regional groundwater modelling of the cambrian … groundwater modelling of the cambrian limestone...

Regional Groundwater Modelling

of the Cambrian Limestone Aquifer System

of the Wiso Basin, Georgina Basin and Daly Basin

Report No.: 29/2006A

Anthony Knapton

Land and Water Division

Alice Springs

Groundwater Modelling of the Tindall Limestone

Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page i

Department of Natural Resources, Environment & The Arts

Technical Report No. 29/2006A

Regional Groundwater Modelling of the Cambrian Limestone Aquifer System

of the Wiso Basin, Georgina Basin and Daly Basin

A report prepared by NRETA Land and Water Division

Author: Anthony Knapton

Department of Natural Resources, Environment & The Arts, Alice Springs


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page ii

Department of Natural Resources, Environment & The Arts

Technical Report No. 29/2006A

Copyright

© 2006 Northern Territory Government Copyright resides with the Northern Territory Government, 2006. Information contained in this publication may be copied or reproduced for study, research, information, or educational purposes, subject to inclusion of an acknowledgment of the source.

Cover Image: Looking downstream along the Katherine River at Galloping Jacks.


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page iii

Executive Summary Background

The Cambrian Limestone aquifer system is the major aquifer in the Wiso, Georgina and Daly

Basins. It represents the source of the majority of the baseflow in the Katherine, Roper, Flora and

Douglas Rivers.

The impending development of nearby horticultural districts reliant on water from the Tidal

Limestone aquifer of the Daly Basin, represent a threat to the environmental flow regime of the

rivers particularly in low flow periods.

This report documents an ambitious attempt at modelling the entire Cambrian Limestone Aquifer

System, with an emphasis on the area surrounding Katherine.

Conceptual Model of the Tindall Limestone

The conceptual model was developed by the Water Resources section of NRETA from the

available groundwater and surface water data and observations. It can be summarized as:

• The Limestone aquifer may be represented as a single unconfined layer.

• Mapped occurrence of the Jinduckin Formation confines the Limestone aquifer and is expected

to have lower transmissivities and storage coefficient than the unconfined Limestone aquifer.

• The limestone aquifer was expected to have greatest permeability within the weathered zone,

confined to the upper 150 metres from the surface. For the purposes of this exercise the

aquifer was considered to have a constant thickness below the groundwater table. That is a

single layer of variable transmissivity was used instead of varying hydraulic conductivity and

aquifer thickness. This is considered valid as the variations in the groundwater level are

considered small compared to the saturated thickness of the Aquifer.

• Aquifer transmissivity of 5,000 m2/d based on the Water Studies modelling results.

• A single estimate of transmissivity in the Venn region indicate that lower values (around

2000 m2/d) may be applicable in this area.

• The confined regions of the Aquifer were assigned a single value of 100 m3/d/m.

• The unconfined aquifer storage coefficient was 0.04. This is considered a reasonable estimate

as previous experience (Jolly, pers comm.) indicates that this value should be between 0.01

and 0.07.

• Confined aquifer storage coefficient was assigned 0.0001 based on typical confined aquifer

storage coefficients.


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page iv

• The dominant recharge mechanism is via sinkholes, however, this is not well understood, so

the recharge was estimated as diffuse recharge.

• The main influence of the Cretaceous sediments is to reduce the recharge to the Tindall

Limestone aquifer. This is based on the subdued response of hydrographs for bores located in

areas with the Cretaceous cover (eg RN22006).

• Initial estimates of the steady state annual recharge over the two areas were 150 mm/yr

(0.41 mm/d) for the outcropping limestone and 40 mm/yr (0.11 mm/d) for the Cretaceous cover.

• Based on the steady state recharge estimates the potential recharge model developed using

the Katherine rainfall record (Jolly et al, 2000) was scaled to provide transient recharge values.

The recharge model estimates the soil moisture deficit and daily evapotranspiration to derive

the potential recharge due to deep drainage.

• Recharge occurs from the Katherine River when stage height > groundwater level.

• The dominant discharge from the aquifer is through the streambed and via springs. Discharge

occurs along the length of the Katherine River where it intersects the Aquifer.

• Over the long term the late dry season discharge to the river via spring flows range from 1 to 2

cumecs, with an average discharge rate of 1.33 cumecs, with approximately 17% or

0.23 cumecs discharging downstream of the low level weir.

• The Cretaceous rocks in the King River area where they overlie impermeable basement rocks

act as a constant flux boundary, which, during transient conditions with no stresses manifest as

a relatively constant head.

• Evapotranspiration from the riparian zone is estimated at approximately 3 mm/day. The ET

has not been explicitly considered in this model. Based on the ET value a riparian zone width

of approximately 200 metres and a length 12 kilometres of river the total ET is 720 m3/d.

Model Calibration Results

Based on the conceptual model developed, calibration of both the steady state model and the

transient model to the observation data was possible. Relatively good fits were obtained for both

the head and discharge data available.

Conclusions

Initial modeling of the Aquifer indicates that the conceptual model is largely valid:

• Initial modeling of the Limestone Aquifer indicates that the conceptual model is largely

valid, however;

• The mapped occurrence of the Limestone Aquifer in the area of the Flora River has been

identified as being incorrect. Initial re-examination of the geological information in the area


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page v

indicates that the mapped geology used in the modelling is incorrect this issue will be

resolved by re-examining the mapped geology in the area and with future investigation

drilling.

• The regional model has been developed to the level that provides the boundary conditions

for the more detailed modelling in the Katherine study area.

• The recharge distribution and the proportionate values proposed within this report are

indicated to be plausible. Improved definition of rainfall distribution over the basin will

enable refinement of this aspect of the model.

• The Cretaceous rocks in the King River area where they overlie impermeable basement

rocks act as a constant flux boundary, which, during transient conditions with no stresses

manifest as a relatively constant head.

• The modelling of the development Scenario A indicate the primary effect of groundwater

extraction in the Katherine area is to reduce flows in the Katherine River (refer Table 3).


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page vi

Contents

Executive Summary ........................................................................................................................... iii 1 Introduction ................................................................................................................................11

1.1 Background ........................................................................................................................11 1.2 Objectives...........................................................................................................................11 1.3 Model Steps........................................................................................................................11 1.4 Location..............................................................................................................................12 1.5 Climate ...............................................................................................................................14 1.6 Geomorphology..................................................................................................................15

1.6.1 Topography ................................................................................................................15 1.7 Previous Modelling ............................................................................................................16

2 Hydrogeological Setting and Conceptual Model .......................................................................18 2.1 Regional Geology...............................................................................................................18

2.1.1 Cretaceous Rocks of the Dunmarra Basin..................................................................18 2.1.2 Oolloo Dolostone .......................................................................................................18 2.1.3 Jinduckin Formation and Anthony Lagoon Beds.......................................................19 2.1.4 Tindall Limestone, Gum Ridge Formation and Montejinni Limestone.....................19 2.1.5 Antrim Plateau Volcanics...........................................................................................20 2.1.6 Groundwater Flow......................................................................................................21 2.1.7 Discharge from the Cambrian Aquifer System..........................................................22

2.2 Study Area Hydrogeology..................................................................................................23 2.2.1 Limestone Aquifer Saturated Thickness ....................................................................23 2.2.2 Hydraulics of the Cretaceous Sediments in the Area of the King River....................24 2.2.3 Groundwater Flow......................................................................................................24 2.2.4 Groundwater Discharge..............................................................................................24

2.3 Study Area Observation Data.............................................................................................25 2.3.1 Rainfall and Potential Recharge.................................................................................25 2.3.2 Observation Bores ......................................................................................................25 2.3.3 Groundwater Level Hydrographs...............................................................................26 2.3.4 Potentiometric Head Distribution...............................................................................27 2.3.5 River Stage Height Data.............................................................................................28 2.3.6 River Gauging Data....................................................................................................29

2.4 Conceptual Model ..............................................................................................................32 3 Model Development ...................................................................................................................34

3.1 Model Specifications..........................................................................................................34 3.1.1 Numerical Model Code ..............................................................................................34 3.1.2 Spatial Discretisation..................................................................................................35

3.2 Layers .................................................................................................................................36 3.3 Boundary Conditions..........................................................................................................36

3.3.1 Recharge (Specified Flux at the Model Surface) .......................................................37 3.3.2 Transfer (Cauchy) Boundary......................................................................................37 3.3.3 Well Boundary Conditions.........................................................................................38

3.4 Hydraulic Parameters .........................................................................................................38 3.4.1 Transmissivity Distribution........................................................................................39 3.4.2 Recharge Distribution ................................................................................................40

4 Steady State Model Development ..............................................................................................42 4.1 Steady State Model Calibration..........................................................................................42 4.2 Steady State Calibration Results ........................................................................................43

4.2.1 Calibrated Water Levels.............................................................................................43 4.2.2 Steady State Water Budget.........................................................................................44


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page vii

5 Transient Model Development...................................................................................................45 5.1 Model Specifications..........................................................................................................45 5.2 Boundary Conditions..........................................................................................................45

5.2.1 Recharge (Specified Flux)..........................................................................................46 5.2.2 Transfer (Cauchy) Boundary......................................................................................46 5.2.3 Well Boundary Conditions.........................................................................................46

5.3 Transient Model Development...........................................................................................46 5.3.1 Temporal Discretisation .............................................................................................46

5.4 Transient Model Calibration ..............................................................................................47 5.5 Calibrated Model Results ...................................................................................................47

5.5.1 Groundwater Level Hydrographs...............................................................................47 5.5.2 Groundwater Discharge Hydrographs........................................................................49

6 Discussion ..................................................................................................................................50 6.1 Introduction ........................................................................................................................50

6.1.1 Surface-water / Groundwater Interaction Considerations..........................................50 6.2 Development Scenarios......................................................................................................51

6.2.1 Scenario with No Pumping ........................................................................................51 6.2.2 Effect of Pumping Proximity on Groundwater Discharge to River ...........................52 6.2.3 Effects of Current (2004) Licensed Entitlements on Groundwater Discharge to the Katherine River ..........................................................................................................................53 6.2.4 Steady State Analysis Scenario “A” Current - (2004) Pumping................................53 6.2.5 Transient Analysis of Scenario “A” – 2004 (Current) Pumping................................55

6.3 Bore Capture Zones............................................................................................................55 7 Conclusions ................................................................................................................................58 8 Recommendations ......................................................................................................................59 9 References ..................................................................................................................................60


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page viii

List of Figures

Figure 1 Location and extent of the model domain. .....................................................................13 Figure 2 Variation in the average annual rainfall across the model domain. ...............................14 Figure 3 3sec (90 metre) Shuttle Radar Topographic Model of the model domain showing

drainage. .........................................................................................................................16 Figure 4 Regional surface geology and the locations of Tindall Limestone Aquifer discharge

zones. ..............................................................................................................................21 Figure 5 Groundwater flow and average late dry season discharge from the Tindall Limestone.

