Waitaki Water Quality Catchment Modelling

Prepared for Environment Canterbury

May 2015

Prepared by : Christopher Palliser Sandy Elliott Sharleen Yalden Ude Shankar

For any information regarding this report please contact: Christopher Palliser Scientist Catchment Processes +64-7-856 1748 [email protected]

National Institute of Water & Atmospheric Research Ltd PO Box 11115 Hamilton 3251

Phone +64 7 856 7026

NIWA CLIENT REPORT No: HAM2015-002 Report date: May 2015 NIWA Project: ENC14202

Quality Assurance Statement Dr Annette Semadeni-Davies Reviewed by: Urban aquatic scientist NIWA Auckland

Formatting checked by: A. Bartley

Approved for release by: D. Roper

© All rights reserved. This publication may not be reproduced or copied in any form without the permission of the copyright owner(s). Such permission is only to be given in accordance with the terms of the client’s contract with NIWA. This copyright extends to all forms of copying and any storage of material in any kind of information retrieval system.

Whilst NIWA has used all reasonable endeavours to ensure that the information contained in this document is accurate, NIWA does not give any express or implied warranty as to the completeness of the information contained herein, or that it will be suitable for any purpose(s) other than those specifically contemplated during the Project or agreed by NIWA and the Client.

Contents

Executive summary ...... 5

1 Introduction ...... 6

2 Method ...... 9 2.1 The CLUES model ...... 9 2.2 Measured water quality and flow data ...... 25 2.3 Determination of suitable calibration sites ...... 36 2.4 Calibration ...... 39

3 Results ...... 43 3.1 Model fit and uncertainty ...... 43

4 Summary and conclusions ...... 52

5 References ...... 53

Appendix A Waitaki sub-catchments ...... 55

Appendix B Percentage of land use by area for each of the watersheds of the calibrated sites ...... 58

Appendix C TN and TP leaching rates ...... 64

Appendix D Flow diversions ...... 87

Appendix E Detrended concentrations ...... 91

Tables Table 2-1: Map key and annual nutrient loads for point sources. 12 Table 2-2: Combinations of land use, rainfall and soil types in the Upper and Lower catchments used in the customised version of CLUES (ECan, pers. comm .). 15 Table 2-3: Soils in the Upper Waitaki (ECan, pers. comm.). 17 Table 2-4: Soils in the Lower Waitaki (ECan, pers. comm .). 18 Table 2-5: Measured, default CLUES and customised CLUES flows for the water quality calibration sites in the Upper Waitaki. 20 Table 2-6: Measured, default CLUES and customised CLUES flows for the water quality calibration sites in the Lower Waitaki. 23 Table 2-7: Estimated flow diversions in the Waitaki River catchment. 25 Table 2-8: Measured median concentrations for TN and TP in the Upper Waitaki catchment. 28

Waitaki Water Quality Catchment Modelling

Table 2-9: Measured median concentrations for TN and TP in the Lower Waitaki catchment. 32 Table 2-10: Map key for calibration sites. 37 Table 2-11: Adjusted TN and TP leaching rates for the Upper Waitaki. 40 Table 2-12: Adjusted TN and TP leaching rates for the Lower Waitaki. 42 Table 3-1: TN measured and modelled loads and yields at the calibration sites in the Upper Waitaki catchment. 44 Table 3-2: TP measured and modelled loads and yields at the calibration sites in the Upper Waitaki catchment. 45 Table 3-3: TN measured and modelled loads and yields at the calibration sites in the Lower Waitaki catchment. 47 Table 3-4: TP measured loads and yields at the calibration sites in the Lower Waitaki catchment (see Figure 2-4). 49

Table A-1: Figure 1-1 key for sub-catchments. 55 Table B-1: Percentage of land use by area for each of the watersheds of the calibrated water quality monitoring sites. 58 Table C-1: TN and TP leaching rates for the Upper Waitaki (ECan, pers. comm.). 64 Table C-2: TN leaching rates for the Lower Waitaki ( ECan, pers. comm .). 71

Figures Figure 1-1: Waitaki River sub-catchments showing streams of order ≥ 3 and lakes. 8 Figure 2-1: Schematic of SPARROW sources and transport. 10 Figure 2-2: Waitaki River catchment showing streams of order ≥ 3, lakes and point sources. 11 Figure 2-3: The land use in the customised CLUES (ECan, pers. comm .). 14 Figure 2-4: Waitaki River sub-catchments showing streams of order ≥ 3, lakes and calibration sites. 38 Figure 3-1: Measured vs modelled TN and TP loads for the calibration sites in the Upper Waitaki. 50 Figure 3-2: Measured vs modelled TN loads for the calibration sites in the Lower Waitaki. 51

