assigned Surety 4 to 8.

Modelled – No entitlements (Natural) : The TimeStudio surface water model run in a natural state. That is, all references to water entitlements have been set to zero. Additionally any man made structures such as dams, power stations and diversions have been omitted and the modelled flow is routed, uncontrolled through the catchment. This is also referred to as Scenario 1.

Modelled – No entitlements (Modified) : The TimeStudio surface water model run with no water entitlements extracted. That is, all references to water entitlements have been set to zero. Where human structures are identified that significantly affect the flow regime, such as large dams, power stations and diversions, the TimeStudio model contains custom code to estimate the flow effect on the downstream subareas. This custom code takes effect from the completion date of the structure. Where there are no significant human structures in the catchment or the model is run before the completion of these structures this model will produce the same output as “Modelled – No entitlements (Natural)”. This option is not available within the user interface and is one of several inputs used to derive a modelled flow specifically for calibration purposes. It is also referred to as MNEM in Section 4.4.

Modelled – with entitlements (extracted): The TimeStudio surface water model with water entitlements removed from the catchment flow. Where human structures are identified within a catchment that significantly affect the flow regime, such as large dams, power stations and diversions, the TimeStudio model contains custom code to estimate the flow effect on the downstream sub-catchments. This custom code takes effect from the completion date of the structure. This is also referred to as Scenario 2.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Modelled – environmental flows and entitlements (extracted ): The TimeStudio surface water model with water entitlements removed. However, low priority entitlements are only removed when sub-catchment flow exceeds a specified environmental threshold. Where man made structures are identified within a catchment, such as dams, power stations and diversions the TimeStudio model contains code to estimate the flow effect on the downstream subcatchments, commencing on the completion date of the structure. This is also referred to as Scenario 3.

Time Period Reduction Factor (TPRF): A reduction factor applied to current levels of water extracted from a catchment. The TPRF was applied to satisfy the assumption that the amount of water extracted from Tasmanian catchments (e.g. for agriculture) has increased over time. The TPRF was calculated by a method developed in the Tasmanian State of the Environment report. This states that water demand has increased by an average of 6% annually over the last 4 decades. This factor is applied to current water entitlements to provide a simple estimate of water entitlements historically. However, following discussions with DPIW the TPRF was capped at 50% of the current extractions if the mid year of the calibration period was earlier than 1994.

Water entitlements: This refers generally to the potential water extraction from the catchment. Included are licensed extractions documented in WIMS (Dec 2006), estimates of additional unlicensed extractions and estimates of unlicensed farm dams. Unless specified otherwise, Hydro Tasmania dams and diversions are not included.

WIMS (Dec 2006): The Department Primary Industries and Water, Water Information Management System, updated to December 2006.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

APPENDIX A

Accounting for the Mersey Hydro Scheme in the Model

Lake Mackenzie

Lake Mackenzie and the Fisher Power station were deemed to have commenced operation in January 1973. The dam diverts the Fisher River through a canal and then through Fisher Power Station, discharging into Lake Parangana. Fisher Canal picks up a small (but unknown) fraction of runoff from SC20, aided by a number of artificial diversions. The only flows into Fisher River below the dam are from spill events and from pickup within subarea 20 (less any pickup diverted into Fisher Canal). Mean daily spill was calculated by month from the Lake Mackenzie Dam record (TSM 629.1/130.00/10) on the Hydro Tasmania data base (Table A-1). When the model is running either scenario 2 or 3 (i.e., not the natural flow scenario), the model uses these mean monthly spills as inflow to subarea 20 from January 1973 onward. A good flow record for Fisher Canal below Lake Mackenzie (TSM 630.1/100.00/1) was available from the Hydro Tasmania database. Daily Mean flows were calculated by month for the period 01/04/1997 – 01/07/2007. In the model these mean flows were fed directly into Lake Parangana (SC2) when running either scenario 2 or 3. This is likely to understate flow through Fisher Power Station as it neglects the small amount of pickup flowing into Fisher canal from SC20, but this was considered preferable to using the less reliable Fisher power station record. Conversely, the model directs pickup that would flow into Fisher Canal into Fisher River. This means that modelled flows into the Fisher River are likely to be slightly greater than observed flows. As the pickup diverted into Fisher Canal is small relative to the pickup from the remainder of the subarea, modelled flows are not likely to exceed actual flows greatly (notwithstanding the accuracy of the model calibration). When run in scenario 1 (natural flow), the model does not account for the existence of Lake Mackenzie or any other artificial diversions.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table A-1 Hydro Tasmania Sites used for long term mean model inputs (ML/day)

Hydro Infrastructure Fisher Canal/ Lake Mackenzie Lake Rowallan Lake Parangana accounted for in Powerstation model

