cr 2010-211 upper motueka catchment finalicm.landcareresearch.co.nz/knowledgebase... · gns science...

Project Number: 631W0403

DISCLAIMER

This report has been prepared by the Institute of Geological and Nuclear Sciences Limited (GNS Science) exclusively for and under contract to Motueka Integrated Catchment Management (ICM) Programme, Landcare and Tasman District Council. Unless otherwise agreed in writing by GNS Science, GNS Science accepts no responsibility for any use of, or reliance on any contents of this Report by any person other than Landcare and shall not be liable to any person other than Landcare, on any ground, for any loss, damage or expense arising from such use or reliance.

The data presented in this Report are available to GNS Science for other use from September 2010.

BIBLIOGRAPHIC REFERENCE

Hong, T.; Minni, G.; Ekanayake, J.; Davie, T.; Thomas, J.; Daughney, C.; Gusyev, M.; Fenemor, A.; Basher, L. 2010. Three-Dimensional Finite-Element Transient Groundwater-River Interaction Model in a Narrow Valley Aquifer System of the Upper Motueka Catchment, GNS Science Consultancy Report 2010/211. 82p.

Confidential 2010

GNS Science Consultancy Report 2010/211 i

CONTENTS PREFACE ............................................................................................................................... III 1.0 INTRODUCTION ..........................................................................................................1 2.0 PURPOSE AND SCOPE OF PROJECT......................................................................1 3.0 DESCRIPTION OF THE STUDY AREA ......................................................................2

3.1 Location and area.......................................................................................................... 2 3.2 Climate .......................................................................................................................... 2 3.3 Hydrogeology ................................................................................................................ 2 3.4 Soils............................................................................................................................... 3

4.0 TRANSIENT GROUNDWATER-RIVER MODEL FOR THE UPPER MOTUEKA CATCHMENT...............................................................................................................4 4.1 General features of the three-dimensional finite-element groundwater flow modelling

approach........................................................................................................................ 4 4.2 Model grid and boundaries............................................................................................ 5 4.3 Hydraulic characteristics and aquifer properties ........................................................... 6 4.4 Rainfall and irrigation scheduling recharge model ........................................................ 6 4.5 Groundwater abstraction model .................................................................................... 8 4.6 River model ................................................................................................................... 9 4.7 Side river model............................................................................................................. 9 4.8 Hill slope recharge model.............................................................................................. 9

5.0 CALIBRATION OF THE MODEL...............................................................................13 6.0 SUMMARY .................................................................................................................15 7.0 REFERENCES ...........................................................................................................17

FIGURES Figure 1. Location map showing the extent of the groundwater model of the Upper Motueka River

catchment. .......................................................................................................................................19 Figure 2. Upper Motueka River catchment showing locations of river monitoring stations..............................20 Figure 3. Annual rainfall isohyets for the Upper Motueka valley......................................................................21 Figure 4. Simplified geological map of the Upper Motueka River Catchment, showing locations of

groundwater monitoring sites. ..........................................................................................................22 Figure 5. Two-dimensional view of the Upper Motueka groundwater model domain, showing aquifer

discretization using a finite-element mesh. ......................................................................................23 Figure 6. Aquifer discretization and finite-element mesh at the area of the confluence of the Tadmor

River and Motueka River..................................................................................................................24 Figure 7. Aquifer discretization and finite-element mesh at the area of the confluence of the Motueka

River and Motupiko River.................................................................................................................25 Figure 8. Aquifer discretization and finite element mesh in the upper valley of the Motueka River. ................26 Figure 9. Aquifer discretization and finite element mesh in the Motupiko River valley. ...................................27 Figure 10. Aquifer discretization and finite-element mesh in the Motupiko River valley. ...................................28 Figure 11. Three-dimensional view of the ground elevation structure from the bottom of the aquifer,

showing aquifer discretization by finite-element mesh.. ...................................................................29 Figure 12. Distribution of hydraulic conductivity values before the model calibration process. .........................30 Figure 13. Estimation of actual evapotranspiration according to the soil water deficit.......................................31 Figure 14. Distribution of three different rainfall recharge zones and their irrigated areas. ...............................32 Figure 15. Daily time series of rainfall and calculated recharge in the P1 recharge zone. ................................33 Figure 16. Daily time series of rainfall and calculated recharge in the P2 recharge zone. ................................34 Figure 17. Daily time series of rainfall and calculated recharge in the P3 recharge zone. ................................35 Figure 18. Locations of groundwater abstraction wells in the Upper Motueka Catchment. ...............................36 Figure 19. Total estimated groundwater abstraction and direct surface water abstraction for the period

from 1st July 2001 to 30th June 2003. ...............................................................................................37 Figure 20. Location of river cross-section surveys and river monitroing sites in the Upper Motueka

Catchment........................................................................................................................................38 Figure 21. Six selected river cross-sections in the Upper Motueka Catchment. ...............................................39 Figure 22. Cross-sections in the Motupiko River...............................................................................................40

Confidential 2010

GNS Science Consultancy Report 2010/211 ii

Figure 23. Cross-sections in the Motueka River................................................................................................41 Figure 24. River width and elevation of lowest part of river bed in the Motueka River. .....................................42 Figure 25. River width and elevation of lowest part of river bed in the Motupiko River. ....................................42 Figure 26. Depiction of the interaction between river and aquifer as modelled by the head-dependent,

third-kind Cauchy’s boundary condition. ..........................................................................................43 Figure 27. Locations of four side river recharge sources implemented in the model.........................................44 Figure 28. Daily time series of side river recharge at four sites implemented in the model...............................45 Figure 29. Location of eleven hill slope recharge zones implemented in the model..........................................46 Figure 30. Locations of the A, B and K hill slope recharge zones implemented in the model. ..........................47 Figure 31. Locations of the C, D, E and J hill slope recharge zones implemented in the model. ......................48 Figure 32. Location of the F, G, H and I hill slope recharge zones implemented in the model..........................49 Figure 33. Korere hill slope recharge boundary and experimental site. ............................................................50 Figure 34. Locations of capacitance probes and TDRs along the slope to measure the phreatic level at

Korere site........................................................................................................................................50 Figure 35. Maximum and minimum phreatic levels at Korere site during 2006-2007. .......................................51 Figure 36. Maximum and minimum phreatic levels at Paratiho site during 2006-2007. ....................................51 Figure 37. Approximation of the water flow at the toe of a hillslope by two dimensional flow net analysis........52 Figure 38. Measured and estimated hillslope phreatic levels for summer and winter seasons at the

Paratiho site. ...................................................................................................................................52 Figure 39. Measured phreatic level recession curves for summer and winter at the Paratiho and Korere

hill slope sites...................................................................................................................................53 Figure 40. Scaling factor (S) represents the ratio of catchment area to hill slope boundary length...................53 Figure 41. Ground water recharge at Korere hill slope site. ..............................................................................54 Figure 42. Ground water recharge at Paratiho experimental site. .....................................................................55 Figure 43. Daily time series of hill slope recharge at sites A to F as implemented in the model. ......................56 Figure 44. Daily time series of hill slope recharge at sites G to K as implemented in the model.......................57 Figure 45. Contour map of initial groundwater level used to initiate calibration of the Upper Motueka

groundwater-river interaction model.................................................................................................58 Figure 46. Contour map of initial groundwater hydraulic head used to initiate calibration of the Upper

Motueka groundwater-river interaction model. .................................................................................59 Figure 47. Observed mean daily river stage at five river monitoring sites. ........................................................60 Figure 48. Intermediate river sites defined in the lower part of the model to improve simulation of river

stage. ...............................................................................................................................................61 Figure 49. Intermediate river sites defined in the upper part of the model to improve simulation of river

stage. ...............................................................................................................................................62 Figure 50. Calibrated hydraulic conductivity distribution in the lower part of Upper Motueka

groundwater-river interaction model. Insert map shows full scale of the model domain. .................63 Figure 51. Calibrated hydraulic conductivity distribution in the upper part of Upper Motueka

groundwater-river interaction model. Insert map shows full scale of the model domain. .................64 Figure 52. Model calibration results for groundwater level at Quinney’s Bush. .................................................65 Figure 53. Model calibration results for groundwater level at North Bridge. ......................................................65 Figure 54. Model calibration results for groundwater level at Crimps................................................................66 Figure 55. Model calibration results for groundwater level at Hyatts. ................................................................66 Figure 56. Model calibration results for groundwater level at Vue Mount..........................................................67 Figure 57. Observed and modelled river flow loss and gain for 9th February 2002.. .........................................68 Figure 58. Modelled daily river gain and loss over the entire study area...........................................................69

TABLES Table 1. Summary of hydrological monitoring sites in the model domains.....................................................71 Table 2. Hydraulic characteristics of the aquifer in the Upper Motueka Catchment. ......................................72 Table 3. Statistical summary of the rainfall recharge models developed for the period from 1st July

2001 to 30th June 2003. ...................................................................................................................73 Table 4. Groundwater abstraction wells in the model domain........................................................................74 Table 5. Summary of river cross-section surveys in the Motueka, Motupiko, and Tadmor Rivers. ...............75 Table 6. Flow information for the period of 1st July 2001 to 30th June 2003 for four side river recharge

sources implemented in the model. .................................................................................................76 Table 7. Initial groundwater levels used to initiate calibration of the Upper Motueka groundwater- river

interaction model..............................................................................................................................76 Table 8. Summary of calibration results for five groundwater monitoring wells..............................................77

Confidential 2010

GNS Science Consultancy Report 2010/211 iii

PREFACE

An ongoing report series, covering components of the Motueka Integrated Catchment Management (ICM) Programme, has been initiated in order to present preliminary research findings directly to key stakeholders. The intention is that the data, with brief interpretation, can be used by managers, environmental groups and users of resources to address specific questions that may require urgent attention or may fall outside the scope of ICM research objectives.

We anticipate that providing access to environmental data will foster a collaborative problem-solving approach through the sharing of both ICM and privately collected information. Where appropriate, the information will also be presented to stakeholders through follow-up meetings designed to encourage feedback, discussion and coordination of research objectives.

Confidential 2010

GNS Science Consultancy Report 2010/10 1

1.0 INTRODUCTION

The area upstream of the Wangapeka confluence of the Motueka River catchment has a sizeable area of fertile alluvial river terrace land that is suitable for irrigated agriculture. Since the mid 1990’s there has been an increasing demand for irrigation water especially from groundwater in these terraces.

Tasman District Council (TDC), water users, and developers are confronting questions about the effects of increasing groundwater abstraction on groundwater levels and stream flow in the Upper Motueka catchment. Due to strong aquifer-surface water interactions in this region, there is concern that abstraction of more groundwater for irrigation may lower groundwater levels and consequently reduce the baseflow of streams during the summer season. Increased groundwater abstraction could therefore reduce the availability and life-supporting capacity of surface water for aquatic life, recreation and other uses. It could also add extra pressure in terms of water allocation and maintenance of minimum baseflow for the Motueka River and tributaries.

2.0 PURPOSE AND SCOPE OF PROJECT

This study is conducted in cooperation with Landcare Research and TDC as a component of the Integrated Catchment Management (ICM) project coordinated by Landcare Research. The purpose of this study is to gain a better understanding of the relationships and interactions between the shallow aquifer and surface water systems in the Upper Motueka catchment so that defensible water management policies can be put in place.