........................................................................................................................................22 Figure 6 Groundwater / surface water interactions indicating the bank storage effects during a)

dry season and b) wet season. ........................................................................................24 Figure 7 Relationship between rainfall at Katherine and potential recharge modified from Jolly,

(2000). ............................................................................................................................25 Figure 8 Location of observation bores and bores used to calibrate the model. Hydrographs of

selected bores indicate the effect of the Cretaceous sediments......................................26 Figure 9 Typical hydrographs across the study area. Green traces indicate bores located where

the Limestone Aquifer is overlain by Cretaceous sediments, Blue traces indicate bores where minimal cover exists over the Limestone Aquifer...............................................27

Figure 10 Groundwater level contours, November 2003. ..............................................................28 Figure 11 Katherine Railway Bridge stage height data (G8140001)..............................................29 Figure 12 Current Gauging Station Locations along the Katherine River .....................................30 Figure 13 Gauged flows (cumecs) in the Katherine River at the Railway Bridge (G8140001) and

the total estimated groundwater discharge to the Katherine River from the aquifer at Galloping Jacks (G8140301)..........................................................................................31

Figure 14 Gauged flows at Seventeen Mile Creek (G8140159), indicating flows in the Katherine River not due to discharge from the Limestone Aquifer................................................31

Figure 15 Model mesh geometry showing region of mesh refinement in the study area...............36 Figure 16 Transfer boundary conceptualization for a losing stream (Diersch, 2004). ...................37 Figure 17 Distribution of transmissivity across the study area based on the mapped occurrence of

the Limestone Aquifer. Region of greater than expected transmissivity in the Flora River identified...............................................................................................................40

Figure 18 Recharge zones, higher recharge rates associated with the outcropping Limestone Aquifer, lower recharge rate in areas where Cretaceous cover exists and no recharge where the Jinduckin exists..............................................................................................41

Figure 19 Steady state water levels for calibrated model. ..............................................................43 Figure 20 Study area location, based on Puhalovich, (2005) TLA model extents. ........................45 Figure 21 Comparison of modelled heads vs observed heads for RN007821. The response from

the calibrated model is in blue........................................................................................47 Figure 22 Comparison of modelled heads vs observed heads for RN022006. The response from




the calibrated model is in blue........................................................................................48 Figure 26 Comparison of modelled discharge vs observed discharge along the Katherine River. 49 Figure 27 In a schematic hydrologic setting where ground water discharges to a stream under

natural conditions (a), placement of a well pumping at a rate (Q1) near the stream will intercept part of the ground water that would have discharged to the stream (b). If the


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page ix

well is pumped at an even greater rate (Q2), it can intercept additional water that would have discharged to the stream in the vicinity of the well and can draw water from the stream to the well (c). .....................................................................................................51

Figure 28 Effects of pumping distance on groundwater discharge to the Katherine River............52 Figure 29 Production bore locations for Scenario “B” ...................................................................53 Figure 29 Comparison of the steady state water balance components for no pumping and

Scenario “A”...................................................................................................................54 Figure 31 Reduction in the dry season flows in the Katherine River due to extraction based on

pumping scenario “A”. ...................................................................................................55 Figure 32 Particle tracking isochrones, NTG and Commonwealth land holdings are identified for

comparison. ....................................................................................................................57


Date printed: 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page x

List of Tables

Table 1 Discharge areas of the Tindall Limestone Aquifer and estimated end of dry average discharge rates and associated elevation of the river along the discharge zone.............23

Table 2 Steady state water balance for the calibrated model with no pumping. .........................44 Table 3 Steady state water balance for the calibrated model with no pumping and with pumping

using Scenario “A”.........................................................................................................54


Date printed 30 July 2008 Status: FINAL Department of Natural Resources, Environment and The Arts Date Last Modified: 30 July 2008 Page 11

1 Introduction

1.1 Background

The Cambrian Limestone aquifer system is the major aquifer in the Wiso, Georgina and Daly

Basins. It represents the source of the majority of the baseflow in the Katherine, Roper, Flora and

Douglas Rivers.

Proposed development in nearby horticultural districts reliant on water from the Tindall Limestone

aquifer, represent a threat to the environmental flow regime of the river particularly in low flow

periods.

This report documents an ambitious attempt at modelling the entire Cambrian Limestone aquifer

system within the Northern Territory, with an emphasis on the area surrounding Katherine.

1.2 Objectives

The objectives of this study were to:

• Develop a regional steady state model that will provide a framework for the development of a

regional transient groundwater model for the whole of the Cambrian Limestone Aquifer System.

• Extend the steady state model to incorporate transient conditions in the area of interest

identified by Puhalovich, (2005). The regional steady state model will provide the boundary

conditions for the detailed transient model.

• Provide a basis for the assessment of development with respect to pumping scenarios and the

effects on the dry season flows in the Katherine River ie during periods when base flows are

lowest

A groundwater model has been developed based on the conceptual hydrogeological model

proposed by Water Studies, (2001) and Puhalovich, (2005).

This work has provided affirmation of a viable hydrogeological model and a tool to be applied

under various development scenarios to assess impacts on groundwater levels and spring flows.

This report presents the model’s basis for development and identifies areas in which data

deficiencies exist.

1.3 Model Steps

The groundwater model was developed using the following steps:

1) Conceptual model development;

2) Numerical model implementation;



3) Development of a regional steady state model;

4) Calibration of the regional steady state model;

5) Extension of the calibrated steady state model to the transient domain in the Katherine region

of the study area;

6) Calibration of the transient model to the hydrologic data from 1994 to 2004 including

rainfall/recharge data, water levels hydrographs, stream flow data and pumping data;

7) Sensitivity analysis of the calibrated model to determine what are the key assumptions which

have a significant impact on the model;

8) Prediction of effects of various pumping scenarios on the Katherine River discharge for the flow

record from July 1963 to June 2004.

1.4 Location

The study area for the groundwater modelling comprises the full extent of the Cambrian Limestone

in the Daly Basin, the northern Wiso Basin and the northern Georgina Basin (Figure 1). The

Cambrian Limestone aquifer covers an area of greater than 159,000 km2 and is bounded

approximately by the latitudes -13.41°S and -20.48°S and the longitudes 130.72°E and 137.74°E.

In the Daly Basin the Cambrian Limestone is termed the Tindall Limestone Aquifer and provides

dry season baseflow to the Katherine, Edith, Flora, Douglas, Ferguson, and Daly Rivers.

The study area is located in the vicinity of Katherine and is predominantly located within the

catchment of the Daly Basin.



"

"

"

Arafura Sea

Wiso Basin

Georgina Basin

Dunmarra Basin

Daly Basin

DAL Y RIVER

VICTORIA RIVER

ROBI

NSO

N R

IVE

R

MANN RIV

ER

MAR

Y R

IVER

CAL

VE

RT RIVER

KATHERINE R

IVER

SANDOVER RIVER

DRY

RIVE

R

H

ANSON RIVER

EAST ALLIGATOR RIVER

KING

RIV

ER

WICKHAM RIVER

COX RIVER

MAINORU RIVER

REYNOLDS RIVER

FOE

LSC

HE

RIV

ER

ADELAIDE RIVER

MCA

R

THUR RIVER

TOWNS RIVER

WALKER RIVER

CATO RIVERWOOLEN RIVER

ARMSTRONG RIVER

ROPER RIVER

WILDMAN RIVER

FISH R

IVE

R

BUNDEY RIV

ER

ARNO

LD R

IVER

FINNISS RIVER

GO

SSE

RIVE

R

JAM

ES R

IVER

NATH

AN R

IVER

GLY

DE

RIV

ER

MOYLE RIVER

TOMKINSON

RIVE

R

FREW RIVER

ELKEDRA RIVER

CU

LLE

N R

IVE

R

HOW

ARD RIVER

GLYD

E RIVER

DARWIN

KATHERINE

TENNANT CREEK

130°0'0"E

130°0'0"E

135°0'0"E

135°0'0"E

20°0

'0"S

20°0

'0"S

15°0

'0"S

15°0

'0"S

" Town

Major Drainage

Major Road

Model Extent

NT Coastline

Arafura Sea

0 50 100 150 200 25025Kilometres

Geological RegionsOther Regions

Daly Basin

Dunmarra Basin

Georgina Basin

Wiso Basin²

Figure 1 Location and extent of the model domain.



1.5 Climate

The study area falls within the wet-dry tropics in the northwest and semi-arid to the southeast. In

the wet-dry tropics there are two distinct seasons, the wet season from December to April and the

dry season spans the remainder of the year. Annual rainfall increases to the northwest from an

average of 980 mm at Katherine (119 yrs of record) to 1,156 mm at Oolloo Crossing (44 yrs of

record) and 1,328 mm at Adelaide River (66 yrs of record). The variation in annual rainfall across

the whole model domain is presented in Figure 2.

"

"

300 - 400

400 - 500

600 - 700

500 - 600

700 - 800

800 - 900

1000 - 1100

1100 - 1200

900 - 1000KATHERINE

TENNANT CREEK

0 50 100 150 200 25025Kilometres

²

Median Rainfall (mm)100101 - 200201 - 300301 - 400401 - 500501 - 600601 - 700701 - 800801 - 900901 - 10001001 - 11001101 - 12001201 - 14001401 - 1500

Figure 2 Variation in the average annual rainfall across the model domain.

Analysis of rainfall vs recharge indicates that minimum rainfall of approximately 700 mm/yr is

required before appreciable annual recharge (Jolly et al., 2000).