Waitaki Water Quality Catchment Modelling

Executive summary

Environment Canterbury requires a whole-of-catchment model for the Waitaki River catchment to provide a catchment-wide view over the likely impacts of land use changes on nutrient (total nitrogen and total phosphorus) loads. While the CLUES (Catchment Land Use for Environmental Sustainability) model was identified as being suitable for this purpose, the model required customisation to increase the spatial representation of current land use and soil drainage characteristics because of the diverse nature of the land uses, soils and rainfall. The development of the customised version of CLUES and its calibration against measured nutrient loads are described in this report. Additionally, the methods used to calculate loads for calibration are described, the sites suitable for calibration are identified, and the modelled or predicted loads are compared with the measured ones.

The model was applied and calibrated separately for the upper and lower sections of the Waitaki Catchment. The calibration results for the model varied with the Root Mean Square Error (RMSE) being 3 m 3/s for flow over the entire catchment, 34.35 t/y for total nitrogen in the Upper Waitaki catchment, 10.38 t/y for total nitrogen in the Lower Waitaki catchment, and 2.00 t/y for total phosphorus in the Upper Waitaki catchment. Environment Canterbury instructed that total phosphorus calibration was not to be done in the Lower Waitaki.

Uncertainty in the modelled loads arises from a number of sources, most notably limited data availability for nitrogen and phosphorus concentrations, both in terms of the number of sites for which water quality is also recorded and the percentage of water quality readings from these sites that were censored. With respect to the latter, techniques were used to enable the use of these concentration data. There is also uncertainty around the flows connected to hydro schemes within the catchment that were used to calculate loads. It was therefore considered wisest to focus on modelled percentage changes when comparing the results from various scenarios.

The CLUES model does not simulate groundwater and assumes that groundwater lags are zero (i.e., that stream concentrations reflect current land use). This means that the model calibration discussed here adjusted key coefficients (e.g., TN and TP yields and stream attenuation coefficients) to match current nutrient load observations. If there are significant groundwater lags in the region, then the CLUES results are likely to under-predict stream loads – that is, the effects of the recent growth in dairying will not yet be fully shown in the measured loads.

Waitaki Water Quality Catchment Modelling 5

1 Introduction Environment Canterbury (ECan) requires a whole-of-catchment model for the Waitaki River catchment to provide a catchment-wide view of the likely impacts of land use changes on nutrient (total nitrogen or TN, and total phosphorus or TP) annual loads in the river drainage network. The Catchment Land Use for Environmental Sustainability model (CLUES; Elliott et al. 2008, Semadeni- Davies et al. 2011) was identified as a suitable model. This model has been set up on a regional basis and has been calibrated nationally. However, to update and improve the spatial representation of the catchment, the default CLUES model was customised and recalibrated specifically for this application. The model was customised (called customised CLUES in this report) in order to provide ECan with a model that is better able to provide robust, defensible estimates of nutrient loads under current and future scenarios at a limited number of nodes in the Waitaki River catchment.

For the purposes of this report, the Waitaki River catchment was divided into Upper and Lower Waitaki catchments, and each was calibrated separately. For the latter, the measured nutrient loads at the bottom of the Upper Waitaki were entered as pseudo point sources into the top of the Lower Waitaki. Both TN and TP loads were calibrated for the Upper Waitaki catchment, whereas just TN load was calibrated for the Lower Waitaki catchment1.

Customisations included:

° Reconfiguration of the drainage network in the upper catchment to take account of the power scheme diversions. This involves prescribing the fraction of load that is diverted via the hydro canals.

° Reconfiguration of the drainage network to re-direct flows in the Kelland Pond/Wairepo Creek area to account for the groundwater divide not following surface topography in that area.

° Introduction of new land-use classes to match classes that are used in scenarios and associated N and P leaching lookup tables and developed by the Land Information Group and Landcare Research, and modification of the GIS user interface to import land-use maps with these classes. These classes split land use into sub-classes according to soil characteristics and rainfall.

° Modification to use N and P leaching lookup tables (developed by Landcare Research and ECan) instead of the simplified version of Overseer 6.1 used in default CLUES.