The project scope involves development of a three-dimensional finite-element transient groundwater-river interaction model. The model will assist with management of the water resources in this region in a holistic manner, for example to analyse the effects of more groundwater abstraction on stream flow and groundwater levels.

The specific objectives of this study were:

• To develop a transient river flow model by using surveyed river cross section data and river flow gauging data from the Motueka and Motupiko Rivers provided by TDC;

• To develop a transient rainfall-recharge model in the Upper Motueka groundwater-river interaction model;

• To develop a transient model of hillslope recharge in the Upper Motueka groundwater-river interaction model;

• To generate a transient groundwater abstraction model for the Upper Motueka groundwater-river interaction model;

• To calibrate the Upper Motueka groundwater-river interaction model (incorporating the river model structure and rainfall-recharge model), using historical data sets of groundwater and river levels; and

• To use the calibrated model to simulate the effects of different water and land management scenarios on the water resources of the Upper Motueka catchment (to be reported later).

Confidential 2010


3.0 DESCRIPTION OF THE STUDY AREA

3.1 Location and area

The Motueka Catchment is located in the northwest part of the South Island of New Zealand (Figure 1). The catchment drains an area of 2180 km2 and is dominated by mountains and hill country, with about 67% of the catchment area having land surface slopes greater than 15°.

The study area is located in the upper part of the Motueka Catchment (Figure 2) and has an area of 887 km2. The Upper Motueka Catchment is composed of three main river valleys: the Motupiko River (344 km2), the Tadmor River (124 km2) and the Motueka River (419 km2). The main morphological features of the Upper Motueka River catchment are steep, narrow headwater channels and broad floodplains and terrace systems within hilly Moutere gravel terrain below the Upper Motueka Gorge to the Wangapeka River confluence. The main stem of the Motueka River flows north for about 110 km to the sea, and is joined from the west by a series of generally much larger tributaries, which drain both hill terrain on Moutere gravel (e.g. Motupiko and Tadmor catchments west of the study area) and mountainous terrain underlain by a complex assemblage of sedimentary and igneous rocks (e.g. Wangapeka and Baton catchments downstream of the study area).

3.2 Climate

Mean annual rainfall for the catchment is estimated at 1600 mm. There is a strong spatial pattern of rainfall variation, correlated to topography. However, the rainfall distribution does not vary dramatically within the study area (Figure 3), although there is a decreasing gradient both northwards and southwards away from the Tapawera gauge. At the margin of the study area, the annual precipitation is approximately 14% higher than at Tapawera.

Annual open pan evaporation downstream at Motueka is 1106 mm and is strongly seasonal, with mean monthly values ranging from 27 mm in July to 179 mm in January. Evaporation is expected to be slightly lower at Tapawera. While annual evaporation is less than annual rainfall, soil moisture deficits are common in summer when evaporation exceeds rainfall and irrigation is required for many crops in the catchment.

3.3 Hydrogeology

Groundwater in the Upper Motueka River catchment is abstracted from shallow unconfined alluvial aquifers that occur in the Quaternary river terrace formations and modern river deposits. In hydrogeological terms, these aquifers are considered as a single unit because they are hydraulically highly connected. Five gravel formations have been identified within the study area upstream of the Wangapeka River confluence (Figure 4). These are (from oldest to youngest) the Moutere Gravel, Manuka, Tophouse, Speargrass, and modern river gravel formations. The Quaternary gravels are underlain by the Moutere Gravel Formation throughout the whole study area (Stewart et al., 2003).

The Moutere Gravel Formation consists of rounded greywacke clasts up to 0.6 m diameter (most less than 0.2 m diameter) in a yellowish-brown, silty, clay matrix. The formation contains minor clasts of very weathered ultramafics in the Motueka River catchment upstream of the Motupiko River confluence. Moutere gravel is widespread throughout the

Confidential 2010


Upper Motueka River catchment and forms the hill country between the valleys of the Motueka, Motupiko, and Tadmor rivers (Johnston, 1983; Stewart et al., 2003).

Manuka Formation surfaces lie between approximately 65 m and 70 m above the river level, and terrace remnants are numerous in the Motupiko and Tadmor River valleys. No Manuka Formation surfaces are recognised in the Motueka River valley, except for small remnants at the confluence of the Motupiko and Tadmor rivers. The thickness of the Manuka Formation is estimated to be between 60 and 100 m. The saturated thickness is estimated to be between 0 and 20 m. No bores have been drilled into the formation to confirm whether it is groundwater bearing but it is likely to be less permeable than the younger lower level Speargrass Formation due to the greater degree of weathering and higher clay content. The Manuka Formation is likely to receive a component of groundwater recharge from the Moutere Gravels where they are in direct contact.

The Tophouse Formation is widespread throughout all of the river valleys in the Upper Motueka River catchment. Tophouse Formation surfaces lie approximately 17 m to 35 m above river level. The potential saturated thickness of the Tophouse Formation is estimated to be between 0 and 12 m. Bores drilled into the formation just south of Stanley Brook Hill and just north of Kohatu were abandoned due to insufficient groundwater. This suggests that either the permeability of the formation is very low or the base of the Tophouse Formation is above the groundwater level in these areas. The top surface of the Tophouse Formation is typically between 10 and 20 m above the top surface of the Speargrass Formation (Stewart et al., 2003).

The Speargrass Formation is widespread in the upper reaches of the valleys but absent in the lower reaches. The formation forms the lowest terrace at approximately 8 m above the river level. An aggradation surface occurs on the Speargrass Formation terrace that is approximately one to two metres higher than the degradation surface. Groundwater is abstracted from the Speargrass Formation within the study area. The average saturated thickness of the Speargrass Formation is estimated to be between 5 and 8.5 m.

Thin modern gravel deposits of Holocene age form the floodplains of the Motueka River and its tributaries. The deposits are more extensive in the lower reaches of the valleys. The composition is similar to the older quaternary river formations, except the modern gravels are better sorted and tend to lack clay (Johnston, 1983). The saturated thickness of the modern gravels ranges from 5.5 to 9 m. These modern gravels supply most of the groundwater that is abstracted in the Upper Motueka River valley.

3.4 Soils

Soil information is derived from a regional soil survey (Chittenden et al., 1966). The most common soils in the area are Tapawera, Motupiko and Atapo loams. There are also small areas of Dovedale, Sherry and Kikiwa loams and Korere and Spooner soils. The distribution of the various soil types within the study area is discussed in Section 4.4.

Tapawera sandy loam soils are formed on the young floodplain. These are well drained soils with high potential rooting depth and no slowly permeable zones in the soil, moderate permeability in the topsoil and rapid in the subsoil, moderate infiltration, high profile available water (220 mm), and moderate levels of stones in the profile (17%). They are rated as class

Confidential 2010


2s for irrigation suitability (cf. Griffiths, 1975), with permeability and stoniness providing the greatest limitations for irrigation.

Motupiko loams are found on slightly older terrace surfaces. These are well drained soils with high potential rooting depth, slowly permeable deep subsoil (below 1 m), moderate permeability in the topsoil and slow permeability below 1 m, moderate infiltration, high profile available water (220 mm), and low levels of stones in the profile (7%). They are rated as class 3s for irrigation suitability with permeability the greatest limitation for irrigation.

Atapo stony silt loam soils are formed on intermediate age terraces. They are well drained soils with high potential rooting depth and no slowly permeable zones in the soil, moderate permeability in the topsoil and rapid in the subsoil, moderate infiltration, moderate profile available water (105 mm), and higher amounts of stones in the profile (24%). They are rated as class 3s for irrigation suitability with stoniness the greatest limitation for irrigation.

There are also small areas of other soil types. Dovedale gravelly loam on slightly older terrace surfaces is a shallow soil with rapid permeability and moderate rooting depth. It is rated as class 3s for irrigation suitability with rooting depth the greatest limitation for irrigation. Sherry sand and sandy loam is formed from coarse granitic alluvium. It is rated as class 3s for irrigation suitability with rooting depth and permeability the greatest limitations for irrigation. Kikiwa silt loam and rolling phase are found on high terraces and are moderately and marginally suitable for irrigation, respectively. Slope steepness and the presence of slowly permeable subsoils are the greatest limitations for irrigation. Korere and Spooner soils are found on the hilly Moutere gravel terrain and are unsuitable for irrigation because of slope steepness.

4.0 TRANSIENT GROUNDWATER-RIVER MODEL FOR THE UPPER MOTUEKA CATCHMENT

4.1 General features of the three-dimensional finite-element groundwater flow modelling approach

Modelling requires analysis of water flux interchanges between shallow aquifers and stream systems across the entire study area. Over the years, many methods have been devised to model such interactions between shallow aquifers and streams, ranging from relatively simple analytical methods to site-specific three-dimensional numerical models. The simpler modelling approaches cannot accurately represent the complex hydrogeological environments found in the Upper Motueka River catchment nor address the spatial requirements for water management.

In this study, a state-of-the-art finite-element groundwater model is developed using FEFLOW software (Diersch, 2005) to simulate the relationships and interactions between shallow aquifers and stream systems in the Upper Motueka River catchment. The main advantages of using a finite-element groundwater modelling system are that a Galerkin finite element structure for mesh generation can be used to define the model grid in areas where intensive interactions are likely to occur between shallow aquifers and streams, and fully adaptive mesh refinement or derefinement can be used to enhance the reliability of the numerical solution.

Confidential 2010


4.2 Model grid and boundaries

The model domain incorporates the Motueka River from 3 km downstream of the confluence with the Tadmor River to approximately 8 km upstream of the confluence with the Motupiko River (Figure 5). The model domain also incorporates 3 km of the Tadmor River upstream of the confluence with the Motueka River, and approximately 15 km of the Motupiko River upstream of the confluence with the Motueka River. The model boundaries encompass almost all of the areas of groundwater abstraction in the Upper Motueka River catchment. The model boundaries are therefore appropriate for simulating the effects of groundwater abstraction on stream flows and groundwater levels.

Figure 5 shows the full extent of the Upper Motueka groundwater flow model domain with aquifer discretization using a finite element mesh. The aquifer is discretized using one layer of 6-node triangular prisms, with 29299 nodes and 30386 elements of varied vertical thickness comprising the entire model domain. After the finite-element mesh was generated by the FEFLOW software, the mesh structure in some parts of the model domain was refined to improve simulation of the interactions between aquifers and rivers. Mesh refinement is easily performed in the FEFLOW program, and it enables the finite-element grid to be made denser (i.e. elements are smaller and more numerous) in areas where it is more important to model rapid changes in water flux or steep hydraulic gradients, for example, near the area of the confluence of the Motueka and Motupiko rivers.

Figures 6 to 10 show the extent of aquifer discretization at the confluence of Tadmor River and Motueka River, the confluence of the Motupiko River and Motueka River, and the upstream areas of the Motupiko River and Motueka River, respectively. In Figures 6 to 10, rivers are displayed in blue, and the size of the mesh representing the rivers is specified to correspond to locations of the measured river cross sections. For example, the measured (and modelled) cross section of the Tadmor River is approximately 23 m wide and it is represented by a fine mesh because it is narrow compared to other rivers in the study area, particularly the Motueka River. The mesh for the Motueka and Motupiko rivers is coarser due to the wider measured river cross sections. Within the model domain, the mesh is much finer in areas where aquifer-river interaction occurs. Table 1 summarises locations of all river flow gauging sites and groundwater reference sites used during the calibration process.