1.6 Geomorphology

The major drainage within the study area shows a rectangular drainage pattern, where both the

main stream and its tributaries exhibit right-angle bends, indicating that geological structures (ie

faulting and jointing) have strongly influenced the development of the drainage, especially where

the drainage incises the Tindall Limestone. The Daly River is orientated sub-parallel to the strike

of the Daly Basin. The ephemeral drainage shows a more dendritic pattern.

1.6.1 Topography

The study area varies in topography from approximately 30 to 300 metres above Australian Height

Datum (Figure 3). The low lying areas are along the main drainage, and the highest topography is

located in the central Wiso and Georgina Basin. Topography is relatively rugged on the dissected

flanks of the plateaux, where steep gullies have been incised into the soft Cretaceous rocks. In

contrast the areas where the Tindall Limestone is exposed have low undulating karstic topography

with generally sparse outcrop.



Bulman

Ngukurr

KATHERINE

Mataranka

Roper Bar

Daly Waters

TENNANT CREEK

Adelaide River

0 60 120 180 240 30030Kilometres

TownRoadHighwayCreekRiverArafura SeaModel ExtentNT Boundary

Elevation (mAHD)High : 500 Low : 0

²

Figure 3 3sec (90 metre) Shuttle Radar Topographic Model of the model domain showing drainage.

1.7 Previous Modelling

Documented groundwater models of the Tindall Limestone in the Katherine area have been

developed by Water Studies, (2001) and Puhalovich, (2005). Some of the salient components of

these models are discussed below:

Water Studies developed a detailed model centred on the Katherine River where the occurrence of

the Tindall Limestone was within the Daly River surface water catchment. The objective of the

model was to determine the effects of pumping in the Venn Horticultural sub-division on

groundwater levels and flows in the Katherine River.



Puhalovich, (2005) developed a model based on the work of Water Studies, (2001). The model

was extended to the south east to include the groundwater divide between the Katherine and

Roper River and the discharge zones at the Roper River. It incorporated the contribution to the

Tindall aquifer from the inflows from the Cretaceous aquifers in the vicinity of the head waters of

the King River using a General Head Boundary condition.

• Groundwater levels in monitoring bores in the Venn Horticultural area and the Cretaceous

rocks in the King River area were derived from incorrect surface RL values, this resulted in

data which could not be adequately calibrated against.

• Calibration of water levels to very wet years using discharge estimates from average years.

Calibration was conducted using 2001 water levels and average groundwater flows.

Hydrograph data and flows in the Katherine River during this period are well above the

average (approx. 2.5 cumecs compared with an estimated 1.33 cumecs for an average

year).

• The input from the Cretaceous rocks to the east of the Venn Horticultural area has been

simulated using a general head boundary. However, when examining the components of

the water budget for the model it can be seen that the extraction due to pumping is not

reflected by a reduction in the discharge to the Katherine River. It is inferred that the

shortfall in the water balance is due to water being supplied to the model at the general

head boundary. It is suggested that the Cretaceous rocks act as a constant flux boundary

ie controlled by recharge to the Cretaceous rocks, which, would still manifest as a relatively

constant head in this region.



2 Hydrogeological Setting and Conceptual Model

2.1 Regional Geology

The major hydrogeological features of the region are the Cambrian-Ordovician Daly Basin , Wiso

Basin and Georgina Basin. Early Cretaceous rocks of the Dunmarra Basin overlie much of the

region and obscures the contact between the basins.

The surface geology of the Daly Basin is depicted in Figure 4. The major unit of interest in the

study area is the, Tindall Limestone, which, is the lower unit of the Daly River Group and the major

aquifer with respect to base flows in the Roper, Flora, Katherine, Douglas and Daly Rivers.

2.1.1 Cretaceous Rocks of the Dunmarra Basin

The Cretaceous aged Mullaman Beds of the Dunmarra Basin forms a mantle of lateritised

claystone and sandstone covering approximately 50% of the study area. The beds are sub-

horizontal and may be divided into an upper cream coloured claystone and siltstone unit and a

basal marine sandstone unit.

Outcrop is generally sparse due to the soft nature of the rock but in places silicification has altered

them to porcellanite and quartzite which outcrop reasonably well. The thickest accumulations in

the Daly Basin are preserved along its axis, running from the north side of the King River, through

Florina Station and then following the north east side of the Daly River as far as Stray Creek

(Tickell, 2002).

The Mullaman Beds are thickest on the Sturt Plateau in the Dry River area and across the

southern map area parallel to the Buchanan Highway, the formation may be up to 75 metres thick

with the clayey upper unit comprising 60m of its thickness. The thickness of the sandy unit is

variable and ranges from less than 5 metres thick, up to 25 metres thick in parts of the central

plateau area. The sandstone is generally friable, however, siliceous outcrops of the unit are

located in the vicinity of Gorrie Station. Where the upper claystone is thin and eroded, the potential

recharge to the underlying limestone aquifer is increased. In all places, the Mullaman Beds are

above the regional water level.

The main influence of the Cretaceous sediments is to reduce the recharge to the Tindall Limestone

aquifer. This assertion is based on the lithology of the unit, which is predominantly clay/clayey

sand and the subdued response of groundwater hydrographs for the bores located in areas with

Cretaceous cover (eg RN022006 - Figure 8 and Figure 9).

2.1.2 Oolloo Dolostone

The Oolloo Dolostone is the uppermost formation in the Cambrian-Ordovician Daly Basin, a largely

undeformed sequence of shallow water carbonate rocks. Outcrop is generally poor due to the



extensive cover of Cretaceous rocks. The main exposures occur at the northwestern and

southeastern ends of the Daly Basin (Tickell, 2002).

2.1.3 Jinduckin Formation and Anthony Lagoon Beds

The laterally equivalent formations of the Jinduckin Formation and Anthony Lagoon Beds occur in

the Wiso Basin, Daly Basin and Georgina Basins respectively.

The Jinduckin Formation overlies the Tindall Limestone and Montejinni Limestone and the Anthony

Lagoon Beds overlie the Gum Ridge Formation. They are mainly of dolomitic siltstone, interbeds of

dolomitic sandstone-siltstone and dolostone. The Jinduckin Formation has eroded off over most of

the Sturt Plateau and only exists in the Daly Basin and in the north of the Wiso Basin, where it is

overlain by Mullaman Beds. Similarly, a partial section of the Anthony Lagoon Beds is seen in the

Larrimah area where highly weathered remnants may be detected in gamma logs. The formation

continues to thicken towards the south-east into the Georgina Basin where approximately 60m of

its lower section may be identified in bore RN27958 east of Dunmarra.

These formations overlie and confine the major limestone aquifers of the region. Where they exist

below the water table, they may host viable aquifers, however, are generally of low permeability

and yield. Dissolution of evaporite beds within these formations result in water with significant

levels of sulphate and sodium chloride salts.

2.1.4 Tindall Limestone, Gum Ridge Formation and Montejinni Limestone

The time equivalent Cambrian limestone formations – the Tindall Limestone, the Montejinni

Limestone and Gum Ridge Formations host the vast majority of the water resources in the region

and have many stratigraphic similarities.

The Tindall Limestone of the Daly Basin is a massive, thinly bedded, multi-coloured crystalline,

dolomitised limestone with some chert nodules and mudstone bands, particularly in the lower

layers. The Tindall Limestone becomes shaley to the northwest in the Douglas River area.

The Montejinni Limestone of the Wiso Basin consists of limestone, dolomitic limestone, dolomite

and calcareous mudstone and siltstone. In many parts of the basin, a threefold division has been

recognised with an upper and lower limestone unit each approximately 25 metres thick and an

intervening red/brown mudstone about 10 metres thick.

The Gum Ridge Formation of the Georgina Basin, although similarly sequenced to the Tindall

Limestone Formation, is generally described as consisting of limestone, fine grained sandstone

and siliclastic mudstone and nodular chert. The depositional environment of this formation has

resulted in a greater proportion of carbonate sediment.



For the purposes of this study, no hydrogeological distinction is made between each of the

formations as they represent a single, extensive aquifer system and are referred to generally in this

report as the ‘Limestone Aquifer’.

2.1.5 Antrim Plateau Volcanics

The early Cambrian Antrim Plateau Volcanics is generally a flat lying, dark grey/green coarse

grained tholeiitic basalt that underlies the Tindall Limestone, Montejinni Limestone and Gum Ridge

Formations and forms the hydrogeological basement. The overall thickness of the volcanics is in

the order of 150 metres. The Helen Springs Volcanics are also flat-lying tholeiitic basalts and are

correlatives of the Antrim Plateau Volcanics. These basalts exist beneath the great majority of the

Sturt Plateau but due to lack of outcrop and distinguishing features, it is not possible to determine

the location of the different basalt units.

The basement high creates division of groundwater flow to the Flora River and Roper River from

the Wiso Basin and Georgina Basin (refer Figure 5).



"

"

"

Daly River

Roper River

Katherine River

Douglas River

Flora River

Elsey Creek

Hayes Creek

KATHERINE

Mataranka

Daly Waters0 20 40 60 80 10010Kilometres

" TownMajor RoadCreekRiverDischarge Zones

CretaceousUpper Oolloo DolostoneLower Oolloo DolostoneJinduckin FormationTindall LimestoneAntrim Plateau VolcanicsUndifferentiated Proterozoic

²

Figure 4 Regional surface geology and the locations of Tindall Limestone Aquifer discharge zones.

2.1.6 Groundwater Flow

The basement high where the Antrim Plateau Volcanics occurs above the water table creates a

division of groundwater flow to the Flora River and Roper River from the Wiso Basin and Georgina

Basin.

Groundwater divide between the Katherine and Roper Rivers effectively separate the two regions.



"

"

"

Douglas River(1.0 cumecs)

Daly River(1.0 cumecs) Katherine River

(1.33 cumecs)

Flora River(2.3 cumecs)

Roper River(3.1 cumecs)

KATHERINE

Mataranka

Daly Waters

Flow Direction

[mAHD]5051 - 6061 - 7071 - 8081 - 9091 - 100101 - 110111 - 120121 - 130131 - 140141 - 150151 - 160161 - 170171 - 190

" TownsMajor RoadDischarge ZonesRiverModel Extent

0 20 40 60 80 10010Kilometres

²

Figure 5 Groundwater flow and average late dry season discharge from the Tindall Limestone.