Commencement Jan 1973 Jan 1973 June 1967 April 1969 date in model

FISHER CANAL LAKE MACKENZIE MERSEY RIVER Site Record [B/L MERSEY RIVER [AT LIENA] [AT DAM] [A/B ARM] L.MACKENZIE]

TSM TSM TSM Data Source TSM(60.1/100.00/1) (630.1/100.00/1) (629.1/130.00/10) (153.1/100.00/1)

01/04/1997 to 01/04/1997 to 01/04/1997 to Period 01/04/2002 to 01/04/2007 01/04/2007 01/04/2007 01/04/2007

SC8 Subarea No. SC2 SC20 SC2 Affected Area scaled Site record model input

January 124.6 0.0 779.3 182.5 155.7

February 110.9 0.0 777.6 169.8 139.0 March 59.6 0.0 493.9 119.4 107.6 April 84.4 0.0 723.7 148.2 131.1 May 251.2 64.4 927.7 185.0 136.2 June 353.2 57.1 1338.5 265.0 162.1 July 473.6 95.1 1828.8 1246.2 1061.9 August 435.8 122.1 1658.3 1086.2 916.5 September 493.8 207.8 2045.6 1810.9 1635.2 October 467.5 84.9 2034.6 603.8 487.5 November 340.7 12.0 1637.5 250.9 180.4 December 187.3 12.4 1004.0 170.5 130.3

Lake Rowallan

Lake Rowallan was deemed to commence in June 1967. A good flow record is available from Mersey above Arm (TSM 153.1/100.00/1). Mean Daily flows were calculated for each month from the period 01/04/1997 – 01/04/2007 (Table A-1). The model uses these means as inflows into SC2 from June 1967 onward when the model is run under scenario 2 or 3.

Lake Parangana

Lake Parangana dam and diversion were deemed to have commenced in April 1969. Mean daily inflows into the subarea downstream of Lake Parangana were calculated for each month over the period 01/04/2002 – 01/04/2007 from the Mersey at Liena (TSM 60.1/100.00/1) flow record on the Hydro Tasmania data base. The period 2002- 2007 was chosen for Parangana, as mandatory environmental flows have been released downstream since 2002. Before 2002 the Mersey only spilled over

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Parangana dam during large flood events. As environmental flows from Lake Parangana are mandated for the foreseeable future, the shorter record better reflects future flows into the Mersey from Parangana, despite its brevity. Mersey at Liena flows into SC22, rather than the subarea immediately downstream of Lake Parangana (SC8). Thus the pickup from SC8 had to be accounted for. The mean daily pickup was calculated by month for the period 01/04/2002 – 01/04/2007 and these values were subtracted from the inflows calculated from the Mersey at Liena record (Table A-1). The adjusted inflows were then used as inputs into SC8 when the model is run under scenario 2 or 3.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

APPENDIX B

This appendix investigates the reliability of Tascatch catchment models in predicting river flow throughout these catchments. One of the difficulties in assessing model reliability is the lack of observed data: there is often only one reliable gauging site within the catchment. Five Tascatch catchment models developed for catchments that have more than one gauging site were selected and investigated with the results presented in Table B-2. The analysis undertaken is outlined below.

• The relationship between catchment area of the calibration site (primary site) and the secondary site was determined. Good variability is represented within this selection, with the secondary site catchment area ranging between 6.6% and 41.5% of the calibration site.

• The catchment area relationship was used to derive a time series at the secondary site based on scaled observed data from the calibration site. This was used in subsequent analysis to assess the suggestion that an area scaled time series derived from a primary site was a good representation of sub- catchment flow in the absence of a secondary gauging site.

• For concurrent periods, estimated monthly volumes (ML) were extracted at both the calibration site and the secondary site.

• R2 values were calculated on the following data sets for concurrent periods:

o Correlation A: The correlation between the calibration site observed data and calibration site modelled data . This provides a baseline value at the calibration site for comparison against the other correlations.

o Correlation B: The correlation between the calibration site observed data (which has been reduced by area) and secondary site observed data . This shows the relationship of area scaled estimates as a predictor of sub-catchment flows, in this case by comparison with a secondary gauge.

o Correlation C: The correlation between the calibration site observed data (which has been reduced by area) and secondary site modelled data . This compares modelled data with an area scaled data set derived from observed data. This has been done because in the absence of a gauging site, observed data from another site is often

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

assumed as a good indication of flow within the sub-catchment (Correlation B addresses this assumption). Where this assumption is applied, this correlation provides a statistical comparison of the models ability to predict comparable volumes to that of an area scaled estimate.

o Correlation D: The correlation between the secondary site observed data and secondary site modelled data . This has been done to assess how well the calibration undertaken at the primary site directly translates to other subcatchments within the model.