The three-dimensional ground elevation model for the Upper Motueka groundwater-river interaction model is shown in Figure 11. Ground elevation is generated using a 20 m resolution DTM data set (Stewart et al., 2003). The modelled ground elevation ranges from 110 m above sea level (asl) at the downstream model boundary to 240 m asl at the Goldpine river gauging site and 305 m asl upstream of the Korere river gauging site.

The three-dimensional view of the ground elevation model shows the valley structure throughout the model domain (Figure 11). For example, the ground elevation in the area below the Tapawera Bridge ranges from approximately 113 m asl in the centre of the valley to 155 m asl on the valley sides. Likewise, the ground elevation across the valley bottom at the confluence of the Motupiko River and the Motueka River ranges from approximately 185 to 200 m asl. The average ground elevation at the upstream ends of the Motupiko and Motueka rivers in the model is approximately 310 m asl and 245 m asl, respectively. These valley structures and the hill slopes beyond imply that hill slope recharge and side valley recharge may be important components of the water balance in the modelled area.

Confidential 2010


Aquifer thickness is defined based on the previous hydrogeological work (Stewart et al., 2003). The smallest aquifer thickness occurs in the northern part of the model (from Wangapeka to Tapawera) and is estimated to be 9 to 10 m. Greater aquifer thickness occurs in the southern part of the model (Kohatu to Goldpine/Korere), where the aquifer is estimated to be 14 to 15 m thick. The bottom slice of the model represents the bottom of the aquifer; modelled aquifer thickness is shown in Figure 11. The top of the aquifer is modelled as a freely movable surface that represents the water table. The adjustment of the moving water table surface is achieved in FEFLOW as follows: (1) compute groundwater head, (2) determine the new free surface of the water table according to computed groundwater head, and (3) compute groundwater head at points between the water table and the bottom of the aquifer.

4.3 Hydraulic characteristics and aquifer properties

In order to simulate a groundwater flow system with FEFLOW, the hydraulic characteristics of the aquifer and confining beds must be specified for each of the nodal points. The hydraulic characteristics normally required to simulate a groundwater flow system are aquifer thickness and hydraulic conductivity (K). The aquifer thickness in the model domain is described in Section 4.2. Hydraulic conductivity values were provided by TDC, based on analysis of slug tests and constant rate pump tests (Table 2). At some localities where no pump tests have been undertaken, the hydraulic conductivity was assumed to be the same as at the closest pump tested site.

The distribution of the hydraulic conductivity values within the model (prior to calibration) is shown in Figure 12. The model domain is divided into regions, each representing a different hydraulic conductivity based on the geological map (Figure 4) and the slug and pump test results (Table 2). The distribution of horizontal hydraulic conductivity in each region is based on the Akima inter/extrapolation technique, which employs a cubic transfer function to interpolate between points where hydraulic conductivity is known. Horizontal hydraulic conductivity is assumed to be the same in the X and Y directions, but the vertical hydraulic conductivity (Z direction) is assumed to be lower. An anisotropy ratio of 10 is assumed for horizontal vs. vertical hydraulic conductivity, meaning that the vertical hydraulic conductivity at each point is assumed to be one tenth of the horizontal hydraulic conductivity value. The assumed anisotropy ratio for hydraulic conductivity is based on the typical values reported in the literature (Freeze and Cherry, 1979). Note that the hydraulic conductivity values initially assigned to each zone (as shown in Figure 12) were adjusted during the model calibration.

4.4 Rainfall and irrigation scheduling recharge model

Rainfall and irrigation recharge to the top surface of the finite element mesh was estimated using a daily soil water balance model based on daily evapotranspiration estimates and water-holding capacities for the main soils in the Upper Motueka Catchment (Ekanayake et al., 2009). The soil water balance model was based on the following equation:

renn IRAETSDSD −−+= −1 (1)

where:

Re = Effective rain (mm)

Confidential 2010


AET = Actual evpotranspiration (mm)

SDn-1 = Soil water deficit at the start of the day (mm)

SDn = Soil water deficit at the end of the day (mm)

Ir = Irrigation (mm)

Effective rain (Re) is estimated by assuming that a fraction of actual daily rainfall (R) drains through macropores if the daily rain total exceeds a predefined threshold, which is taken as 2 mm. If the daily rain total is larger than the threshold value, 90% of the total daily rain is assumed to be effective rain, and the remaining 10% of the total daily rain is assumed to be drained through macropores.

Actual evapotranspiration (AET) is a function of soil water deficit (SD). AET reaches potential evapotranspiration (PET) if SD is less than 1 mm, whereas AET = 0 if SD is greater than the readily available water capacity of the soil (RAW). For RAW < SD > 1, AET varies between 0 and PET according to the function shown in Figure 13. This shows that when SD < 0, the soil is saturated, and when SD = 1, the soil is unsaturated but is assumed to have the potential to meet PET. AET is taken as equal to PET if the soil is irrigated, regardless of the magnitude of the soil water deficit.

Soil water deficit at the start and end of the day (SDn-1 and SDn, respectively) are determined from the algebraic sum of irrigation (Ir), actual evapotranspiration (AET) and effective rainfall (Re) based on Equation 1 above.

Irrigation (Ir) is applied in the model if the soil water deficit exceeds 50% of the RAW of the soil. Five mm of water is modelled as remaining in the soil if the irrigation exceeds the soil water deficit by 5 mm, and the excess water is removed as drainage. Soil drainage takes place only if the irrigation plus any rainfall exceeds the soil moisture deficit:

Drainage = -0.1R + (SD - Ir) +5 if (SD - Ir) < -5 mm

Drainage = -0.1R if (SD - Ir) > -5 mm

where 0.1R is the macropore drainage if the rainfall exceeds the 2 mm threshold value defined above. Note that the sign of the drainage is negative.

Three different rainfall recharge zones (P1, P2, and P3) were defined based on the distribution of the dominant soil types within the study area (Figure 14, Table 3). Rainfall recharge in each zone was estimated using the daily soil water balance model described above. Each recharge zone was further divided into irrigated and non-irrigated areas. For all recharge zones (P1, P2, P3), RAW was taken as 100 mm and plant available water capacity (PAW) was taken as 220 mm for all soil types. The rainfall at each farm within each zone was scaled according to location using the annual rainfall isohyets for the Upper Motueka valley (Figure 3).

The soil moisture balance model was run with a daily time step for the period from 1st July 2001 to 30th June 2003 (Figures 15-17). Table 3 shows a statistical summary of the rainfall recharge models developed for each recharge zone. For example, the annual recharge rate

Confidential 2010


for the non-irrigated areas of P1 is estimated at 35.8% of rainfall without bypass flow, compared to 40.3% for the irrigated areas of P1. The calculated recharge for the three recharge zones is applied at every model cell in the relevant areas of the model.

4.5 Groundwater abstraction model

A groundwater abstraction model was developed to simulate the effect of groundwater abstraction on river flows and groundwater levels. Figure 18 shows the location of the 39 groundwater abstraction wells within the model domain that have water permits from the TDC. The location and consented abstraction rate for each well are shown in Table 4. Each well is positioned at the nearest node in the model grid.

Daily groundwater abstraction at each well was estimated based on irrigated land uses. For calibrating the transient river-aquifer model, estimates of actual irrigation water usage are needed. Under a conservative approach, recharge must be estimated for the groundwater extraction related to the actual irrigated areas. The Landcare Research irrigation scheduling model is used to estimate the irrigation demand and drainage for actual irrigated areas in use under the 2001-2003 climatic conditions (Ekanayake et al., 2009). Steps involved in the calculation of conservative groundwater abstraction are as follows:

1. Use the soil water balance and irrigation scheduling model (described above) based on the 2007-2008 climatic conditions to estimate the climate-driven irrigation demand. Soil water deficit was allowed to increase up to 50% of the available water capacity under the conservative irrigation scenario.

2. Estimate the actual irrigated area by comparing the estimated irrigation demand and measured water usage for 2007-2008. Only the wells with water usage records were used to calibrate the actual irrigated land areas, although their data were also used to estimate actual irrigated areas for any unmetered wells. This approach was required because it is only over the past few years that most of the consented wells within the study area have been installed with flow meters to monitor the actual water usage.

3. Use the estimated actual irrigation areas determined in Step 2 to estimate the groundwater abstraction and drainage under the climate conditions of 2001-2003. Drainage from each zone under the conservative irrigation scenario was estimated by taking the average of all the farms in each zone.

Figure 19 shows the daily time series of total estimated groundwater abstraction across all 39 wells implemented in the model for the period from 1st July 2001 to 30th June 2003, based on the conservative abstraction approach described above. The majority of the groundwater abstraction occurs from deep levels of the Quaternary gravels in the area between Tapawera bridge and the confluence of the Motupiko and Motueka rivers.

Daily peak groundwater abstraction is modelled at 30,683 m3/day (355 L/s) excluding 6,410 m3/day (74.2 L/s) of direct surface water abstraction. For the 2001-02 year, average groundwater abstraction over the nominal irrigation season October to April inclusive is 51 L/s, while for the drier 2002-03 year, average groundwater extraction is 129 L/s. Corresponding surface water extraction is 10 and 26 L/sec respectively.

Confidential 2010


4.6 River model

The river system in the model is composed of three rivers: the Motueka River, the Motupiko River and the Tadmor River. Thirty-eight river cross-sections were surveyed in these three rivers (Figure 20, Table 5). As stated above, the river cross-sections were used to determine the appropriate finite element size at corresponding locations within the model domain. River widths, depths and cross-sectional profiles vary significantly across the study area (Figures 21 to 23). For example, the average slope of the Motueka River bed between Wangapeka and Goldpine is calculated at 0.007 m/m, whereas the average slope of the Motupiko River bed is estimated to be 0.017 m/m (Figures 24 and 25).

The three rivers are implemented in FEFLOW using a head-dependent, third-kind Cauchy’s boundary condition. This Cauchy’s boundary condition allows head-dependent flow into and out of a stream, similar to the MODFLOW Stream Package (Diersch, 2005). Figure 26 shows conceptually the interaction between the river and the aquifer as modelled. Modelled stream flow is increased in areas where the river gains water from the aquifer and reduced in areas where river water seeps out of the river into the aquifer. The river bed is assumed to be sealed by a layer of thickness d and hydraulic conductivity of K0. Normally, the layer K0 is much smaller than the hydraulic conductivity K1 of the adjacent aquifer to be modelled. The flux (q) between the aquifer and the stream (or vice versa) is controlled by the difference between river stage hR and groundwater head h at each node and each time step:

d

hhKqR −

−= 20 (2)

The riverbed resistance coefficient (Φ), sometimes also called leakance coefficient, represents the sealing of the riverbed in comparison to the adjacent aquifer.

dK 0≈Φ in [d-1] (3)

It is possible to set differing leakance values for inflowing and outflowing conditions, but that has not been done in the upper Motueka model due to lack of data confirming any difference. The leakance values are one of the major variables to be adjusted during the model calibration process.