2.1.7 Discharge from the Cambrian Aquifer System

Major discharge from the Cambrian – Ordovician aquifer system occurs predominantly from the

Tindall Limestone and Oolloo Dolostone units of the Daly Basin. The majority of the observed

discharges from the Limestone Aquifer in the Daly Basin occur at the Douglas River and Hayes

Creek, the Daly River south of Beeboom Crossing, Flora River and along the Katherine River

(Tickell et al., 2002).

Major discharges from the Limestone Aquifer also occur in the vicinity of Mataranka at the

Mataranka Hot Springs, The Bitter Springs and along the Roper River and Elsey Creek. The

locations of the major discharge zones are indicated in Figure 5. A summary of the estimated



average discharge from each of the identified areas and their estimated Stage Height elevations

derived from the SRTM data are presented in Table 1.

Table 1 Discharge areas of the Tindall Limestone Aquifer and estimated end of dry average discharge rates and associated elevation of the river along the discharge zone.

Location Gauging

Station

Estimated Average Discharge

[cumecs]

Estimated RL of River Stage Height

[mAHD]

Seventeen Mile

Creek*

G8140159 0.3 N/A

Katherine at Low

Level Crossing

G8140001 1.14 86.5

Katherine River at

Galloping Jacks

G8140301 1.33 81

Roper River / Elsey

Creek

G9030176 3.1 120 - 130

Douglas River /

Hayes Creek

G8140063 1 55

Flora River G8140044 2.3 80

Daly River at Mt

Nancar

G8140040 1 30

* Dry season flows in Seventeen Mile Creek are not sourced from the Tindall Limestone Aquifer and have been

added to identify major end of dry season flow inputs to the Katherine River.

2.2 Study Area Hydrogeology

The study area is within the Katherine River catchment (Figure 1). The major geological units in

the study area are the Cretaceous sediments, Jinduckin Formation and Limestone Aquifer.

2.2.1 Limestone Aquifer Saturated Thickness

The hydrogeological units within the study area comprises the Limestone Aquifer overlain by the

Jinduckin Formation and Cretaceous rocks.

Pidsley, (1987) provides an estimate of the saturated thickness in the region of the Venn

Horticultural Subdivision. Puhalovich, (2005) identified from geological cross-sections that the

saturated thickness of the Limestone Aquifer varies across the area. Based on the cross-sections

presented and the work by Pidsley, (1987) the saturated thickness is seen to vary from greater



than 100 metres along the southwestern margin of the Limestone Aquifer, to less than 30 metres to

the northeast of the Venn Horticultural Subdivision.

It has been assumed, based on current knowledge, that in the area to the west of the Venn area

the transmissivity are assumed to be at a maximum of approximately 5000 m2/d. In the Venn

region the transmissivity is assumed to be approximately 2000 m2/d.

2.2.2 Hydraulics of the Cretaceous Sediments in the Area of the King River

Hydrograph dynamics indicate that the recharge to the Cretaceous sediments is relatively constant

and that the inputs to the Limestone Aquifer are similarly constant.

2.2.3 Groundwater Flow

In the study area groundwater flows from the groundwater divide in the King River area and

discharges to the southeast at the Roper River and to the northwest at the Katherine River.

Groundwater also flows to the Roper River from the Georgina Basin.

2.2.4 Groundwater Discharge

Jolly et al., (2000), identified two components to the dry season recessions in the stream discharge

hydrographs. These recessions were interpreted to signify two different sources for the water, river

bank storage and regional groundwater flow. Bank storage is described as where a large

permanent stream undergoes an increase in river stage under the influence of an arriving

floodwave, flow may be induced into the stream banks and the adjoining aquifer. As the stage

declines, the flow is reversed (Freeze and Cherry, 1979). River bank storage effects along the

Katherine River are expected to be most evident during years of above average rain, flood heights

and recharge events.

Dry Season Dynamics

a)

Dry SeasonFlow Height

River Bank

Flow from the aquiferinto the river.

Groundwater level

Wet Season Dynamics

b)

Wet SeasonFlow/Flood Height

River Bank

Flow from the riverinto the aquifer.

Groundwater level

Figure 6 Groundwater / surface water interactions indicating the bank storage effects during a) dry season and b) wet season.



2.3 Study Area Observation Data

2.3.1 Rainfall and Potential Recharge

Rainfall data is available in the Katherine area from the Katherine Post Office site, DR014902.

Data is available from 1887 to present. The methodology for determining the potential recharge to

the aquifers in the Katherine region, based on these rainfall records, was developed by Jolly et al.,

(2000). The potential recharge record derived from this work was used as a basis for the recharge

to the Limestone Aquifer system.

The potential recharge was calculated from the daily rainfall record, using estimates for the end of

dry season soil moisture deficit and daily losses (evapotranspiration etc). A soil moisture deficit of

150 mm and wet season evapotranspiration (ET) of 5 mm/day were chosen. It was also assumed

that there was little surface runoff from the ground overlying the Limestone Aquifer (Jolly et al.,

2000).

The relationship between the daily rainfall and daily potential recharge from 1960 to 2004 is

presented in Figure 7.

Jan-

60

Jan-

62

Jan-

64

Jan-

66

Jan-

68

Jan-

70

Jan-

72

Jan-

74

Jan-

76

Jan-

78

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

Date

0

50

100

150

200

250

Rai

nfal

l / P

ot. R

echa

rge

(mm

)

Daily RainfallDaily Potential Recharge

Figure 7 Relationship between rainfall at Katherine and potential recharge modified from Jolly, (2000).

2.3.2 Observation Bores

25 bores within the Study Area have time series water level data. The list of bores is provided in

Appendix A. The hydrographs of 6 bores RN007821, RN022002, RN022006, RN022397,

RN023427 and RN029429, form a transect away from the Katherine River, were used to provide

the basis for the transient calibration.



"

"

KATHERINE RIV ER

KIN

G R

IVER

WATERHO

USE R

IVER

EDITH RIVER

ROPER RIV ER

RN022287

RN029429

RN024613

RN023428RN023427

RN023425

RN023424

RN022006RN022002

RN022001RN007821

RN005032

KATHERINE

Mataranka

0 10 20 30 40 505Kilometres

" TownMonitoring BoresRoadMajor RoadCreekRiverCretaceous

GraniteJinduckinOolloo_lowerOolloo_upperTindall

²

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

Date

100

110

120

130

140

150SW

L (m

AHD

)

RN007821 (unconfined, bore collar surveyed)

Measuring Point RL = 196.465 mAHD

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

Date

100

110

120

130

140

150

SWL

(mAH

D)

RN022006 (unconfined, bore collar surveyed)

Measuring Point RL = 188.25 mAHD

Figure 8 Location of observation bores and bores used to calibrate the model. Hydrographs of selected bores indicate the effect of the Cretaceous sediments.

2.3.3 Groundwater Level Hydrographs

Groundwater hydrographs of the study area are presented in Appendix B. The six hydrographs

used to calibrate the model are discussed below and presented in Figure 9.

RN007821 is located approximately 17.5 kilometers to the east southeast of the Katherine River.

The bore is located in an area where the Limestone Aquifer outcrops.

RN022002 is located to the southeast of the Venn Horticultural sub-division and is in an area with

approximately 10-20 metres Cretaceous cover (Britten, 1983), the subdued response to the

seasonal recharge events is interpreted to be due to lower recharge rate through the Cretaceous

layer.

RN022006 is located to the southeast of the Venn Horticultural sub-division and is in an area with

with approximately 10-20 metres Cretaceous cover (Britten, 1983), the subdued response to the

seasonal recharge events is interpreted to be due to lower recharge rate through the Cretaceous

layer.

RN022397 is located near the Katherine River, the groundwater level response shows strong

influence from the seasonal levels in the river.



RN023427 is located to the east of the Venn Horticultural sub-division and is constructed in

Cretaceous sediments (Yin Foo, 1985). The response is relatively flat and shows no evidence of

seasonal recharge events.

RN029429 is located between the Venn Horticultural sub-division and the Katherine River. It is

located in an area where the Limestone Aquifer outcrops, which, like RN007821, is reflected in the

groundwater level response to recharge.

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

80

90

100

110

120

130

140

150

160

Wat

er L

evel

(mA

HD

)

RN023427RN022006RN007821RN029429RN022397

Figure 9 Typical hydrographs across the study area. Green traces indicate bores located where the Limestone Aquifer is overlain by Cretaceous sediments, Blue traces indicate bores where minimal cover exists over the Limestone Aquifer.

2.3.4 Potentiometric Head Distribution

The groundwater levels for November 2003 were collated and contours of the head distribution are

presented in Figure 10.



"

"

Katherine River(1.33 cumecs)

Roper River(3.1 cumecs)

KATHERINE

Mataranka

Flow Direction

[mAHD]5051 - 6061 - 7071 - 8081 - 9091 - 100101 - 110111 - 120121 - 130131 - 140141 - 150151 - 160161 - 170171 - 190

" TownsMajor RoadDischarge ZonesRiverModel Extent

0 10 20 30 40 505Kilometres

²

Figure 10 Groundwater level contours, November 2003.

2.3.5 River Stage Height Data

River stage height data is available at the Katherine Railway Bridge (G8140001) from 02/03/1959

to 07/08/2004. There are some gaps in the early portion of the record (1960 – 1962 and 1971) as

can be seen from Figure 11. Total discharge from the Limestone Aquifer to the Katherine River is

derived from Galloping Jacks (G8140301).

It should be noted that the river stage heights of the two stations during the dry season differ by

approximately 5-6 metres. The stage height during the end of the dry season, upstream of the

Katherine Railway Bridge, is approximately 86.5 mAHD (due to the weir), whilst the stage height at



Galloping Jacks at the end of the dry season has been estimate to be approximately 81 mAHD

(from the 1:5000 Flood Mapping for Katherine, March 2005).

Jan-

1960

Jan-

1965

Jan-

1970

Jan-

1975

Jan-

1980

Jan-

1985

Jan-

1990

Jan-

1995

Jan-

2000

Jan-

2005

Date

85

90

95

100

105

110

Sta

ge H

eigh

t(m

etre

s A

HD)

Figure 11 Katherine Railway Bridge stage height data (G8140001)

Maximum daily recorded height data has been extracted from the corporate database Hydsys for

G8140001 at daily intervals. The lowest levels of 86.54 mAHD occurred in late 1962. The flow

measured at this time was 0.69 cumecs. During 1996 the level reached a minimum of

86.56 mAHD, which, corresponds to a flow of 0.94 cumecs.