° Modification of the soil types to match types used in the N and P lookup tables, and importation of this information into the CLUES base information for each subcatchment. For P in the Lower Waitaki, the standard CLUES methods will be used (based on the simplified version of Overseer 6.1 used in default CLUES).

° Adjustment of model parameters where appropriate to improve the fit with measurements.

° Adjustment of flow rates to reflect the diversions.

1 ECan instructed that TP calibration not to be done in Lower Waitaki.

6 Waitaki Water Quality Catchment Modelling

This report describes the development and calibration of the customised version of CLUES for current land use. CLUES was calibrated against measured loads which were determined from nutrient concentrations and flow data from a number of sites in the catchment. This, in addition to model development, the water quality and quantity measurements used are described, the sites suitable for calibration are identified, and the modelled or predicted loads are compared with the measured ones. A brief overview of CLUES is also given.

The sub-catchments, streams of order ≥ 3 and the lakes are mapped in Figure 1-1, details about the nutrient point sources and calibration sites are given later in the report.

Waitaki Water Quality Catchment Modelling 7

Figure 1-1: Waitaki River sub-catchments showing streams of order ≥ 3 and lakes. See Table A-1 in Appendix A for the list of sub-catchments (small-case numbers). The border between the upper and lower catchments is shown in bold.

8 Waitaki Water Quality Catchment Modelling

2 Method

2.1 The CLUES model CLUES is a steady-state, spatially-distributed modelling system within ArcGIS which operates at a minimum scale of sub-catchments (~10 km 2 and above). CLUES simulates the effects of land use and farm practices on water quality as indicated by annual loads of total nitrogen, total phosphorus, total suspended sediments and E. coli . The basic spatial unit within CLUES is the River Environments Classification (REC; Snelder et al. 2010) river reach and surrounding contributing area, called a REC unit in this report. CLUES was developed by NIWA for the Ministry of Primary Industries (MPI) and the Ministry for the Environment (MfE) in partnership with AgResearch, Lincoln Ventures, Harris Consulting, Plant and Food Research and Landcare Research. CLUES is provided to users as a front- end interface for ESRI ArcGIS which queries a geo-spatial database that is provided with the model software. The geo-spatial input data comes from a range of national sources including the Land Cover Database, Agribase, and the Land Resource inventory (LRI, Newsome et al. 2008). CLUES has been calibrated nationally against available contaminant load data. The CLUES interface contains tools which allow users to develop land use change and farm practice scenarios. Farm practices are applied to dairy, sheep and beef and deer farming. CLUES results are provided as maps and tables which can be exported to other applications for further analysis or reporting.

The CLUES modelling system couples modified versions of several water quality models which are used for specific modelling tasks, these are:

° OVERSEER® (AgResearch, Wheeler et al. 2006) – a customised, pre-parameterised, and simplified version of OVERSEER® 6 is provided within CLUES which computes nutrient leaching for dairy, sheep and beef, and deer farming. It provides annual average estimates of nutrient losses from these land uses, given information on rainfall, soil order and topography. For other variables, such as stocking and fertiliser application rates, typical values are used based on the region and land use.

° SPARROW (SPAtially Referenced Regression On Watershed attributes) – predicts E. coli and sediment yields from all land uses and nutrient yields from sources other than pasture and horticulture. SPARROW is also used to simulate contaminant transport in the drainage network (i.e., annual average stream loads) and includes provisions for stream routing and loss processes (storage and attenuation). The modelling procedure was originally developed by the United States Geological Survey (Smith et al. 1997) and has since been applied and modified in the New Zealand context with extensive liaison with the developers. SPARROW has been applied to nitrogen and phosphorus routing in Waikato (Alexander et al. 2002) and subsequently to the whole New Zealand landscape (Elliott et al. 2005). The SPARROW sediment transport routines were assessed by (Elliott et al. 2008) and simulations compared favourably with measured sediment load data.

° SPASMO (Soil Plant Atmosphere System MOdel) – calculates the nitrogen budget for a range of horticultural enterprise scenarios. Detailed simulations for many cases (combinations of crops, climate, fertiliser use) have been run (using a daily time step) to build lookup tables that CLUES queries. It has been validated against data from grazed pasture (Rosen et al. 2004) and pasture treated with herbicide (Close et al. 2003, Sarmah et al. 2005).

Waitaki Water Quality Catchment Modelling 9

Further details on the modelling framework can be found in (Woods et al. 2006) and information on setting up and running CLUES scenarios can be found in the model manual (Semadeni-Davies et al. 2011).