4.7 Side river model

Four small streams entering the model domain (from the hills) are defined as side river recharge sources that would influence river flow (Figure 27, Table 6): (A) Clarke River junction with Motupiko River; (B) Glen Rae stream junction with Motueka River; (C) Long Gully junction with Motupiko River; and (D) Norris Gully junction with Motueka River. The daily time series of river flow at these four side river sites was estimated and implemented in the model using a head-dependent, third-kind Cauchy’s boundary (Figure 28).

4.8 Hill slope recharge model

Eleven zones of potential hill slope recharge to the aquifer have been identified (Figures 29 to 32). This hill slope recharge has been found not to have a significant impact on river flows

Confidential 2010


or groundwater levels, compared to rainfall recharge and river recharge to the aquifer (Ekanayake et al., 2009). However, the hill slope recharge does influence total water budget, and hence the zones of potential hill slope recharge are implemented in the model using a flux boundary condition.

Capacitance probes and time domain reflectometry (TDR) soil moisture instruments were installed at the foot of the hill slopes at Korere and Paratiho to measure climate-induced and topography-driven changes in phreatic levels. Topography of the hill slope recharge interface at Korere is shown in Figure 33. Instruments were configured to measure and record hourly readings of rainfall and phreatic levels. Instrument locations at Korere are shown in Figure 34.

Measurements at the Korere and Paratiho sites (Figures 35 and 36) show that time taken for phreatic levels to fall to base flow levels exceed 30 days for larger rain events. Therefore a soil water balance model with a daily time step and piston-type instantaneous recharge (as described above) is not appropriate to estimate the hill slope recharge. Since the hydraulic gradient of the phreatic surface at the hill slope boundaries determines the recharge rate, a method which estimates the recharge from the transient phreatic levels is the most appropriate to estimate hill slope recharge.

Flow net analysis was selected to estimate the hill slope recharge. Flow net analysis is appropriate for this study because it requires much less spatial and temporal data compared to other approaches such as numerical groundwater models. A flow net is a network of stream lines and equipotential lines (Figure 37). The assumptions behind flow net analysis are as follows:

1. flow is two dimensional;

2. flow is in steady state for a given time step;

3. soil is homogenous and isotropic, and hence stream lines are perpendicular to the equipotential lines;

4. the bottom boundary is impermeable; and

5. the pattern of flow nets are drawn as approximate squares with Nd representing the number of equipotential drops and Nf representing the number of flow tubes.

Flow through a unit thickness of one square element in m3/day is given by the Darcy equation:

zxhkq Δ

ΔΔ

= (4)

Where Δh = Ht/Nd and Ht is the total head drop (m) and k is the hydraulic conductivity (m/day). Δz is the depth of an element and Δx is the potential drop across an element. Because the flow elements were constructed to have approximate square shape, Δz ≈ Δx and so:

Confidential 2010


d

t

NHkq = (5)

The total flow Q in m3/boundary length/day is given by:

d

ftf N

NkHqNQ == (6)

Hill slope phreatic level change is mainly determined by the surface and bedrock topography, antecedent moisture status, soil properties and climatic conditions. According to Equation 6, hill slope recharge is governed by the hydraulic head and the saturated hydraulic conductivity of the soil at the hill slope boundary. The parameters Nf and Nd are selected to form approximate square elements in the flow net. Hill slope water recharge to the ground water catchment takes place through an interface determined by the perched water table depth, which is typically around 3 to 4 m deep for most of the hill slopes in the study area. Recharge from the hill slope remains active as long as phreatic levels create a positive hydraulic gradient at the interface. Daily water flow across the hill slope interface can be estimated if phreatic level records are available. Climatic data are available for all 11 hill slope recharge sites, but phreatic level records are only available for the Paratiho and Korere sites and only for a limited time period.

In this study, it was necessary to develop a model to estimate the hill slope phreatic level where direct measurements were not available. Measured data at Paratiho hill slope site in the 2003-2005 summer and winter seasons were used to develop and calibrate the model equations and parameters. All other measured data at both the Paratiho and Korere sites were used to validate the model. Steps involved were as follows:

1. estimate the initial phreatic level for a given rain event;

2. apply a master recession formula to calculate the transient decline of the phreatic level; and

3. use flow net analysis (Equation 6) to estimate the transient recharge.

The initial phreatic level cannot be estimated from daily rainfall amount as change in phreatic level does not occur for all rain events. Furthermore, infiltration is controlled by many factors such as hill slope geometry, thickness of the unsaturated zone, antecedent moisture levels, and climatic variables. Application of the irrigation scheduling model (with irrigation switched off) indicates that drainage takes place for rain events large enough to overcome the soil water deficit. This suggests that infiltrated water is available to raise the phreatic level only under certain conditions.

A reasonable correlation between the drainage and daily rainfall was found as shown in Figure 38. Drainage remains relatively unchanged for large rain events due to overland flow generation. Hence for a known daily total drainage (X, in mm), the mean daily phreatic level (H, in m) can be estimated for summer (Equation 7) and winter (Equation 8) rain events as follows:

Confidential 2010


]1[5.2

)( XeSH η−−= (7)

where S is a scaling factor (described below) and η = 0.2557.

])

18.41(2

11[18.4 5.0

2

XSSH

ε⎥⎦⎤

⎢⎣⎡+

−= (8)

where S is a scaling factor (described below) and ε = 1.006.

Once the initial peak phreatic level for a rainfall event is estimated, the change in the phreatic level over time can be calculated using the master recession curve. Master recession formulas are usually derived by using a linear or power type relationship between the phreatic level elevation and phreatic level decline rate for a given site. For the present situation, we need to develop a master recession curve to represent not only one hill slope site but to cover all of the hill slopes of interest across the whole study area. Therefore, all the phreatic levels measured during winter and summer at both sites (25 km apart) were used to create a master recession formula. This was developed assuming the drainage is mainly controlled by the soil hydrology, which is region specific and independent of the seasonal variation. Identical drainage patterns of the measured recession data for all seasons at both sites suggest that this assumption is reasonable (Figure 39). The following empirical relationship was developed to predict the recession behaviour of the phreatic level:

]][

[ 5.0tSWlevel

δβ += (9)

where β and δ are found to be site and season independent and are given as follows:

]479.0029.0[ 2H+−=β (10)

]5248.0495.1[ H+=δ (11)

The β and δ parameters reflect the initial phreatic level (H) at day = 0 found in Equations 5 and 6. The time step (t) is taken as 1 day. The scaling factor (S) used in Equations 7, 8 and 9 is site dependent and represents the ratio of catchment area to interface boundary length at a hill slope site. The relationship between the scaling factor and the catchment/boundary length was developed using the data available from the Korere and Paratiho sites, and it is assumed to be linear as shown in Figure 40. The catchment area and interface boundary length for each hill slope site were manually plotted on ARC-Views GIS maps to estimate the scaling factors. Saturated hydraulic conductivities were measured at Korere and Paratiho sites up to 1.5 m depth and scaled to appropriate values according to the permeability characteristics given in GIS soil maps.

Confidential 2010


Once the transient phreatic levels were estimated, the hill slope recharge was estimated using Equation 6. For a fixed flow net geometry, recharge is proportional to the hydraulic head dictated by the phreatic level at the boundary interface. In spite of the large distance between the Paratiho and Korere sites (25 km), the combined model reproduced the phreatic levels at Korere hill slope site reasonably well. Recharge estimated using the measured phreatic levels and predicted recharges for the Korere and Paratiho sites are given in Figures 41 and 42.

Predicted daily hill slope recharge along boundaries A to K is shown in Figures 43 and 44. These are input into the groundwater model at hill slope recharge boundary nodes as prescribed flux boundary conditions.

5.0 CALIBRATION OF THE MODEL

The observation data of the groundwater level records, rainfall and the river flow shows that the aquifer and river systems are highly dynamic. Hence the Upper Motueka groundwater-river interaction model was calibrated as a transient model using a daily time step. The model was calibrated using visual and numerical comparison of the predicted and observed groundwater levels at five groundwater monitoring wells (Quinney’s Bush, North Bridge, Crimps, Hyatts and Vue Mount) (Figure 5 and Table 1).

The following parameters were adjusted during the calibration process:

1. river stage between locations of gauging measurements;

2. spatial distribution of hydraulic conductivity;

3. value of the transfer coefficient between river and aquifer, also known as the river bed leakance coefficient (a single global value was assumed for the whole model domain).

Generally the groundwater levels in the model are sensitive to a change in the bottom elevation of the unconfined aquifer. In our study, the bottom elevation of the aquifer was not changed during the calibration process because these values are relatively well known.

The implementation of accurate river stage in a daily time step is crucial to simulate the water volume exchange between the aquifer and river and match the groundwater levels in wells near the rivers. The initial groundwater level values used to initiate the calibration process are shown in Figures 45 and 46. The daily time step for observed river stage is accurately known at the six river gauging sites shown in Figure 47. In between these six sites at which river gauging measurements had been made, the stage at each river cell in the model was initially estimated using the Akima inter/extrapolation technique based on a cubic transfer function. However, interpolated river stage in some areas did not allow for accurate simulation of the exchange of water between the aquifer and river, particularly during irrigation periods. To improve the accuracy of interpolated river stage throughout the model domain, six “intermediate” river stage sites were defined within the model domain (Figures 48 and 49). The interpolated river stage values at the intermediate sites were lowered in the reaches where groundwater abstraction occurred (river bed levels were not adjusted), in order to produce lower modelled groundwater levels to more closely match observed

Confidential 2010


groundwater levels during the irrigation period. An alternate approach to achieve the same end would be to leave the river stage values unmodified but adjust the leakance coefficients for only those river reaches where groundwater abstraction occurred. However, this approach was discarded because there is no physical reason to expect the leakance coefficients to differ between river reaches where groundwater abstraction is or is not occurring. On the other hand, it is physically reasonable to expect that river stage levels will be affected by groundwater abstraction. Thus, model calibration was achieved by adjusting river stage for reaches with significant groundwater abstraction, while assuming that a single value of the leakance coefficient applied to all river reaches.

After the regeneration of interpolated river stage in the model, spatial distributions of hydraulic conductivity and the river transfer coefficient were adjusted to achieve a match between modelled and observed groundwater levels. Calibrated horizontal hydraulic conductivity distribution is shown in Figures 50 and 51. The hydraulic conductivity values from pump test data vary by less than one order of magnitude, from 103 m/day (12 × 10-4 m/s) to 432 m/day (40 × 10-4 m/s). The calibrated hydraulic conductivity values covered a similar range. The global value for the river bed leakage coefficient was manually adjusted until the loss and gain of various river sections, simulated for 9th February 2002 (Day 224), were as close as possible to the observations (see further discussion below). The final value of the leakance coefficient (Φ) used in the model is 60 day-1.