2.3.6 River Gauging Data

Stream discharge flow gaugings are available at two locations Katherine Railway Bridge

(G8140001) and Galloping Jacks (G8140301). Continuous gauging data exist at the Katherine

Railway Bridge generated from the stage height data rated to the manual gauged flows. The

locations of these two river flow stations are depicted in Figure 12.



#

#

#

#

"

"

Seve

ntee

n Mile

Cre

ek

KATHER IN

E RIVER

KIN

G R

IVER

DRY R

IVER

EDITH RIVER

WATERHOUSE RIVER

ROPER RIVER

G8140301

G8140159

G8140002G8140001KATHERINE

Mataranka

0 10 20 30 40 505Kilometres

²" Towns# Gauging Stations

Major RoadCreekRiverModel Extent

Cretaceous RocksUpper Oolloo DolostoneLower Oolloo DolostoneJinduckin FormationTindall LimestoneUndifferetiated

Figure 12 Current Gauging Station Locations along the Katherine River

The difference between the limited number of coincident dry season flow gaugings at the Katherine

Railway Bridge (G8140001) and at Galloping Jacks (G8140301) has been used to determine an

extended estimate of the total discharge from the Limestone Aquifer to the Katherine River at

Galloping Jacks. Jolly et al, (2000) identified that the late dry season discharge to the Katherine

River at Galloping Jacks is generally 1.17 times the flow recorded at the Katherine Railway Bridge

(Figure 13).

Gauging data from Seventeen Mile Creek (G8140159), indicate that considerable flows during the

dry season can be expected during wetter periods (Figure 14). Typically flows of 0.2-0.3 cumecs

occur during dry periods (1960 – 1972 and 1986 – 1997) whilst, flows greater than 1 cumecs have

been recorded during the recent wet period from 1999 to present.



Jan-

1960

Jan-

1962

Jan-

1964

Jan-

1966

Jan-

1968

Jan-

1970

Jan-

1972

Jan-

1974

Jan-

1976

Jan-

1978

Jan-

1980

Jan-

1982

Jan-

1984

Jan-

1986

Jan-

1988

Jan-

1990

Jan-

1992

Jan-

1994

Jan-

1996

Jan-

1998

Jan-

2000

Jan-

2002

Jan-

2004

0.1

1

10Ka

ther

ine

Riv

er D

isch

arge

(cum

ecs)

G8140001- Measured FlowG8140301- Estimated Flow

G8140001 - Gauged Flows

Figure 13 Gauged flows (cumecs) in the Katherine River at the Railway Bridge (G8140001) and the total estimated groundwater discharge to the Katherine River from the aquifer at Galloping Jacks (G8140301).

Jan-

60

Jan-

62

Jan-

64

Jan-

66

Jan-

68

Jan-

70

Jan-

72

Jan-

74

Jan-

76

Jan-

78

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

040.1

1

10

Seve

ntee

n M

ile C

reek

Dis

char

ge(c

umec

s)

G8140159 - Gauging StationG8140159 - Gauged Flows

G8140159 - Gauged Flows

Figure 14 Gauged flows at Seventeen Mile Creek (G8140159), indicating flows in the Katherine River not due to discharge from the Limestone Aquifer.



2.4 Conceptual Model

The conceptual model for the Limestone Aquifer was developed by the Water Resources section of

NRETA from the available data and observations outlined in the previous sections. The

conceptual model for the Limestone system can be summarized as:

• The Limestone aquifer may be represented as a single unconfined layer. At the regional scale

the karstic nature of the aquifer can be represented as an equivalent porous medium, with

effective transmissivity and storage values used to simulate the overall aquifer characteristics.

• Mapped occurrence of the Jinduckin Formation confines the Limestone aquifer and is expected

to have lower transmissivities and storage coefficient than the unconfined Limestone aquifer.

• The Limestone aquifer was expected to have greatest permeability within the weathered zone,

confined to the upper 150 metres from the surface. For the purposes of this exercise the

aquifer was considered to have a constant thickness below the groundwater table. That is the

system can be represented as a single layer with zones of constant transmissivity instead of

varying the hydraulic conductivity and aquifer thickness. This is considered valid as the

variations in the groundwater level are considered small compared to the saturated thickness

of the Limestone Aquifer.

• Aquifer transmissivity of 5,000 m2/d based on the Water Studies modelling results.

• A single estimate of transmissivity in the Venn region indicate that lower values (around

2000 m2/d) may be applicable in this area.

• The confined regions of the Limestone Aquifer were assigned a single value of 100 m3/d/m.

• Unconfined aquifer storage coefficient was 0.04. This is considered a reasonable estimate as

previous experience (Jolly, pers comm.) indicates that this value should be between 0.01 and

0.07.

• Confined aquifer storage coefficient was assigned 0.0001 based on typical confined aquifer

storage coefficients.

• The dominant recharge is mechanism is via sinkholes, however, this is not well understood, so

the recharge was estimated as diffuse recharge.

• The main influence of the Cretaceous sediments is to reduce the recharge to the Limestone

Aquifer. This is based on the subdued response of hydrographs for bores located in areas with

the Cretaceous cover (eg RN22006).

• Initial estimates of the steady state annual recharge over the two areas were 150 mm/yr

(0.41 mm/d) for the outcropping limestone and 40 mm/yr (0.11 mm/d) for the Cretaceous cover.



• Based on the steady state recharge estimates the potential recharge model developed using

the Katherine rainfall record (Jolly et al, 2000) was scaled to provide transient recharge values.

The recharge model estimates the soil moisture deficit and daily evapotranspiration to derive

the potential recharge due to deep drainage.

• Recharge occurs from the Katherine River when stage height > groundwater level.

• The dominant discharge from the aquifer is through the streambed and via springs. Discharge

occurs along the length of the Katherine River where it intersects the Limestone Aquifer. Bank

storage mechanism, beyond scope of this model.

• Over the long term the late dry season discharge to the river via spring flows range from 1 to 2

cumecs, with an average discharge rate of 1.33 cumecs, with approximately 17% or

0.23 cumecs discharging downstream of the low level weir.

• Contributions to the groundwater from the Cretaceous rocks near King River are considered as

being relatively constant and as such constitute a constant flux boundary.

• Evapotranspiration from the riparian zone is estimated at approximately 3 mm/day. The ET

has not been explicitly considered in this model. Based on the ET value a riparian zone width

of approximately 200 metres and a length 12 kilometres of river the total ET is 720 m3/d

(0.0083 cumecs or <0.001% of the flow in the river). Assuming this is all derived from

groundwater this is a relatively small component of the water balance.



3 Model Development

3.1 Model Specifications

The model encompasses the mapped occurrence of the Cambrian Limestone Aquifer of the Daly,

Wiso and Georgina Basins. The entire model covers an area of approximately 195,000 km2. The

south western boundary in the Georgina Basin was selected to coincide with the interpreted

groundwater divide evident from groundwater levels (Tickell, 2003).

The model mesh was developed to facilitate two levels of detail in the modelling.

• The first being a regional model encompassing the entire Limestone Aquifer. The regional

mesh was refinement along the major rivers where baseflow from the Limestone Aquifer has

been identified. The coarse nature of the model at this regional level was considered

reasonable as these areas are to provide steady state boundary conditions outside of the study

area.

• The second area with the greater level of detail encompasses the Limestone Aquifer within the

area reported by Puhalovich, (2005). The extent of TLA model is presented in Figure 20. The

model mesh in this region was refined to allow for the incorporation of transient boundary

conditions such as pumping bores and groundwater / surface water interactions along the

Katherine River.

The difference in the density of mesh elements between the regional area and the study area can

be seen in Figure 15.

3.1.1 Numerical Model Code

The finite element package FEFLOW® v5.205 from WASY was used to simulate the saturated flow

processes. FEFLOW® is a fully three dimensional finite-element package capable of simulating

unsaturated and saturated flow and contaminant transport. FEFLOW® also has built-in mesh-

design, problem editing and graphical post processing display modules that allow rapid model

development, execution and analysis (Diersch, 2004). A 32-bit PC laptop under Windows XP was

used as the platform for the numerical simulations (transient simulations over 41 years typical took

70-80 minutes).

The high-level graphical interface, the Geographic Information System (GIS) capabilities, and the

capacity for detailed mesh generation built into FEFLOW are important features that have allowed

the rapid development and testing of the models described in this report.

Finite elements provide greater flexibility in the mesh design than the rectilinear grids employed by

finite difference code, allowing for the refinement of the mesh around points such as bores and

linear features such as rivers. The code proprietary and as such has limitations because the



software requires a licence to run – unlike the core code for Modflow which is “freeware” from the

US Geological Survey.

3.1.2 Spatial Discretisation

The superelement, mesh and model were developed with the FEFLOW® package. The mesh was

generated using the automatic Triangle option (Shewchuk, 2002). This feature offers the ability to

define the local variation of mesh density by allowing for the refinement of the mesh around

specified point and line features. Node placement and refinement of the mesh was defined at the

locations of the production bores and monitoring bores. Refinement of the model mesh was also

defined along the major drainage features previously identified, where significant discharge from

the Limestone Aquifer occurs. Geological boundaries were also incorporated to provide accurate

assignment of model parameters using the “JOIN” feature in the FEFLOW.

The regional mesh was generated using the following settings for the Triangle (Delaunay)

generator in the Mesh Generator Options:

• Quality mesh, minimum angle <= 30 degrees

• Force all triangles to be Delaunay

• Fill all possible holes in mesh

• Divide-and-conquer meshing algorithm

An initial mesh density of 1000 elements was used in the Generate Automatically option to

generate the mesh.

The regional mesh was then refined in the vicinity of the study area using the Mesh Geometry

Editor.

The resultant mesh used in the modelling is presented in Figure 15 and comprises 29,258

elements and 15,068 nodes.



Figure 15 Model mesh geometry showing region of mesh refinement in the study area.

3.2 Layers

It was considered that the Cambrian Limestone Aquifer could be approximated as a single layer

system, with spatially variable transmissivity. This is considered valid as the variations in the

groundwater level are considered small compared to the saturated thickness of the Limestone

Aquifer.