2.1.1 Routing along the drainage network As noted above, in addition to simulating sediment and E. coli loads, the SPARROW component of CLUES is used to route contaminants along the drainage network. A schematic of the routing routine in Sparrow is shown in Figure 2-1. Contaminant loads generated by each reach’s REC unit from diffuse sources are added to the instream load for reach. The load of contaminant generated for a particular source type ( source load ) is the product of the amount of source (area of land cover) times a source coefficient (yield for diffuse sources). This source load is then modified by a land-to-water delivery term , which is an exponential function of a number of delivery variables (such as rainfall or land drainage class) and delivery coefficients. An additional source term for TP associated with mass erosion of sediment was also added (Elliott et al. 2008). These modified sources are then summed for a given REC reach to give the total load entering the associated stream reach. Annual loads from any point sources located within the REC unit (as shown in Figure 2-2 and Table 2-1 for Waitaki) are added to total load as a model input. In-stream losses are modelled by a first-order decay term, and the load is then accumulated and attenuated during movement down the reach or stream network. A separate attenuation factor for reservoirs (e.g., lakes), is also calculated (Elliott et al. 2005). The result of the calculations is the modelled mean annual load for each stream reach.

Source Source

Land -to -water Sub -catchment delivery

Lake Point source Monitoring site

Figure 2-1: Schematic of SPARROW sources and transport.

10 Waitaki Water Quality Catchment Modelling

Figure 2-2: Waitaki River catchment showing streams of order ≥ 3, lakes and point sources. See Table 2-1 for a list of point sources (denoted by P).

Waitaki Water Quality Catchment Modelling 11

Table 2-1: Map key and annual nutrient loads for point sources. See Figure 2-2. Loads obtained from ECan which were based on a review of the discharge consents.

Label Name TN load (t/y) TP load (t/y)

P1 Salmon farm at Wairepo Arm/Ohau B 6.000 –

P2 Salmon farm at Ohau B-C 40.000 –

P3 Salmon farm at Ohau A 14.000 –

P4 Salmon farm at Tekapo B 19.400 –

P5 WWTP at Kurow 1.291 0.244

P6 Lake Ohau 0.100 0.025

P7 Twizel 3.371 0.884

P8 Omarama 1.604 0.360

P9 2.129 0.779

P10 Mount Cook Village 0.921 0.200

P11 Otematata 0.921 0.200

2.1.2 CLUES customisation for Waitaki catchment This section describes how the default version of CLUES summarised above was adjusted for application to the Waitaki catchment by refining the characterisation of land use, reconfiguring some of the drainage network and adding flow diversions to simulate the effect of hydropower flow regulation.

Characterisation of geo-spatial data

CLUES represents land use in each REC unit by the percentage of the REC unit area covered by various land use classes including agriculture, crops and horticulture. Soil within each REC unit is represented by the area weighted average drainage class (i.e., poor to well drained) and by the dominant NZ soil order (e.g., ultic, podzols etc.,) taken from the LRI. Rainfall is represented by the reach’s annual average rainfall.

While the spatial scale of the model has remained unchanged, and the model still uses soil and rainfall as described above as model inputs, the customised version of CLUES has refined the characterisation of land use to take into account the influence of intra-unit variations in soil and rainfall on nutrient losses from different land use classes. Moreover, the default stock land uses were extended to include the type of irrigation, dairy for example, has been replaced by dairy and dairy support with no irrigation, with border dykes or with spray irrigation. That is, each REC unit was divided into unique combinations of land use (inclusive of irrigation) soil and rainfall class each with their own set of nutrient generated yields. There are 114 and 531 unique combinations in the Upper and Lower Waitaki catchments respectively. The land use proportions in the default model were then replaced by the proportions of each combination. The spatial disaggregation leading to the various

12 Waitaki Water Quality Catchment Modelling

combinations and their respective nutrient yields were provided by ECan. The proportion of each combination aggregated from REC units by river subcatchment for the calibrated water quality sites is given in Appendix B and the leaching rates for each combination are discussed further below and are given in Appendix C. These refinement enables the customised CLUES to better represent the Waitaki catchment in terms of land use/soil/rainfall with the hope of improving model fit. For example, in the customised Waitaki model, a REC unit which was originally classified in the default version as half and half dairy and arable land may, may be reclassified as 50% by area of dairy with spray irrigation/a medium-deep soil/800 mm of annual rainfall and 50% arable with no irrigation/a light soil/650 mm of annual rainfall. Another example could be a REC unit originally classed as dairy being split in the customised version into a combination of three dairy areas:

° dairy/no irrigation/a poorly drained soil/500 mm of annual rainfall

° dairy/spray irrigation/a light soil/650 mm of annual rainfall

° dairy/borderdyke irrigation/a medium-deep soil/800 mm of annual rainfall. The land use types in the default CLUES have been replaced in the customised version by land use and irrigation classes mapped in Figure 2-3. These were further split by the soil and rainfall classes given in Table 2-2. Table 2-3 and Table 2-4 describe the soil characteristics in Table 2-2 for the Upper and Lower Waitaki catchments respectively. Annual rainfall has been categorised into four bands relating to 475, 500, 650 and 800 mm/year. As described above, this breakdown or disaggregation is for the purpose of applying nutrient yields to each of the unique combinations.

Waitaki Water Quality Catchment Modelling 13

Figure 2-3: The land use in the customised CLUES (ECan, pers. comm .). See Table 2-2 for descriptions of land use, Table B-1 in Appendix B for the percentage of land use by area for each of the watersheds of the calibrated water quality monitoring sites, and Table A-1 in Appendix A for the sub-catchment.

14 Waitaki Water Quality Catchment Modelling

Table 2-2: Combinations of land use, rainfall and soil types in the Upper and Lower Waitaki River catchments used in the customised version of CLUES (ECan, pers. comm .). UW Land use in Upper Waitaki (UW). LW Land use in Lower Waitaki (LW). 1 475, 500, 650 and 800 are millimetres of annual rainfall. See Table 2-3 and Table 2-4 for a description of the soils.

Land use Description Rainfall for Rainfall Soils for UW Soils for LW UW 1 for LW 1

Arable UW Arable UW475, Light, Medium- UW650, deep, Poorly UW800 drained

Arable_Bdyke_Irr LW Arable, borderdyke Blank, Class 4 to LW500, irrigation 6, Class 6, Class LW650, 7, H, L, M, Pd, LW800 PdL, VL, XL

Arable_Dry LW Arable, no irrigation LW500, Blank, Class 4 to LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

Arable_Spray_Irr LW Arable, spray LW500, Blank, Class 4 to irrigation LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

Dairy_Bdyk UW, LW Dairy, borderdyke UW475, LW500, Light, Medium- Blank, Class 4 to irrigation UW650, LW650, deep, Poorly 6, Class 6, Class UW800 LW800 drained 7, H, L, M, Pd, PdL, VL, XL

Dairy_Dry LW Dairy, no irrigation LW500, Blank, Class 4 to LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

Dairy_Spra UW, LW Dairy, spray irrigation UW475, LW500, Light, Medium- Blank, Class 4 to UW650, LW650, deep, Poorly 6, Class 6, Class UW800 LW800 drained 7, H, L, M, Pd, PdL, VL, XL

DairySuppB UW, LW Dairy support, UW475, LW500, Light, Medium- Blank, Class 4 to borderdyke irrigation UW650, LW650, deep, Poorly 6, Class 6, Class UW800 LW800 drained 7, H, L, M, Pd, PdL, VL, XL

DairySuppS UW, LW Dairy support, spray UW475, LW500, Light, Medium- Blank, Class 4 to irrigation UW650, LW650, deep, Poorly 6, Class 6, Class UW800 LW800 drained 7, H, L, M, Pd, PdL, VL, XL

Waitaki Water Quality Catchment Modelling 15

Land use Description Rainfall for Rainfall Soils for UW Soils for LW UW 1 for LW 1

DairySuppD UW, LW Dairy support, no UW475, LW500, Light, Medium- Blank, Class 4 to irrigation UW650, LW650, deep, Poorly 6, Class 6, Class UW800 LW800 drained 7, H, L, M, Pd, PdL, VL, XL

Deer_BDyke LW Deer, borderdyke LW500, Blank, Class 4 to irrigation LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

Deer_Dry LW Deer, no irrigation LW500, Blank, Class 4 to LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

Deer_Spray LW Deer, spray irrigation LW500, Blank, Class 4 to LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

SBF_BDyke UW Sheep & beef on the UW475, Light, Medium- flat, borderdyke UW650, deep, Poorly irrigation UW800 drained

SB_BDyke LW Sheep & beef, LW500, Blank, Class 4 to borderdyke irrigation LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

SBF_Dry UW Sheep & beef on the UW475, Light, Medium- flat, no irrigation UW650, deep, Poorly UW800 drained