The final fit between observed and simulated groundwater levels at the five calibration wells is shown in Figures 52 to 56. Figures 52 to 56 indicate the good agreement between measured and simulated groundwater levels at the five observation wells. The observed groundwater level records indicate that groundwater levels decrease during the irrigation season in the period of November to April, and recover to average levels each winter (by August). The model prediction shows a good match to the observed drawdown at the five groundwater wells during the irrigation season as well as to winter recovery. Table 9 summarizes the statistical results for the model calibration, based on the root mean square error (RMSE) between observed and predicted groundwater levels.

The model predictions match observed groundwater levels most closely at Quinney’s Bush and Hyatt’s (Table 8). The difference between observed and predicted groundwater level at Crimps and Vue Mount was higher than other three sites because of the complexity of groundwater-river interaction with groundwater abstraction, particularly during the irrigation season. The maximum difference between observed and predicted groundwater levels at five calibration wells changes only slightly during the irrigation season compared to the winter season. This result shows that the calibrated model can simulate hydrological processes over the whole year, giving a reasonably accurate balance of prediction for both the summer and winter seasons.

A check on the robustness of the model can be made by evaluating the flux (gain and loss) of water from the river system, compared to field measurements (concurrent gaugings) made on the 9th February 2002 (Stewart et al., 2003). For this day and over the entire model domain (from Quinney’s Bush and North Bridge to the Wangapeka confluence), the model predicts that the river system gains 73,788 m3 water (Figure 57). In comparison, the gauging measurements for the same day indicate that the entire river system gains 28,602 m3 water. For some individual river reaches, the model prediction of flux into or out of the river differs from the field measurement by a factor of two or three, but the predicted fluxes are within the same order of magnitude as the field measurements.

Confidential 2010


Overall then, the model reproduces the general dynamics of river-groundwater interchange acceptably well. The degree of misfit between the model predictions and the field observations is considered acceptable for two reasons. Firstly, the model’s primary aim is to match groundwater level, which it achieves very well; the ability to match river-groundwater flux exchange is considered to be a secondary calibration target, because it is limited by the accuracy of the interpolation of river stage between measurement points. Secondly, the concurrent gauging measurements themselves may not be of sufficient accuracy to warrant any additional emphasis on matching them by the model. The measurements of river loss and gain are available for only one day, the 9th February 2002, a day on which substantial rainfall of about 40 mm was recorded (greater rainfall was observed on only six other days in the entire two year period for which the model is developed). The model simulations indicate that the volume of water lost and gained by the river is quite variable on a day-to-day time scale in response to rainfall (Figure 58). It is therefore quite possible that the river levels may have changed significantly in the time taken to complete the gauging survey, even within individual river reaches. This in turn suggests that the one-day record of gauged river gain and loss should be used only to guide to the model development, but not be used as a hard calibration target. A recommendation for future improvement of the model would be to repeat the concurrent river gauging surveys, to provide additional data to constrain the calibration.

6.0 SUMMARY

• The effects of water abstraction from aquifers and rivers is a growing issue for water management in the Upper Motueka River Catchment. The purpose of this study was to develop a groundwater-river interaction model that is capable of simulating the interaction between the shallow aquifer and the surface water system. The overall aim was to simulate the effects of dynamic groundwater abstraction on stream flow and groundwater levels, so that defensible water allocation policies can be put in place for the water resources in the Upper Motueka area.

• A three-dimensional finite-element groundwater model was developed to simulate the interaction between the shallow aquifer and the river system in the Upper Motueka River catchment. The model incorporates the Motueka River from 3 km downstream of the confluence with the Tadmor River to approximately 8 km upstream of the confluence with the Motupiko River. The model domain also incorporates 3 km of the Tadmor River upstream of the confluence with the Motueka River, and approximately 15 km of the Motupiko River upstream of the confluence with the Motueka River.

• The aquifer in the model is discretized using 6 modal triangular prism elements. The entire model domain is composed of 30386 finite elements and 29299 mesh nodes. The modelled thickness of the aquifer ranges from 6.5 m to 13.5 m.

• Three different rainfall recharge zones (P1, P2, and P3) are defined based on soil type. Each recharge zone is divided into irrigated and non-irrigated areas. Rainfall recharge in each zone is estimated by Landcare Research using a daily soil water balance model based on daily evapotranspiration estimates and water-holding capacities for the main soils in the Upper Motueka River catchment. For each of the three recharge zones, the rainfall recharge model is run on a daily time step for the period from 1st July 2001 to 30th June 2003.

• A daily-time step groundwater abstraction model is developed using a first-kind Dirichlet-type boundary condition. There are 39 groundwater abstraction wells identified in the

Confidential 2010


model domain. Daily groundwater abstraction at each well was calculated based on the land use pattern in the period from 1st July 2001 to 30th June 2003.

• The river system in the model is composed of three rivers: Motueka River, Motupiko River and Tadmor River. Thirty-eight river cross-sections were surveyed by TDC in these three rivers. The three rivers are implemented in the model using a head-dependent, third-kind Cauchy’s boundary condition. This boundary condition was used to dynamically simulate the leakage to and/or gain from rivers based on the head differences between the river and underlying groundwater.

• Four small streams that drain the surrounding hill country are defined as side river recharge sources that influence river flow. The four streams are implemented in the model using as head-dependent, third-kind Cauchy’s boundary conditions for: (A) Clarke River junction with Motupiko River; (B) Glen Rae stream junction with Motueka River; (C) Long Gully junction with Motupiko River; and (D) Norris Gully junction with Motueka River.

• Eleven zones of potential hill slope recharge to the aquifer were identified by Landcare Research. Hillslope recharge rates were calculated from a new model developed by Landcare Research and implemented in the groundwater model using a flux boundary condition with a daily time step for the period 1st July 2001 to 30th June 2003.

• The Upper Motueka groundwater-river interaction model is calibrated as a transient model using a daily time step. The observation data of groundwater levels, rainfall and the river flow show that the aquifer and river systems are highly dynamic. The model has been calibrated using visual comparison of the match between predicted and observed values at five monitoring wells (Quinney’s Bush, North Bridge, Crimps, Hyatts and Vue Mount). The period from 1st July 2001 to 30th June 2003 was used for calibration. The calibration process involved adjustment of interpolated values of river stage, spatial distribution of hydraulic conductivity and the river bed leakage coefficient. The calibrated model successfully represents variations and trends in groundwater levels across the model domain. The model also matches observed river-groundwater flux exchange (river gain or loss) to a reasonable degree.

• This phase of the project has demonstrated that a transient groundwater-river interaction model can be developed in a three-dimensional finite-element structure. The model can be used to investigate the effects of dynamic abstraction of groundwater and surface water on river flow and groundwater levels in the Upper Motueka Catchment. This model will be used to assess different water allocation management regimes in the Upper Motueka River Catchment in the final phase of the project.

Confidential 2010


7.0 REFERENCES

Chittenden, E.T., Hodgson, L., Dodson, K.J. 1966. Soils and agriculture of Waimea County, New Zealand. Soil Bureau Bulletin 30. DSIR, Wellington.

Diersch, H.J.G. 2005. FEFLOW finite element subsurface flow and transport simulation system reference manual, Version 5.4. WASY Institute for Water Resources Planning and System Research, Berlin, Germany.

Ekanayake, J, Davie, T, Payne, J. 2009. Estimating ground water recharge from rain and irrigation to the Upper Motueka groundwater catchment. Landcare Research ICM draft report.

Freeze, R.A., Cherry, J.A. 1979. Groundwater. Prentice-Hall, New Jersey.

Griffiths, E. 1975. Classification of land for irrigation in New Zealand. New Zealand Soil Bureau Scientific report 22, DSIR, Wellington.

Hong, T., Thomas, J., Davie, T. 2005. River-aquifer interaction modelling in the Upper Motueka River catchment: three-dimensional finite-element groundwater flow model, Motueka Integrated Catchment Management (Motueka ICM) Programme Report Series, Landcare ICM Report No. 2004-05/02.

Johnston, M.R. 1983. Geological map of New Zealand 1:50,000 Sheet N28 AC Motupiko. 1st Ed. New Zealand Geological Survey, Lower Hutt.

Stewart, M., Hong, T., Cameron, S., Daughney, C., Tait, T., Thomas, T. 2003. Investigation of groundwater in the Upper Motueka River Catchment, Institute of Geological and Nuclear Sciences Report 2003/32.

Confidential 2010


FIGURES

Confidential 2010


Figure 1. Location map showing the extent of the groundwater model of the Upper Motueka River catchment.

Confidential 2010


Figure 2. Upper Motueka River catchment showing locations of river monitoring stations.

Confidential 2010


125012

00

1150

1100

Figure 3. Annual rainfall isohyets for the Upper Motueka valley. Model area is shown in green (with

river channels displayed in blue). Location of Tapawera gauge also shown. Isohyet data from Tasman District Council.

Tapawera

Confidential 2010


Figure 4. Simplified geological map of the Upper Motueka River Catchment, showing locations of

groundwater monitoring sites.

Confidential 2010


Figure 5. Two-dimensional view of the Upper Motueka groundwater model domain, showing aquifer discretization using a finite-element mesh.

Confidential 2010


Figure 6. Aquifer discretization and finite-element mesh at the area of the confluence of the Tadmor River and Motueka River.

Confidential 2010


Figure 7. Aquifer discretization and finite-element mesh at the area of the confluence of the Motueka River and Motupiko River.

Confidential 2010


Figure 8. Aquifer discretization and finite element mesh in the upper valley of the Motueka River.

Confidential 2010


Figure 9. Aquifer discretization and finite element mesh in the Motupiko River valley.

Confidential 2010


Figure 10. Aquifer discretization and finite-element mesh in the Motupiko River valley.

Confidential 2010


Motupiko River

Motueka River

Motueka River

Tadmore

River

Figure 11. Three-dimensional view of the ground elevation structure from the bottom of the aquifer, showing aquifer discretization by finite-element mesh. Scale shows elevation (m asl).

Confidential 2010


Tam

or Rive

r

valle

y

Motueka River valley

Motueka River valley

Motupik

o Rive

r vall

ey

Figure 12. Distribution of hydraulic conductivity values before the model calibration process.

Confidential 2010


Figure 13. Estimation of actual evapotranspiration according to the soil water deficit. The only

hydraulic property used in the model is the RAW, which is taken as 100 mm for all three soil zones (P1, P2, P3).

Confidential 2010


Figure 14. Distribution of three different rainfall recharge zones and their irrigated areas.

Confidential 2010


0

10

20

30

40

50

60

70

80

90

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Rai

nfal

l (m

m/d

ay)

P1 rainfall

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Rec

harg

e (m

m/d

ay)

P1 irrigated

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Rec

harg

e (m

m/d

ay)

P1 non-irrigated

Figure 15. Daily time series of rainfall and calculated recharge in the P1 recharge zone.