3.3 Boundary Conditions

The following boundary conditions have been used to simulate the various input/output features in

the model.

• The diffuse recharge to the aquifer is a variable flux boundary describing the proportion of the

potential recharge entering the model.

• Groundwater / surface water interactions. The conceptual model assumes that the dominant

mechanism for discharge of groundwater from the system is through spring flow to the rivers.

Based on this assumption the discharge to the rivers has been implemented using a transfer

boundary. As identified in the conceptual that localized recharge occurs along the Katherine

River during periods when the river height exceeds the groundwater level.



• Pumping bores were implemented using Well BC’s

• The boundary of the model domain where the Antrim Plateau Volcanics outcrop or occurs

above the groundwater level is implemented as no-flow boundaries.

3.3.1 Recharge (Specified Flux at the Model Surface)

Recharge was applied to the entire model based on the surface geology. In areas were the

Cretaceous unit occurred the recharge rate was reduced by a factor of 3-4 (refer to Section 2.1.1).

Also, the groundwater table has been modelled to be in direct connection to the recharge from the

surface (ie no time lag has been introduced to simulate the time for the deep drainage to travel

through the Cretaceous unit, which, can be up to 100 metres thick).

3.3.2 Transfer (Cauchy) Boundary

The groundwater/surface water interactions along river features are simulated using transfer

(Cauchy) boundary conditions. The transfer boundary is similar to the RIV package used by

MODFLOW (Anderson and Woessner, 2002). The transfer boundary condition (Figure 16)

describes a reference hydraulic head which has an imperfect hydraulic contact with the

groundwater body caused by a colmation layer (related to the stream bed conductance).

Additionally to the reference head for the Transfer boundary condition you have to assign a

transfer rate (leakage) to describe the hydraulic properties of the colmation layer (Diersch, 2004).

Figure 16 Transfer boundary conceptualization for a losing stream (Diersch, 2004).

The flux through the colmation layer as shown above can be described using the Darcy equation:

dhhK

lhKq

Rinin

nh

−−=

∆∆

−≈ 200



The transfer rate ( inhΦ ) can be estimated by:

dK in

inh

0≈Φ in (d-1)

The reference hydraulic ( Rh2 ) was initially determined from the 3 second (90 metre) digital terrain

model. Steady state modelling assumed that the stage height for the Katherine River upstream of

the Low Level Crossing is 86.5 mAHD and upstream of Galloping Jacks is 81 mHD.

3.3.3 Well Boundary Conditions

Well boundary conditions describe the injection or withdrawal of water at a single node in m3/d.

Pumping rates were applied either as at a steady state value equal to the annual pumped volume

for the bore converted to m3/d by dividing by 365 days or as a variable pumping rate using power

functions to define the transient pumping schedule at each bore.

3.4 Hydraulic Parameters

The hydraulic parameters of interest in the steady state model were the transmissivity and

recharge rate and the reference hydraulic heads and the transfer out rate associated with the

Transfer Boundary condition. The methodology for the choice of initial values for the major

hydraulic parameters is discussed below.

• Transmissivity was zoned according to the identified unconfined and confined areas of the

Aquifer.

o The unconfined Aquifer was initially defined as being consistent across the model with

an estimated initial value of 5,000 m3/d/m.

o Where the Aquifer is confined by the Jinduckin Formation and equivalents, it is thought

that the Aquifer has developed less permeability and as such has been assigned a

considerably lower constant transmissivity value of 100 m3/d/m.

• Recharge was zoned according to the occurrence of the Jinduckin Formation and equivalents,

the Cretaceous cover and outcropping Limestone Aquifer.

o No recharge was applied to the model where the Jinduckin Formation has been

mapped.

o Highest recharge rates were applied where the Aquifer outcrops.

o Lower recharge rates (approximately 3-4 times lower) were applied to areas with

Cretaceous cover directly overlies the Aquifer.

o Initial estimates of the annual recharge over the two areas were 140 mm/yr (0.38 mm/d)

for the outcropping dolostone and 40 mm/yr (0.11 mm/d) for the Cretaceous cover.



• Transfer Rate describes the connection between the Transfer Boundary and the model and in

this instance is considered to be controlled by the surficial geology. Areas where the

Limestone Aquifer outcrop is assigned 1,000,000 d-1.

3.4.1 Transmissivity Distribution

The steady state model initially used a single value across the entire model domain for the

transmissivity. These were then adjusted to more closely represent the mapped occurrence of the

various hydrogeological units (refer Error! Reference source not found.).

The transmissivity distribution for the calibrated model (Figure 17) is based on the mapped

occurrence of the Cambrian Limestone Aquifer.

In the area of the Sturt Plateau the water table is close to the Antrim Plateau Volcanics and the

limestone aquifer has minimal submergence, transmissivities have been reduced to reflect this.

Areas where the Limestone is confined by the Jinduckin Formation transmissivity was set to

100 m3/d/m.

No flow was assigned to the area to the west of the Stuart Highway between Mataranka and Daly

Waters where the Antrim Plateau Volcanics occurs above the groundwater level. This feature has

been mapped by Yin Foo and Mathews, (2003).

The higher than expected transmissivities in the Flora River region raised questions as to the

reason for these values. It has been suggested that the mapped width of the Limestone Aquifer in

this area is less than the actual width, initial appraisal of geological data has confirmed this

assessment. Further work to provide greater control on this contact is planned.



0 50 100 150 200 25025Kilometres

Transmissivity(m2/day)

No Flow171004301,3002,1602,6004,3205,01017,28041,800

²

Figure 17 Distribution of transmissivity across the study area based on the mapped occurrence of the Limestone Aquifer. Region of greater than expected transmissivity in the Flora River identified.

3.4.2 Recharge Distribution

As discussed previously, the steady state recharge to the model was zoned depending on the

occurrence of the Jinduckin Formation and the Cretaceous units. In areas where the Jinduckin

Formation occurs recharge is set to 0 mm. In areas where the Cretaceous layer is absent higher

recharge rates were applied (up to 150 mm/yr), in areas where the Cretaceous cover exists the

recharge was reduced by approximately 3-4 times (~40 mm/yr).

The trend in steady state recharge can be seen to increase from the southeast to the northwest.

This is in line with what would be expected, given the rainfall patterns discussed in section 1.5.



0 50 100 150 200 25025Kilometres

Recharge(mm/d)

00.00010.0020.0030.0040.0050.040.170.20.220.250.3

²

Figure 18 Recharge zones, higher recharge rates associated with the outcropping Limestone Aquifer, lower recharge rate in areas where Cretaceous cover exists and no recharge where the Jinduckin exists.



4 Steady State Model Development The initial conceptual model described in section 2.2 was implemented using a single semi-

unconfined layer with inputs from regional/diffuse recharge (implemented as source/sink on the

upper slice of the model) and due to flows in the Katherine River and outputs as discharge along

the river features (both implemented using the transfer boundary condition applied at the nodes

along the rivers).

The 1996 late dry season water levels were considered as slightly below average and constituted a

reasonable approximation of the system in steady state.

4.1 Steady State Model Calibration

The dependent variables considered in the steady state calibration process were the hydraulic

heads and the groundwater discharge. The 1996 late dry season water levels were considered as

slightly below average and constituted a reasonable approximation of the system in steady state.

The discharge to the Katherine River was estimated from the gauged flows at the Low Level

crossing and multiplied by 1.17 the estimated dry season flow for 1996 was 1.1 cumecs.

The measure of the “goodness” of fit of the heads is the root mean square error (RMS error)

where:

RMS error = ( )

n

hhn

i∑

=

−1

2model(i)obs(i)

and

RMS error is the root mean square error (metres)

hobs(i) is the ith observed water level (metres)

hmodel(i) is the ith modelled water level (metres)

n is the number of observations

The target for calibration was to adjust the transmissivity and recharge rates to minimise the overall

RMS error and provide a discharge from the transfer boundary, in line with the estimate of

1.1 cumecs (section 2.3.6).

As stated previously the steady state model has been calibrated against the observed heads for

the late dry season of 1996.



4.2 Steady State Calibration Results

4.2.1 Calibrated Water Levels

The results of the calibrated steady state model provides a reasonable fit to the observation data,

with the RMS error for the observed versus modelled heads of 8 metres. This is an error of

approximately 6% considering the head distribution across the site ranges from 60 – 180 metres

above Australian Height Datum.

40 60 80 100 120 140 160 180 200Observed Head (mAHD)

40

60

80

100

120

140

160

180

200

Mod

elle

d H

ead

(mA

HD

)

RN005032

RN023425

RN023427

RN023428

RN024556

RN024613

RN029429

RN022287

RN020850

RN007876

RN000319

RN005898

RN032950

RN026553

RN006326

RN007821

RN022001

RN022002RN022006

RN022397

RN023424

Modelled vs Observed1:1

Confined Bores

Bores constructedin Cretaceous

Douglas River Bores

Figure 19 Steady state water levels for calibrated model.



4.2.2 Steady State Water Budget

The water budget for the calibrated steady state model is presented in Table 2.

Table 2 Steady state water balance for the calibrated model with no pumping.

Water Budget Component

Recharge [cumecs]

Storage [m3]

Discharge [cumecs]

Recharge – Total 10.17

Storage 2.91 x 1010

Discharge – Daly River

1.75

Discharge – Douglas River

2.03

Discharge – Flora River

1.97

Discharge - Katherine River

1.15

Discharge - Roper River

3.22

Discharge from Cretaceous

0.05

Total 10.17 2.91 x 1010 10.17



5 Transient Model Development

5.1 Model Specifications

The study area is roughly coincident with the model developed by Puhalovich, (2005) and covers

an area of approximately 6,400 km2 (Figure 20).