The catchment model has been calibrated to provide a good fit for monthly and seasonal volumes at the calibration site. Calibration sites are typically selected low in the catchment to represent as much of the catchment as possible. Therefore the calibration fit parameters on average are expected to translate well to other subcatchments. However, where subcatchments vary significantly in terrain or vegetation or rainfall compared to the catchment average, errors are expected to be greater. The analyses undertaken in this section appears to confirm that the models perform acceptably and the conclusions drawn from these analyses are summarised below:

1. Four of the five catchments studied showed fair to good R 2 values between observed and modelled data at the secondary site. (Correlation D).

2. The George secondary site was the worst performing in the study with a fair R2 value of 0.83. It is expected that this is due to localised changes in terrain, vegetation and/or rainfall. This is a known limitation of the model and is therefore expected in some cases.

3. Scaling the calibration site observed data by area to derive a data set at another location is not recommended. Area scaled data does not consistently out perform the model at predicting flow/volumes within catchment. It is demonstrated that the model does (in the majority of cases) a good job of directly predicting the flow/volumes within catchment.

Time Series plots of the monthly volumes in Megalitres for the five catchments studied in this section are shown in Figure B-1 to Figure B-4. These plots show that generally the calibration fit at the primary site translates well as a direct model output at other locations within the catchment, when modelling monthly volumes.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Observed - Forth a/b Lemonthyme Site 450 140000 Site 450 - Modelled - with entitlements Observed- Scaled Forth at Paloona Bdg - site 386 120000

100000

80000

60000 Monthly Volume (ML) Volume Monthly 40000

20000

0 1963 1964 1964 1965 1966 1967 1968

Figure B-1 Forth catchment – monthly volumes at secondary site.

Observed - Ransom Rv Site 2217 5000 Site 2217 Modelled - with entitlements Observed - Scaled George at WS site 2205 4500

4000

3500

3000

2500

2000

Monthly Volume (ML) Volume Monthly 1500

1000

500

0 1983 1984 1987 1989

Figure B-2 George catchment – monthly volumes at secondary site.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Observed - Leven at Mayday Rd - Site 821 20000 Site 821 Modelled - with entitlements 18000 Observed- Scaled Leven at Bannons site 14207

16000

14000

12000

10000

8000

Monthly Volume (ML) Volume Monthly 6000

4000

2000

0 1983 1984 1987 1989 1991 1993

Figure B-3 Leven catchment – monthly volumes at secondary site.

Observed - Swan u/s Hardings F - Site 2219 Site 2219 Modelled - with entitlements 16000 Observed - Scaled Swan at Grange site 2200

14000

12000

10000

8000

6000 Monthly Volume (ML) Volume Monthly 4000

2000

0 1983 1984 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994

Figure B-4 Swan catchment – monthly volumes at secondary site.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Observed - Montagu at Togari - Site 14216 Site 14216 Modelled - with entitlements 20000 Observed- Scaled Monatgu at Montagu Rd Brg - Site 14200

18000

16000

14000

12000

10000

8000

Monthly Volume (ML) Volume Monthly 6000

4000

2000

0 1985 1986 1987 1988 1988 1989 1990

Figure B-5 Montagu catchment – monthly volumes at secondary site.

Mersey River Surface Water Model Hydro Tasmania Version No: 1.1

Table B-2 Tascatch Models’ performance at secondary sites

Catch- Calibration Site Secondary Site Correlation A Correlation B Correlation C Correlation D ment Primary Site

Name Site Name Sub- Catchment Concurrent Site Name Sub- Catchment Catchment Monthly ML Monthly ML Monthly ML Monthly ML

& No. Catchment Area data periods & No. Catchment Area area factor 2 2 2 2 Location used in this Location (compared R Value R Value R Value R Value Km2 Km2 analysis with Calibration site Secondary site Calibration site Secondary calibration observed vs observed vs observed(scale site observed site) Calibration site Calibration site d) vs Modelled vs Modelled modelled observed (scaled) Forth Forth at SC33 1079.6 01/01/1963 to Forth River SC31 310.2 0.2873 0.97 0.95 0.95 0.97 Paloona 01/03/1969 above Bridge – Lemonthym Site 386 e – site 450 George George SC2 397.9 01/03/1983 to Ransom Rv SC3 26.1 0.0656 0.91 0.96 0.86 0.83 River at SH 01/10/1990 at Sweet WS – Site Hill – Site 2205 2217 Leven Leven at SC4 496.4 01/04/1983 to Leven at SC6 37.5 0.0755 0.93 0.87 0.88 0.92 Bannons 01/09/1994 Mayday Rd Bridge – – site 821 Site14207 Swan Swan River SC20 465.9 01/07/1983 to Swan River SC4 35.6 0.0764 0.92 0.95 0.82 0.85 at Grange – 01/10/1996 u/s Site 2200 Hardings Falls – site 2219 Montagu Montagu at SC3 325.9 01/01/1985 to Montagu at SC2 135.4 0.4155 0.98 0.98 0.95 0.94 Montagu 01/01/1990 Togari – Rd Brdge – Site 14216 Site 14200