Confidential 2010


0

10

20

30

40

50

60

70

80

90

2/07/2001 2/10/2001 2/01/2002 2/04/2002 2/07/2002 2/10/2002 2/01/2003 2/04/2003

Date

Rai

nfal

l (m

m/d

ay)

P2 rainfall

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/20021/04/2002 1/07/2002 1/10/2002 1/01/20031/04/2003

Date

Rec

harg

e (m

m/d

ay)

P2 irrigated

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/20021/04/2002 1/07/2002 1/10/2002 1/01/20031/04/2003

Date

Rec

harg

e (m

m/d

ay)

P2 non-irrigated


Confidential 2010


010

2030

405060

7080

90100

2/07/2001 2/10/2001 2/01/2002 2/04/2002 2/07/2002 2/10/2002 2/01/2003 2/04/2003

Date

Rai

nfal

l (m

m/d

ay)

P3 rainfall

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/20021/04/2002 1/07/2002 1/10/2002 1/01/20031/04/2003

Date

Rec

harg

e (m

m/d

ay)

P3 irrigated

0

10

20

30

40

50

60

70

1/07/2001 1/10/2001 1/01/20021/04/2002 1/07/2002 1/10/2002 1/01/20031/04/2003

Date

Rec

harg

e (m

m/d

ay)

P3 non-irrigated


Confidential 2010


Figure 18. Locations of groundwater abstraction wells in the Upper Motueka Catchment.

Confidential 2010


0

5000

10000

15000

20000

25000

30000

35000

40000

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Abs

tract

ion

(m3 /d

ay)

Total groundwater abstractionTotal surface water abstraction

Figure 19. Total estimated groundwater abstraction and direct surface water abstraction for the period from 1st July 2001 to 30th June 2003.

Confidential 2010


Figure 20. Location of river cross-section surveys and river monitroing sites in the Upper Motueka Catchment (Numbers represent the CS name shown in Table 5).

Confidential 2010


258.50

259.00

259.50

260.00

260.50

261.00

Offeset (m)

Alti

tude

(m a

sl)

234.50

235.00

235.50

236.00

236.50

237.00

237.50

238.00

238.50

239.00

Offset (m)

Alti

tude

(m a

sl)

176.00

176.50

177.00

177.50

178.00

178.50

179.00

179.50

180.00

180.50

Offset (m)

Altit

ude

(m a

sl)

148.00

148.50

149.00

149.50

150.00

150.50

151.00

151.50

152.00

0.00 4.92 7.82 8.50 11.48 12.96 16.12 19.28 24.13 28.95 36.31 44.21 50.59 55.38 56.18 57.42 60.16 57.32

Offset (m)

Altit

ude

(m a

sl)

151.00

151.50

152.00

152.50

153.00

153.50

154.00

154.50

155.00

155.50

Offset (m)

Alti

tude

(m a

sl)

109.00

110.00

111.00

112.00

113.00

114.00

115.00

Offset (m)

Altit

ude

(m a

sl)

Korere Goldpine

Kohatu Tapawera

Tadmor Wangapeka

0.00 5.00 5.34 6.06 6.55 8.16 11.79 14.96 17.60 18.39 18.78 21.30 24.37 26.09 28.70 24.260.00 1.80 2.82 3.72 5.18 7.52 8.88 10.30 12.14 14.10 15.74 17.98 22.21 25.22 27.16 29.00 31.05 28.86

6.38 10.98 19.19 25.21 30.14 35.00 40.43 44.69 51.72 53.32 54.83 60.40 55.860.00

0.00 1.44 7.61 12.47 13.38 15.79 15.57-1.42 0.00 2.88 5.88 5.91 7.12 8.76 10.08 14.28 19.76 26.16 33.00 38.04 47.57 52.20 56.08 56.43 57.54 65.29 66.02 69.49

Figure 21. Six selected river cross-sections in the Upper Motueka Catchment.

Confidential 2010


Figure 22. Cross-sections in the Motupiko River.

Confidential 2010


Figure 23. Cross-sections in the Motueka River.

Confidential 2010


0 5 10 15 20 250

100

200

300

400

500 River cross-section width

Riv

er c

ross

-sec

tion

wid

th (m

)

Distance from Wangapeka (m)

100

120

140

160

180

200

220

Minimum river bed

3915 7340 9591 12146 16805

CS name

Wanagpeka

Wanagpeka

18913Tapawera Confluence

Motueka RiverM

in river bed (m)

211

Figure 24. River width and elevation of lowest part of river bed in the Motueka River.

Riv

er c

ross

-sec

tion

wid

th (m

)

Min. river bed (m

)

30 32 34 36 380

20

40

60

80

100

120

140

River cross-section width

180

200

220

240

260

280

Minimum river bed

334 1370 2820 4148 6061

KorereConfluence

Confluence

CS name

Distance from Confluence (m)11854

Motupiko River

256

131

Figure 25. River width and elevation of lowest part of river bed in the Motupiko River.

Goldpine

Confidential 2010


(A) River recharge to groundwater

(B) Groundwater recharge to river

Figure 26. Depiction of the interaction between river and aquifer as modelled by the head-

dependent, third-kind Cauchy’s boundary condition. Source: Diersch (2005).

Confidential 2010


Figure 27. Locations of four side river recharge sources implemented in the model.

Confidential 2010


Rai

nfal

l (m

m/d

ay)

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/20030

10

2030

40

50

60

70

8090

Date

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Clarke River junction with Motupiko River

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Glen Rae stream junction with Motueka River

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Long Gully junction with Motupiko River

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/20030

50000

100000

150000

200000

250000

300000

350000

400000

450000

500000 Flow rate at Norris Gully junction with Motueka River

Side

rive

r flo

w ra

te (m

3 /day

)

Rai

nfal

l (m

m/d

ay)

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/20030

10

2030

40

50

60

70

8090

Date

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Clarke River junction with Motupiko River

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Glen Rae stream junction with Motueka River

0

50000

100000

150000

200000

250000

300000

350000

400000

450000 Flow rate at Long Gully junction with Motupiko River

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/20030

50000

100000

150000

200000

250000

300000

350000

400000

450000

500000 Flow rate at Norris Gully junction with Motueka River

Side

rive

r flo

w ra

te (m

3 /day

)

Figure 28. Daily time series of side river recharge at four sites implemented in the model.

Confidential 2010


Figure 29. Location of eleven hill slope recharge zones implemented in the model.

Confidential 2010


Figure 30. Locations of the A, B and K hill slope recharge zones implemented in the model.

Confidential 2010


Figure 31. Locations of the C, D, E and J hill slope recharge zones implemented in the model.

Confidential 2010


Figure 32. Location of the F, G, H and I hill slope recharge zones implemented in the model.

Confidential 2010


Figure 33. Korere hill slope recharge boundary and experimental site (Photo John Payne).

260

270

280

290

300

310

0 20 40 60 80 100 120 140 160 180

H-Distance (m)

Ele

vatio

n (m

)

G.level

Cap_Probe ECap_Probe D

TDRs C

TDRs B

TDRs C

Figure 34. Locations of capacitance probes and TDRs along the slope to measure the phreatic level at Korere site. Solid line represents ground level.

Confidential 2010


266

268

270

272

274

30 40 50 60 70 80 90 100

H distance (m)

Phr

eatic

leve

l (m

)

G.levelWinter maxWinter minsummer maxsummer min

Figure 35. Maximum and minimum phreatic levels at Korere site during 2006-2007 (levels are reference to an arbitrary datum).

Figure 36. Maximum and minimum phreatic levels at Paratiho site during 2006-2007 (levels are reference to an arbitrary datum).

Confidential 2010


Flow lines

Equipotential lines

Phreatic surface

Ground surfaceWater level recorders

HT(m)

D(m)

Δx

Δz

D = aquifer hillslope interface thickness

Water level recorders

Flow lines

Equipotential lines

Phreatic surface

Ground surfaceWater level recorders

HT(m)

D(m)

Δx

Δz

D = aquifer hillslope interface thickness

Water level recorders

Figure 37. Approximation of the water flow at the toe of a hillslope by two dimensional flow net analysis.

Phreatic Level & Drainage (Summer 03)

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50

Draiange (mm)

Phr

eatic

Lev

el (m

)

Measured Estimated

(a)

r2 =0.96

Phreatic Level & Drainage (winter 04)

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60 70 80 90 100

Draiange (mm)

Phr

eatic

Lev

el (m

)

Measured Estimated

(b)

r2 =0.88

Figure 38. Measured and estimated hillslope phreatic levels for summer and winter seasons at the

Paratiho site. Drainage remains relatively unchanged for large rain events due to overland flow generation.

Confidential 2010


Recession curves for winter and summer rain events

0

0.2

0.4

0.6

0.8

1

1 2 3 4 5 6 7 8 9

Days

Wat

er le

vel (

m)

summerwinter

Figure 39. Measured phreatic level recession curves for summer and winter at the Paratiho and Korere hill slope sites.

Scaling factor

0

1

2

3

4

5

6

0 500 1000 1500

(Catchement area/boundary length) m

Sca

ling

fact

or

Paratiho

korere

Scaling factor

0

1

2

3

4

5

6

0 500 1000 1500

(Catchement area/boundary length) m

Sca

ling

fact

or

Paratiho

korere

Figure 40. Scaling factor (S) represents the ratio of catchment area to hill slope boundary length.

Confidential 2010


Hillsope recharge ( Karore summer-06 )

0.010

0.011

0.012

0.013

0.014

0.015

23/1

0/20

06

6/11

/200

6

20/1

1/20

06

4/12

/200

6

18/1

2/20

06

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120Rain

Measured

Estimated

Hillsope recharge ( Karore w inter-06 )

0.010

0.011

0.012

0.013

0.014

0.015

1/05

/200

6

15/0

5/20

06

29/0

5/20

06

12/0

6/20

06

26/0

6/20

06

10/0

7/20

06

24/0

7/20

06

7/08

/200

6

21/0

8/20

06

TimeR

echa

rge

(m2 /d

ay)

0

20

40

60

80

100

120


0.012

0.012

0.013

0.013

0.014

0.014

0.015

0.015

22/0

5/20

07

5/06

/200

7

19/0

6/20

07

3/07

/200

7

17/0

7/20

07

31/0

7/20

07

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120Rain

Measured

Estimated

Hillsope recharge ( Karore summer-06 )

0.010

0.011

0.012

0.013

0.014

0.015

23/1

0/20

06

6/11

/200

6

20/1

1/20

06

4/12

/200

6

18/1

2/20

06

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120Rain

Measured

Estimated


0.010

0.011

0.012

0.013

0.014

0.015

1/05

/200

6

15/0

5/20

06

29/0

5/20

06

12/0

6/20

06

26/0

6/20

06

10/0

7/20

06

24/0

7/20

06

7/08

/200

6

21/0

8/20

06

TimeR

echa

rge

(m2 /d

ay)

0

20

40

60

80

100

120


0.012

0.012

0.013

0.013

0.014

0.014

0.015

0.015

22/0

5/20

07

5/06

/200

7

19/0

6/20

07

3/07

/200

7

17/0

7/20

07

31/0

7/20

07

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120Rain

Measured

Estimated

Figure 41. Ground water recharge at Korere hill slope site. Blue line indicates the calculated water

recharge using the measured phreatic levels. Green lines show calculated recharge from measured phreatic levels using the flow net analysis.