"

"

K

ATHERINE RIVER

KI

NG RIV

ER

DRY R

IVER

LIMESTONE CREEK

SCO

TT CRE

EK

E LSE Y CREEK

WATE

RHOUS

E RIV

ER

DOOK CREEK

WES

T BR

ANC

H

GUNDI CR

E EK

ROPER CREEK

BESWICK CREEK

LEIGHT CREEK

MATHISON

CR

EEK

CAVE CREEK

FE RGUSSO

N R

IVER

BONDI CREEK

EDITH RIVER

ROPER RIV ER

MCADDENS C REEK

CHAINMAN CREEK

EMIL

Y CREEK

YUJULLOWAN CREEK

SALT

CR

EEK

REDB

ANK CREEK

EMU CR

EEKDIAM

OND CREEK

MARANBOY CREEK

WURUNL UH CREEK

BONE CREEK

BULL CREEK

DURRINYAN CREEK

GRANITE CREEK

KATHERINE

Mataranka

0 10 20 30 40 505Kilometres

" TownRoadMajor RoadCreekRiverTLA model extent

CretaceousUpper Oolloo DolostoneLower Oolloo DolostoneJinduckin FormationTindall LimestoneAntrim Plateau VolcanicsUndifferetiated

²

LOCALITY MAP

Figure 20 Study area location, based on Puhalovich, (2005) TLA model extents.

5.2 Boundary Conditions

Four boundary conditions have been employed in the implementation of the transient model:

• The entire surface of the model is a variable flux boundary describing the recharge to the

aquifer.

• Transfer BC’s which simulate the groundwater/surface water interactions along the rivers.

• Well BC’s.

• No flow boundary along the edges of the model and where the Antrim Plateau Volcanics

occurs above the groundwater table.



5.2.1 Recharge (Specified Flux)

The recharge distribution was applied to the entire model based on the surface geology. In areas

were the Cretaceous unit occurred the recharge rate was reduced by a factor of 3-4 (refer to

Section 2.1.1). Also, the groundwater table has been modelled to be in direct connection to the

recharge from the surface (ie no time lag has been introduced to simulate the time for the deep

drainage to travel through the Cretaceous unit, which, can be up to 100 metres thick).

No recharge is applied to the model where the confining Jinduckin Formation is present along the

south east portion of the study area.

5.2.2 Transfer (Cauchy) Boundary

The groundwater/surface water interactions were implemented using transfer (Cauchy) boundary

conditions. As identified previously, the transfer boundary describes a reference hydraulic head

which has an imperfect hydraulic contact with the groundwater body caused by a colmation layer

(related to the stream bed conductance) (Diersch, 2004). The height of the river is considered to

“drive” the inflow/outflow of water from the Limestone aquifer.

5.2.3 Well Boundary Conditions

Discharge due to bores was modelled using Well boundary conditions with pump rate schedules

Scenario “A” and Scenario “B” as described by Puhalovich, (2005).

5.3 Transient Model Development

The initial conceptual model described in section 2.4 was implemented using a single unconfined

layer with inputs from regional/diffuse recharge (implemented as source/sink on the upper slice of

the model) and outputs, as discharge along the river features using the transfer boundary condition

applied at the nodes along the rivers.

The model was converted from a steady state model to a transient model. The main addition to

the model were the storage coefficients, the time variable recharge rates and the river stage height

along the Katherine River.

5.3.1 Temporal Discretisation

The transient simulation uses the automatic time step control in FEFLOW®, which employs the

forward Adams Bashforth / backward trapezoid time integration scheme (Diersch, 2004) using a

minimum possible time step of 0.001 days. Model inputs such as recharge and river stage height

data were applied on a daily basis.



5.4 Transient Model Calibration

As with the steady state calibration process the dependent variables considered were the hydraulic

heads and the groundwater discharge along the Katherine River where it intersects the Limestone

Aquifer. The transient model was modified to minimize the difference between the observed heads

and the modelled heads at the 5 observation bores (RN007821, RN022006, RN022397,

RN023427 and RN029429), which, are considered typical of the groundwater response in the

aquifer system.

5.5 Calibrated Model Results

5.5.1 Groundwater Level Hydrographs

The resultant calibrated hydrographs for the period 1963 – 2004 for the monitoring bores

RN007821, RN022006, RN022397, RN023427 and RN029429 are presented below.

Jan-

60

Dec

-64

Jan-

70

Jan-

75

Jan-

80

Dec

-84

Jan-

90

Jan-

95

Jan-

00

Dec

-04

100

110

120

130

140

150

Wat

er L

evel

(mA

HD

) RN007821

Figure 21 Comparison of modelled heads vs observed heads for RN007821. The response from the

calibrated model is in blue.

Jan-

60

Dec

-64

Jan-

70

Jan-

75

Jan-

80

Dec

-84

Jan-

90

Jan-

95

Jan-

00

Dec

-04

110

120

130

140

150

160

Wat

er L

evel

(mA

HD

) RN022006





Jan-

60

Dec

-64

Jan-

70

Jan-

75

Jan-

80

Dec

-84

Jan-

90

Jan-

95

Jan-

00

Dec

-04

70

80

90

100

110

120

Wat

er L

evel

(mA

HD

) RN022397



Jan-

60

Dec

-64

Jan-

70

Jan-

75

Jan-

80

Dec

-84

Jan-

90

Jan-

95

Jan-

00

Dec

-04

140

150

160

170

180

190

Wat

er L

evel

(mA

HD

) RN023427



Jan-

60

Dec

-64

Jan-

70

Jan-

75

Jan-

80

Dec

-84

Jan-

90

Jan-

95

Jan-

00

Dec

-04

100

110

120

130

140

150

Wat

er L

evel

(mA

HD

) RN029429

Figure 25 Comparison of modelled heads vs observed heads for RN29429. The response from the calibrated model is in blue.



5.5.2 Groundwater Discharge Hydrographs

The modelled and observed discharges to the Katherine River are presented in Figure 26. The

modelled response shows a reasonable fit to the observed data. It is noted that the model

generally underestimates the higher flows >2 cumecs during the periods 1974 to 1987 and 1999 to

2004. It is suspected that the deviation of the modelled discharges from the observed discharges

during very wet years is due to increased flows in Seventeen Mile Creek (Figure 14), which hasn’t

been subtracted from the flows at Galloping Jacks.

The low flow years are reasonably well predicted, which, is the aim of this model.

Jan-

1960

Jan-

1962

Jan-

1964

Jan-

1966

Jan-

1968

Jan-

1970

Jan-

1972

Jan-

1974

Jan-

1976

Jan-

1978

Jan-

1980

Jan-

1982

Jan-

1984

Jan-

1986

Jan-

1988

Jan-

1990

Jan-

1992

Jan-

1994

Jan-

1996

Jan-

1998

Jan-

2000

Jan-

2002

Jan-

2004

0.1

1

10

Kat

herin

e R

iver

Dis

char

ge(c

umec

s)

G8140001 - Measured Flow x 1.17Katherine River - Modelled Discharge

G8140001 - Gauged Flows x 1.17

Figure 26 Comparison of modelled discharge vs observed discharge along the Katherine River.



6 Discussion

6.1 Introduction

The effects of pumping on the groundwater discharge to the Katherine River due to variables such

as distance to the pumped well from the river have been developed based on hypothetical

pumping scenarios.

6.1.1 Surface-water / Groundwater Interaction Considerations

Withdrawing water from an aquifer that is directly connected to surface-water bodies can have a

significant effect on the movement of water between the groundwater and the surface water. The

effects of pumping a single well or a small group of wells on the hydrologic regime are local in

scale. However, the effects of many wells withdrawing water from an aquifer over large areas may

be regional in scale.

Withdrawing water from aquifers near surface-water bodies can diminish the available surface-

water supply by capturing some of the groundwater flow that otherwise would have discharged to

surface water or by inducing flow from surface-water into the surrounding aquifer system. A

qualitative analysis of the sources of water to a pumping well in an aquifer that discharges to a

stream is provided here to gain insight into how a pumping well can change the quantity and

direction of flow between the shallow aquifer and the stream. Although a stream is used in this

example, the results apply to all surface-water bodies, including lakes and wetlands.

A groundwater system under predevelopment conditions is in a state of dynamic equilibrium—for

example, recharge at the water table is equal to groundwater discharge to a stream (Figure 27a).

Assume a well is installed and is pumped continually at a rate, Q1. After a new state of dynamic

equilibrium is achieved, inflow to the groundwater system from recharge will equal outflow to the

stream plus the withdrawal from the well. In this new equilibrium, some of the groundwater that

would have discharged to the stream is intercepted by the well, and a groundwater divide, which is

a line separating directions of flow, is established locally between the well and the stream (Figure

27b). If the well is pumped at a higher rate, Q2, at a later time a new equilibrium is reached. Under

this condition, the groundwater divide between the well and the stream is no longer present and

withdrawals from the well induce movement of water from the stream into the aquifer (Figure 27c).

Thus, pumpage reverses the hydrologic condition of the stream in this reach from a groundwater

discharge feature to a groundwater recharge feature.

This hypothetical withdrawal of water from a shallow aquifer that discharges to a nearby surface-

water body is a simplified but compelling illustration of the concept that groundwater and surface

water are one resource. In the long term, the quantity of groundwater withdrawn is approximately

equal to the reduction in streamflow that is potentially available to downstream users.



Figure 27 In a schematic hydrologic setting where ground water discharges to a stream under natural conditions (a), placement of a well pumping at a rate (Q1) near the stream will intercept part of the ground water that would have discharged to the stream (b). If the well is pumped at an even greater rate (Q2), it can intercept additional water that would have discharged to the stream in the vicinity of the well and can draw water from the stream to the well (c).

6.2 Development Scenarios

6.2.1 Scenario with No Pumping

The results of the groundwater model calibration (section 5.5) have been used as the “baseline”

data to compare the various scenarios of development pumping.



6.2.2 Effect of Pumping Proximity on Groundwater Discharge to River

The first of these series of hypothetical pumping scenarios will be used to determine the relative

effects of pumping at various distances from the Katherine River. The three scenarios employ a

single bore pumped at 10,000 kL/d for the entire duration of the transient simulation (1963 – 2004).

The results of the three runs are presented in Figure 28.

10 100 1000 10000 100000Time[days]

0

20

40

60

80

100

Dis

char

ge D

eple

tion

Rat

e (a

s %

of p

umpi

ng ra

te)

Bore 1km from Katherine RiverBore 10km from Katherine RiverBore 20km from Katherine River

Figure 28 Effects of pumping distance on groundwater discharge to the Katherine River.

It can be seen that a bore located approximately 1 kilometre from the river has an almost

instantaneous effect on the discharge to the river with a decrease in the discharge of

approximately 60% of the pumping rate after only 30 days.