Confidential 2010


Hillsope recharge ( Paratiho w inter-06 )

0.00

0.01

0.02

0.03

11/0

5/20

06

8/06

/200

6

6/07

/200

6

3/08

/200

6

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated

Hillsope recharge ( Paratiho summer-03/04 )

0.00

0.01

0.02

0.03

0.04

0.05

1/09

/200

3

29/0

9/20

03

27/1

0/20

03

24/1

1/20

03

22/1

2/20

03

19/0

1/20

04

16/0

2/20

04

15/0

3/20

04

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in(m

m)

Rain

Measured

Estimated


0.00

0.01

0.02

0.03

18/0

9/20

04

16/1

0/20

04

13/1

1/20

04

11/1

2/20

04

8/01

/200

5

5/02

/200

5

5/03

/200

5

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated


0.00

0.01

0.02

1/09

/200

6

29/0

9/20

06

27/1

0/20

06

24/1

1/20

06

22/1

2/20

06

19/0

1/20

07

16/0

2/20

07

16/0

3/20

07

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated

Hillsope recharge ( Paratiho w inter-06 )

0.00

0.01

0.02

0.03

11/0

5/20

06

8/06

/200

6

6/07

/200

6

3/08

/200

6

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated


0.00

0.01

0.02

0.03

0.04

0.05

1/09

/200

3

29/0

9/20

03

27/1

0/20

03

24/1

1/20

03

22/1

2/20

03

19/0

1/20

04

16/0

2/20

04

15/0

3/20

04

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in(m

m)

Rain

Measured

Estimated


0.00

0.01

0.02

0.03

18/0

9/20

04

16/1

0/20

04

13/1

1/20

04

11/1

2/20

04

8/01

/200

5

5/02

/200

5

5/03

/200

5

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated


0.00

0.01

0.02

1/09

/200

6

29/0

9/20

06

27/1

0/20

06

24/1

1/20

06

22/1

2/20

06

19/0

1/20

07

16/0

2/20

07

16/0

3/20

07

Time

Rec

harg

e (m

2 /day

)

0

20

40

60

80

100

120

Dai

ly ra

in (m

m)

Rain

Measured

Estimated

Figure 42. Ground water recharge at Paratiho experimental site. Blue lines indicate the estimated

recharge using the measured phreatic levels. Green lines show calculated recharge from measured phreatic levels using the flow net analysis.

Confidential 2010


0 100 200 300 400 500 600 700

0

1

2

3

4

5

Hills

lope

sre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

4

8

12

16

Hill

slop

esre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

2

4

6

8

Hills

lope

sre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

2

4

6

8

Hills

lope

sre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

1

2

3

4

5

Hill

slop

esre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

2

4

6

8

10

Hill

slop

esre

char

ge(m

3 )

Site A Site B

Site D

Site FSite E

Site C

Day Day

Figure 43. Daily time series of hill slope recharge at sites A to F as implemented in the model.

Confidential 2010


0 100 200 300 400 500 600 700

0

1

2

3

4

5

Hill

slop

esre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

1

2

3

4

5

Hills

lope

sre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

4

8

12

Hill

slop

esre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

2

4

6

8

10

Hill

slop

esre

char

ge(m

3 )

0 100 200 300 400 500 600 700

0

2

4

6

8

10

Hills

lope

sre

char

ge(m

3 )

Site G Site H

Site I Site J

Site K

Day Day

Day Day

Day

Figure 44. Daily time series of hill slope recharge at sites G to K as implemented in the model.

Confidential 2010


Figure 45. Contour map of initial groundwater level used to initiate calibration of the Upper Motueka groundwater-river interaction model.

Confidential 2010


Figure 46. Contour map of initial groundwater hydraulic head used to initiate calibration of the Upper Motueka groundwater-river interaction model.

Confidential 2010


Rai

nfal

l (m

m/d

ay)

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/20030

10

20

30

40

50

60

70

80

90

Riv

er s

tage

(m)

Date

River stage at Korere River stage at Goldpine

235.0

235.5

236.0

236.5

259.0

259.5

260.0

River stage at Kohatu River stage at Tapawera

149.5

150.0

150.5

178.0

178.5

179.0

179.5

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003112.0

112.5

113.0

113.5 River stage at Wangaeka

Riv

er s

tage

(m)

Riv

er s

tage

(m)

Figure 47. Observed mean daily river stage at five river monitoring sites.

Confidential 2010


Intermediate river siteRiver monitoring siteGroundwater monitoring site

Figure 48. Intermediate river sites defined in the lower part of the model to improve simulation of

river stage.

Confidential 2010


Intermediate river siteRiver monitoring siteGroundwater monitoring site

Figure 49. Intermediate river sites defined in the upper part of the model to improve simulation of

river stage.

Confidential 2010


Vue Mount

Campbells

Hyatts

Figure 50. Calibrated hydraulic conductivity distribution in the lower part of Upper Motueka groundwater-river interaction model. Insert map shows full scale of the model domain.

Confidential 2010


Crimps

North Bridge

Quinney's Bush

Figure 51. Calibrated hydraulic conductivity distribution in the upper part of Upper Motueka

groundwater-river interaction model. Insert map shows full scale of the model domain.

Confidential 2010


191.0

191.5

192.0

192.5

193.0

193.5

194.0

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Gro

undw

ater

leve

l (m

asl

)

0

20

40

60

80

100

120

140

Rai

nfal

l rec

harg

e (m

m/d

ay)Rainfall recharge P1 non-irrigated

Quinney's Bush referenceQuinney's Bush simulation

Figure 52. Model calibration results for groundwater level at Quinney’s Bush.

202.3

202.8

203.3

203.8

204.3

204.8

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Gro

undw

ater

leve

l (m

)

0

20

40

60

80

100

120

140

Rai

nfal

l rec

harg

e (m

m/d

ay)

Rainfall recharge P3 non-irrigatedNorth Bridge observationsNorth Bridge model predictions

Figure 53. Model calibration results for groundwater level at North Bridge.

Confidential 2010


178.3

178.8

179.3

179.8

180.3

180.8

181.3

181.8

182.3

182.8

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Gro

undw

ater

leve

l (m

)

0

20

40

60

80

100

120

140

Rai

nfal

l rec

harg

e (m

m/d

ay)

Rainfall recharge P1 irrigatedCrimps referenceCrimps simulation

Figure 54. Model calibration results for groundwater level at Crimps.

154.5

155.0

155.5

156.0

156.5

157.0

157.5

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Gro

undw

ater

leve

l (m

)

0

20

40

60

80

100

120

140

Rai

nfal

l rec

harg

e (m

m/d

ay)

Rainfall recharge P1 irrigatedHyatts observationsHyatts model predictions

Figure 55. Model calibration results for groundwater level at Hyatts.

Confidential 2010


136.0

136.5

137.0

137.5

138.0

138.5

139.0

139.5

1/07/2001 1/10/2001 1/01/2002 1/04/2002 1/07/2002 1/10/2002 1/01/2003 1/04/2003

Date

Gro

undw

ater

leve

l (m

)

0

20

40

60

80

100

120

140

Rai

nfal

l rec

harg

e (m

m/d

ay)

Rainfall recharge P1 non-irrigatedVue Mount observationsVue Mount model predictions

Figure 56. Model calibration results for groundwater level at Vue Mount.

Confidential 2010


Figure 57. Observed and modelled river flow loss and gain (m3/day). The observations are based on

the concurrent gauging survey conducted on 9th February 2002. The modelled river loss and gain are derived from the calibrated FEFLOW model for the time step (day) corresponding to 9th February 2002.

Confidential 2010


-400,000

-200,000

0

200,000

400,000

600,000

1/07

/200

1

1/09

/200

1

1/11

/200

1

1/01

/200

2

1/03

/200

2

1/05

/200

2

1/07

/200

2

1/09

/200

2

1/11

/200

2

1/01

/200

3

1/03

/200

3

1/05

/200

3

Date

Mod

elle

d R

iver

Gai

n or

Los

s, W

hole

Mod

el D

omai

n (m

3 /d)

Figure 58. Modelled daily river gain and loss over the entire study area. Positive and negative

numbers indicate predicted volume of water gained or lost, respectively, by the whole river system (between Quinney’s Bush and North Bridge and the Wangapeka confluence).

Confidential 2010


TABLES

Confidential 2010


Table 1. Summary of hydrological monitoring sites in the model domains.

Type Site Name Grid Ref X Y GWL (m)

GW Higgins N28:9638-6996 2496379 5969958 201.51

GW Quinney's Bush N28:9465-7217 2494645 5972174 191.30

GW Crimps N28:9574-7343 2495743 5973426 179.11

GW Hyatts N28:9593-7750 2495927 5977500 154.97

GW Campbells N28:9643-7777 2496432 5977765 155.58

GW Vue Mount (Biggs) N27:9409-8087 2494085 5980869 136.62

GW Oldham N28:9350-7895 2493500 5978950 150.35

GW North Bridge 2496300 5970000 201.67

Discharge (L/s)

River Above Wangapeka confluence ? 2492737 5985290 1855

River Glenrae ? 2492777 5984045 1427

River Smiths ? 2493328 5982558 2419

River Below Tapawera ? 2494040 5981439 1692

River Tapawera bridge ? 2494499 5980096 1880

River Rogers ? 2495102 5978701 1778

River Hyatts ? 2495515 5977426 1584

River Reynolds ? 2496422 5976055 1607

River Downstream Kohatu ? 2496152 5973305 1781

River Kohatu bridge ? 2496187 5972915 1539

Rain Woodstock ? 2495100 5994300

Rain Tapawera Bridge M27:945799

Rain Motueka Gorge N28:028526 2502800 5952600

Confidential 2010


Table 2. Hydraulic characteristics of the aquifer in the Upper Motueka Catchment.

ID Well Number Bore ID X Y GWL RL

(m) Surveyed Ground

Level (m)

Well Depth (m)Hydraulic

Conductivity (m/d)

RL top of Moutere Gravel

(m) 1 4614 Higgins 2496379 5969958 202.51 205.21 7.2 940 197.113

2 4615 Quinney's Bush 2494645 5972174 191.30 193.68 7.5 620 185.682

3 4616 Crimp 2495743 5973426 179.11 182.18 11 465 170.68

4 4617 Hyatt 2495927 5977500 154.97 158.29 12.6 213 145.285

5 4618 Campbell 2496432 5977765 155.58 160.01 11.3 753 148.213

6 4619 Vue Mount 2494085 5980869 136.62 139.03 6.5 620 130.03

7 4620 Oldham 2493500 5978950 150.35 153.32 4.5 54 148.824

Confidential 2010


Table 3. Statistical summary of the rainfall recharge models developed for the period from 1st July 2001 to 30th June 2003.

Recharge (mm/day)

P1 P2 P3

Irrigated 718.3 854.8 914.3

Non-irrigated 692.3 821.3 902.6

Annual recharge rates (% of rainfall)

P1 P2 P3

Irrigated 32.5 35.5 36.3

Non-irrigated 31.3 34 35.8

Total rainfall for 2001-2003 year

P1 P2 P3

Rainfall total 2211.3 2410.3 2520.9

Area by soil type (ha)

P1 P2 P3

Atapo 153 31 3

Dovedale 4 2 0

Kikiwa 26 0 0

Korere 0 2 0

Motupiko 366 257 420

Sherry 0 0 1

Spooner 0 1 6

Tapawera 639 195 299

Total 1189 488 729

Confidential 2010


Table 4. Groundwater abstraction wells in the model domain.

POW ID Eastings Northings Max Daily Take

(m3/day) Irrigation Area

(ha) Landuse Well No. Consent Applicant Names Address

1 2494930 5972736 172 4 Pasture 4550 40504 B J & E M Duncan Bevan & Elaine Motupiko

2 2494288 5971216 1586 37 40548 K J Fry Kevin Motupiko

3 2494103 5971805 373 8.7 4810 40596 A F Morton Andrew Motupiko

4 2494723 5972643 129 3 4611 40631 Motupiko Dairy Farm Ltd Julian Raine C/- J R Raine

5 2493819 5970981 1714 40 Pasture 4551 40632 Motupiko Dairy Farm Ltd Julian Raine C/- J R Raine

6 2495977 5975847 430 15.2 Pasture

7 2495870 5975530 257 6 Pasture 4772 40277 Hubbard Partnership Alan & Olive 4 Glover Place

8 2493241 5982076 429 10 Potato and Pasture 4584 40392 S P & D K Phillips Peter & Joyce Glenrae Road

9 2493528 5981846 429 10 - See map for diff. areas 4583 40393 S P & D K Phillips Peter & Joyce Glenrae Road

10 2495927 5978111 214 5 Blackberries 4559 40416 C R Fry Colin C/- Ross Fry

11 2496412 5970755 360 10 Pasture 4648 40430 M J & H A Gillard Mark & Heather Kohatu

12 2493338 5984308 4671 109 Hops & Blackcurrants 4564 40446 Hinetai Hops Ltd Wai west horticulture P O Box 2244

13 2494458 5979645 322 9 Raspberries 4641 40518 P B Higgins & Sons Ltd Peter P O Box 1

14 2496167 5978388 200 9.8 Barley, Red clover

15 2495509 5979291 407 9.5 4555 40561 D A & J K Cooper David & Julie 4741 Motueka Valley Highway

16 2495683 5978858 407 9.5 Red clover, Barley, Lamb feed 4554 40562 D A & J K Cooper David & Julie 4741 Motueka Valley Highway

17 2496351 5977890 424 9.9 Red clover (grown in different paddocks each year) and Pasture 4558 40566 J W Rodgers John Tapawera

18 2494750 5980580 300 7 4769 40570 K R Haakma Kenneth C/- Tapawera Postal Agency

19 2496219 5972330 250 5 4598 40576 JN McGaveston Julian & Carroll State Highway 6

20 2492665 5983721 2057 48 Pasture 4563 40583 F J & E A O`Connor Fergus & Elizabeth C/- Martin O`Connor

21 2493063 5982691 1307 30.5 Pasture 4476 40584 Fj & Ea & Mp O`Connor & F & T Leov C/- Martin O`Connor

22 2495682 5974438 514 12 Pasture 4581 40610 I D Diack Donald Old School Road

23 2493179 5985135 643 15 Pasture and Clover for seed production and foda crops 4568 40613 David E & Daphne J McGaveston & Hamish Kennedy David, Daphne & Kennedy Main Road Tapawera

24 2496525 5976679 1286 30 Blackcurrants 4548 40616 Toka Farms Ltd Robbie C/- Robbie Reynolds

25 2496821 5976898 429 10 Pasture 4547 40618 Toka Farms Ltd Robbie C/- Robbie Reynolds

26 2493142 5982438 1050 24.5 Pasture 4630 40619 M & J Bonny Martin & Jillian 47 Leicester Street

27 2496619 5976027 643 11.6 Pasture

28 2495644 5972602 1286 30 Pasture (Dairy) 4570 40623 Paul Dewhirst Paul & Tania Boscobel

29 2495866 5971244 2572 60 Pasture (Dairy) 4569 40624 Paul Dewhirst Paul & Tania Boscobel

30 2495768 5973517 1286 30 Pasture 4792 40650 Wairua Farm (2001) Ltd C/- E N & M A Baigent




34 2492940 5982936 150 3.5 Pasture 4567 40715 O B & J E Oxnam and M N Oxnam Owen, John, Murray 345 Tapawera-Baton Road

35 2496291 5977170 1414 33 Blackcurrants 4556 40730 Hyatt & Sons Ltd John Hyatt Sunrise Valley

36 2496519 5977477 772 18 Pasture and lamb fattening crops 4774 40732 Hyatt & Sons Ltd Sunrise Valley

37 2496016 5977533 827 19.3 Blackcurrants 4802 40733 Hyatt & Sons Ltd Sunrise Valley

38 2495362 5978677 424 Red clover, Lucern - This is grown in different paddocks each year. 4541 40751 R J & S A Rodgers Richard & Sharon Tapawera

39 2497448 5968847 47 1.1 Golf green 4650 40849 Golden Downs Golf Club Inc Karl Foulsham C/- Grant R Faulknor

Confidential 2010


Table 5. Summary of river cross-section surveys in the Motueka, Motupiko, and Tadmor Rivers.

CS name Y X Distance

along river (m)

Stage (m asl)

Elevation of lowest part of

river bed (m asl)

River cross section width (m)

Wangapeka 2492625.840 5985979.090 0 114.350 111.596 70.884 CS1 2492815.644 5985631.281 440.16 115.877 113.468 144.169 CS2 2492609.813 5984836.304 1317.58 118.800 116.701 149.397 CS3 2492768.246 5984054.457 2129.58 122.622 119.450 196.457 CS4 2493218.511 5983374.977 2987.41 126.420 124.164 173.883 CS5 2493498.796 5982508.322 3915.51 131.518 127.87 152.845 CS6 2493640.477 5982101.035 4356.92 132.814 130.389 468.656 CS7 2493985.998 5981401.078 5163.74 137.890 133.431 419.990 CS8 2494213.933 5980691.576 5905.38 143.737 136.788 140.870 CS9 2494573.531 5979879.457 6793.30 144.574 142.613 194.737 CS10 2494658.451 5979329.254 7340.56 150.590 144.065 150.308 CS11 2494915.376 5978831.364 7939.67 151.080 147.467 175.432 CS12 2495167.739 5978581.538 8257.98 151.313 149.765 116.073

Tapawera 2495180.530 5978544.39 8294.68 151.299 149.287 57.313 CS13 2495216.566 5978474.018 8373.16 150.050 150.047 140.683 CS14 2495426.596 5977968.328 8943.18 155.068 152.739 184.355 CS15 2495638.420 5977343.198 9591.91 159.790 154.958 270.847 CS16 2496076.218 5976641.406 10581.47 163.672 159.395 136.199 CS17 2496185.132 5976264.155 10938.12 164.761 161.994 155.474 CS18 2496454.829 5975967.068 11332.85 167.303 164.193 107.403 CS19 2496373.527 5975577.052 11754.34 168.710 166.741 104.796 CS20 2496341.186 5975167.294 12146.19 170.877 168.426 212.383 CS21 2496253.229 5974276.924 13120.39 177.165 173.746 143.598 CS22 2496154.493 5973637.354 13809.28 181.164 177.945 147.819

Kohatu 2496104.000 5973688.000 178.600 177.59 55.86 Confluence 2496147.784 5973071.040 14296.53 182.867 180.672 138.243

CS23 2495965.191 5972321.236 15176.21 188.735 186.673 121.937 CS24 2496193.033 5971588.394 15991.13 193.917 191.900 103.393 CS25 2496096.662 5970788.482 16805.55 199.396 196.915 78.235 CS26 2496567.170 5970287.241 17493.74 204.209 201.559 90.089 CS27 2496711.790 5969577.496 18266.40 208.985 206.826 263.602 CS28 2496666.696 5968954.603 18913.57 216.725 211.280 226.416 CS29 180.475 55.532 CS30 2495881.999 5972906.850 334.33 184.409 182.023 55.532 CS31 2495434.829 5972870.624 779.48 186.637 185.266 82.737 CS32 2495098.705 5972409.154 1369.87 192.488 190.020 109.393 CS33 2494631.565 5971829.014 2132.61 196.320 194.215 99.689 CS34 2494395.485 5971192.496 2820.39 200.968 199.711 131.365 CS35 2494213.714 5970658.124 3419.16 205.079 204.423 98.958 CS36 2493903.171 5970071.824 4148.89 209.816 207.734 56.573 CS37 2493655.803 5969949.477 4453.79 211.388 209.329 76.097 CS38 2493482.292 5968478.370 6061.98 221.520 220.111 74.874 Korere 2493255.000 5963182.000 11854.91 260.810 258.870 24.26 Tadmor 2493266.680 5978825.220 3762.57 154.2110 152.462 15.569 Goldpine 2498470.82 5966289.340 22483.43 238.5520 234.993 28.86151936

Confidential 2010


Table 6. Flow information for the period of 1st July 2001 to 30th June 2003 for four side river recharge sources implemented in the model.

Site X Y Average flow (m3/day)

Minimum flow (m3/day)

Maximum flow (m3/day)

Average stage (m)

A 2492410 5961020 31532 1272 363962 269 B 2492555 5984280 27496 2261 384928 127.6 C 2494035 5970510 22696 2110 306610 207.4 D 2496260 5971315 37570 3740 486557 195.1

Table 7. Initial groundwater levels used to initiate calibration of the Upper Motueka groundwater- river interaction model.

ID Groundwater monitoring site X Y Initial head (m)

1 Vue Mount 2494085 5980869 137 2 North Bridge 2496300 5970000 203 3 Quinney’s Bush 2494645 5972174 192 4 Crimps 2495743 5973426 180 5 Hyatts 2495927 5977500 155 6 Campbells 2496432 5977765 156 7 Oldham 2493500 5978950 151

Confidential 2010


Table 8. Summary of calibration results for five groundwater monitoring wells.

Groundwater monitoring site

Calibration RMSE*

(m)

Evaluation RMSE*

(m)

Max. difference between

observed and predicted

(m)

Average difference between

observed and predicted

(m)

Max. difference over

irrigation period 1st Nov-31 Mar.

(m)

Max. difference over

winter period 1st July-30 Sept.

(m)

Vue Mount 0.17 0.15 0.41 -0.11 0.22 0.23

Crimps 0.29 0.2 0.49 -0.11 0.3 0.4

Hyatts 0.12 0.1 0.74 0.01 0.42 0.6

Quinney’s Bush 0.09 0.09 0.45 0.002 0.29 0.45

North Bridge 0.16 0.15 0.67 0.03 0.44 0.48

* Root mean square error

1 Fairway Drive

Avalon

PO Box 30368

Lower Hutt

New Zealand

T +64-4-570 1444

F +64-4-570 4600

Dunedin Research Centre

764 Cumberland Street

Private Bag 1930

Dunedin

New Zealand

T +64-3-477 4050

F +64-3-477 5232

Wairakei Research Centre

114 Karetoto Road

Wairakei

Private Bag 2000, Taupo

New Zealand

T +64-7-374 8211

F +64-7-374 8199

National Isotope Centre

30 Gracefield Road

PO Box 31312

Lower Hutt

New Zealand

T +64-4-570 1444

F +64-4-570 4657

Principal Location

www.gns.cri.nz

Other Locations

cr 2010-211 upper motueka catchment finalicm.landcareresearch.co.nz/knowledgebase... · gns science...

Documents