Increasing the distance between the bore and the river to 10 kilometres results in a time lag of

between 80 – 90 days before the effects of pumping are observed at the river. The results

presented here follow a power rule where the time lag for the effects of the bore to be “felt” by the

river is proportional to the square of the distance between the bore and the river.

In the above example the effect of the pumping at 10 km is 80 days

80 days = x * (10)2

Therefore x = 0.8

For the bore at 20 km the expected time for the effect to be observed is:

0.8 * (20)2 = 320 to 360 days



6.2.3 Effects of Current (2004) Licensed Entitlements on Groundwater Discharge to the Katherine River

The first of the pumping scenarios employed by Puhalovich, (2005) has been considered.

Scenario “A” was applied to the model to predict the effect on the end of dry season flows in the

Katherine River. Each scenario was run from 1963 to 2004 with

Scenario “A” considers the current pump licensing as of November 2004

Scenario “B” considers the current pump licensing (Nov. 2004) and includes new applications

These two pumping scenarios were then compared to a “baseline” scenario where no pumping

was applied.

"

Venn HorticulturalSubdivision

KATHERINE

RN040001

RN034125

RN033830

RN033804

RN033728

RN033343

RN033275

RN033263

RN032681

RN032614

RN032417

RN031879

RN031865

RN031746

RN031625

RN031332

RN031105

RN031104RN031068

RN030947

RN030946

RN030697

RN030662

RN030648

RN030493

RN029739

RN029243

RN029217RN028900

RN028879

RN028348

RN027286 RN026360

RN026356

RN026308

RN026142RN026139

RN026083

RN025975RN025974

RN025768

RN025633

RN025434

RN025089

RN024951

RN024868

RN024724

RN023640RN023638

RN022836

RN022487

RN022398

RN022391

RN022286 RN022027RN022025

RN022001

RN021421

RN021170

RN021096

RN005042

RN005032

RN002377

RN001440

0 2 4 6 8 101Kilometres

" TownRoadMajor RoadCreekRiverCadastre

CretaceousGraniteJinduckinTindall

²

Figure 29 Production bore locations for Scenario “B”

6.2.4 Steady State Analysis Scenario “A” Current - (2004) Pumping

The water budget for Scenario “A” is presented compared with the water budget where no pumping

occurs. The steady state model identifies the long term effects that would be expected over the

entire Daly Basin.



Table 3 Steady state water balance for the calibrated model with no pumping and with pumping using Scenario “A”

Water Budget Component

No Pumping [cumecs]

Scenario A [cumecs]

Recharge – Total 10.17 10.17

Discharge - Wells 0.00 0.62

Discharge – Daly River

1.75 1.75

Discharge – Douglas River

2.03 2.03

Discharge – Flora River

1.97 1.96

Discharge - Katherine River

1.15 0.61

Discharge - Roper River

3.22 3.16

Discharge from Cretaceous

0.05 0.04

Imbalance 0 0

To demonstrate the changes in the water budget components the tabulated results are presented

as a bar graph (Figure 30). It can be seen that the greatest impact due to the pumping is to

reduce discharge to the Katherine River, with the balance of the flow reductions observed in the

Roper and Flora Rivers.

Recharge Wells Douglas Flora Katherine RoperWater Balance Component

-5

0

5

10

15

Wat

er F

lux

[cum

ecs]

0.0% 0.8%

47.0%

1.9%

No PumpingScenario A

Steady State Water Balance

Figure 30 Comparison of the steady state water balance components for no pumping and

Scenario “A”.



6.2.5 Transient Analysis of Scenario “A” – 2004 (Current) Pumping

Scenario “A” is the same as that used by Puhalovich, (2005). The initial head distribution for the

simulation employed the Scenario A steady state head conditions (section 6.2.4). The monthly

pumping rates in ML/month as documented by Puhalovich, (2005), were converted to kL/day and

applied to the model over the period 1963 - 2004. The total extraction for this scenario is

19,861 ML/yr (0.63 cumecs), which, is approximately 47% of the average late dry season flows in

the Katherine River.

The results of this scenario are presented in Figure 31.

Jan-

60

Jan-

62

Jan-

64

Jan-

66

Jan-

68

Jan-

70

Jan-

72

Jan-

74

Jan-

76

Jan-

78

Jan-

80

Jan-

82

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

0

0.2

0.4

0.6

0.8

1

Dis

char

ge (c

umec

s)

Total Pump DischargeReduction in River Discharge

0

10

20

30

40

50

% D

iffer

ence

% difference at late dry season low flows% difference between flows

1

10

Dis

char

ge (c

umec

s)

Discharge - No PumpingDischarge - Scenario "A"Late Dry Season Flow

Scenario "A"

Figure 31 Reduction in the dry season flows in the Katherine River due to extraction based on pumping scenario “A”.

6.3 Bore Capture Zones

The bore capture zone is the surface and subsurface area surrounding a water bore or borefield,

through which contaminants are reasonably likely to move toward and reach the water bore or



borefield within a specified period of time. The delineation of well capture zones is a basic

component of ground water protection. The conventional methodology for capture zone

delineation is backward advective particle tracking, often applied under the assumption of a two-

dimensional aquifer.

The particle tracking computation methods are based on the Darcian velocity distributions

determined from the steady state head distribution (Anderson and Woessner, 2002).

lhKVd δ

δε

.=

where

Vd = darcy velocity

K = hydraulic conductivity

ε = specific yield

lh

δδ

= groundwater gradient

This technique can provide point related information about groundwater age in the form of

isochrones, which are often used to delineate well capture zones. It should be noted that particle

tracking simulates advective transport and neglects to include dispersion processes.

The hydraulic head determined from the steady state simulation is independent of porosity,

however, as noted above, to determine the particle track of the plume migration using the Darcy

velocity a specific yield of 0.04 and aquifer thickness of 150 meters were employed.

The simulations have been presented in isochrones (the distance covered by a “particle” for a

given time) to show the migration of multiple water “particles”. The capture zones are presented in

Figure 32 as shaded zones representing 365, 730, 1095, 1460 and 1825 days.

The model developed here is for a homogeneous and isotropic aquifer system, however, the actual

Limestone Aquifer is karstic in nature and preferential pathways are known to exist. It is therefore

expected that the actual capture zones may be much more extensive than those presented.

Statistical analysis of the effects of preferential pathways on the flow paths and travel times could

be investigated with the inclusion of fracture features.



0 1 2 3 4 50.5Kilometres

Production BoresRoadMajor RoadRiverEphemeral Drainage

All Other TenureCommonwealth of AustraliaCrownNorthern Territory of AustraliaPower and Water AuthorityPower and Water Corporation

²

Scenario "A"Isochrone [days]

365

730

1095

1460

1825

Figure 32 Particle tracking isochrones, NTG and Commonwealth land holdings are identified for comparison.



7 Conclusions Initial modeling of the Limestone Aquifer indicates that the conceptual model is largely valid,

however;

• The mapped occurrence of the Limestone Aquifer in the area of the Flora River has been

identified as being incorrect. Initial re-examination of the geological information in the area

indicates that the mapped geology used in the modelling is incorrect.

The regional model has been developed to the level that provides the boundary conditions for the

more detailed modelling in the Katherine study area.

The recharge distribution and the proportionate values proposed within this report are indicated to

be plausible. Improved definition of rainfall distribution over the basin will enable refinement of this

aspect of the model.

The modelling of the development Scenario A indicate the primary effect of groundwater extraction

in the Katherine area is to reduce flows in the Katherine River (refer Table 3).



8 Recommendations It is recommended that the following work is undertaken in terms of data capture to improve

definition and calibration of the model:

• Strategic groundwater monitoring points in the Tindall Limestone are required, this item will be

addressed by the NWI funded project to develop Daly Basin Integrated Hydrologic Model /

Integrated Hydrologic Monitoring Network;

• Locate and measure spring discharges and productive zones along the river;

• Development of a methodology to utilise the model to inform the planning / allocation process.



9 References Anderson, M. P. and Woessner, W. W., (2002), Applied groundwater modelling: simulation of flow

and advective transport, Elsevier (USA)

Diersch, H.-J.G., (2004), FEFLOW® 5.1 Users Manual, WASY - Institute for Water Resources

Planning and System Research, Berlin, Germany.

Freeze, R.A. and Cherry, J.A., (1979), Groundwater, Prentice Hall of Australia Pty. Limited, Sydney

Jolly, P.J., (2002), Daly River Catchment Water Balance, Technical Report No. 10/2002

Jolly, P.J., George, D., Jolly, I., and Z Spiers, (2000), Analysis of Groundwater Fed Flows for the

Flora, Katherine, Douglas and Daly Rivers, Technical Report No. 36/2000

Pidsley, D., (1987), Investigation and Development of Groundwater Irrigation Source for Lot 3252,

Venn Horticultural Subdivision, Power and Water Authority, Water Resources Group,

Technical Report No. 11/1987

Puhalovich, A., (2005), Groundwater Modelling of the Tindall Limestone Aquifer, Report for the

Department of Infrastructure, Planning & Environment (NT), Technical Report No. 02/2005

Tickell, S.J., (2002), Water Resources of the Oolloo Dolostone, NTG Report No. 17/2002

Tickell, S., Cruikshank, S., E Kerle, E. and Willis, G., (2002), Stream Baseflows in the Daly Basin,

Department of Infrastructure, Planning & Environment, (NT), Technical Report No. 36/2002

Tickell, S., (2003), Water Resource Mapping of the Barkly Tablelands, Natural Systems Division

Department of Infrastructure, Planning & Environment (NT), Technical Report No. 23/2003

Water Studies, (2001), Regional Groundwater Impact Modelling, Venn Agricultural Area, Report for

the Northern Territory Land Corporation, Report No. WSDJ00188/DF1

Yin Foo, D., (1985), Katherine Groundwater Investigations 1984, Cretaceous Sediments Near King

River. Water Resources Division, Department of Mines & Energy, Technical Report No.

03/1985.

Yin Foo, D., and Matthews, I., (2001), Hydrogeology of the Sturt Plateau: 1:250,000 Scale Map

Explanatory Notes, Department of Infrastructure, Planning & Environment, Darwin, NT,

Technical REPORT 17/2000D

regional groundwater modelling of the cambrian … groundwater modelling of the cambrian limestone...

Documents