SB_Dry LW Sheep & beef, no LW500, Blank, Class 4 to irrigation LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

SBF_Spray UW Sheep & beef on the UW475, Light, Medium- flat, spray irrigation UW650, deep, Poorly UW800 drained

SB_Spray LW Sheep & beef, spray LW500, Blank, Class 4 to irrigation LW650, 6, Class 6, Class LW800 7, H, L, M, Pd, PdL, VL, XL

SBH_Dry UW Sheep & beef on hill UW475, Light, Medium- country, no irrigation UW650, deep, Poorly UW800 drained

16 Waitaki Water Quality Catchment Modelling

Land use Description Rainfall for Rainfall Soils for UW Soils for LW UW 1 for LW 1

SBF_Dry_Dev UW Sheep & beef on the UW475, Light, Medium- flat, no irrigation, UW650, deep, Poorly developed land UW800 drained

SBH_Dry_Dev UW Sheep & beef on hill UW475, Light, Medium- country, no UW650, deep, Poorly irrigation, developed UW800 drained land

ExoticFor UW, LW Exotic forest UW475, LW500, Light, Medium- All, Blank UW650, LW650, deep, Poorly UW800 LW800 drained

NativeFor UW, LW Native forest UW475, LW500, Light, Medium- All, Blank UW650, LW650, deep, Poorly UW800 LW800 drained

Other UW, LW Other land covers, UW475, LW500, Light, Medium- All, Blank e.g. ice, bare soil, UW650, LW650, deep, Poorly rock UW800 LW800 drained

Tussock UW, LW Tussock UW475, LW500, Light, Medium- All, Blank UW650, LW650, deep, Poorly UW800 LW800 drained

Scrub UW, LW Scrub UW475, LW500, Light, Medium- All, Blank UW650, LW650, deep, Poorly UW800 LW800 drained

Urban UW, LW Urban UW475, LW500, Light, Medium- All, Blank UW650, LW650, deep, Poorly UW800 LW800 drained

Table 2-3: Soils in the Upper Waitaki (ECan, pers. comm.).

Name Available water to a 60 cm depth (mm)

Light <60

Medium-deep ≥ 60

Poorly drained Poorly drained or very poorly drained

Waitaki Water Quality Catchment Modelling 17

Table 2-4: Soils in the Lower Waitaki (ECan, pers. comm .).

Name Assumption

All A convenience group – used for a set of non-productive land uses.

Blank Gaps in the soil layer, most commonly riverbed/water/rock

Class 4 to 6 Hill soils – equivalent to light soils in Upper Waitaki

Class 6 Hilly/steep soils – equivalent to light soils in Upper Waitaki

Class 7 Steep soils – equivalent to light soils in Upper Waitaki

XL Equivalent to light soils in Upper Waitaki

VL Equivalent to light soils in Upper Waitaki

L Equivalent to light soils in Upper Waitaki

M Equivalent to medium-deep soils in Upper Waitaki

H Equivalent to medium-deep soils in Upper Waitaki

Pd Equivalent to poorly drained soils in Upper Waitaki

PdL Equivalent to poorly drained soils in Upper Waitaki

TN and TP leaching rates

As stated above, leaching rates were provided by ECan and varied according to the land use, soil type and rainfall combinations; these are given in Table C-1 and Table C-2 in Appendix C.

In the default version of CLUES, leaching rate tables are held in a geodatabase that is inaccessible to the user. For the Waitaki customised version, they were made available via a lookup table that the user can access and alter if necessary. The lookup table consists of three columns: the first lists the land use/soil/rainfall combinations, and the next two columns have the corresponding leaching rates for N and P.

Flows

The default version of CLUES uses estimated flows from Woods et al. (2006). The customised version instead uses a combination of measured flows where available and estimated flows. Some of the estimated flows from the default version were also updated in the customised version using more recent data from flow monitoring sites. Corrections between the old flows and updated flows were done by calculating the percentage difference between old and updated values at the flow monitoring station, then propagating this difference to all the downstream reaches, right down to the sea. Table 2-5 and Table 2-6 compare the updated flows in the customised version with the flows in the default version used for the water quality calibration sites.

For example, the flow at bottom of the Upper Waitaki catchment (see Waitaki at Kurow site in Table 2-5) for the default and customised versions was 435 and 354 m 3/s respectively, with the measured flow being 360 m 3/s. Similarly, the flow at the bottom of the Lower Waitaki catchment, i.e., at the

18 Waitaki Water Quality Catchment Modelling

coast, for the default and customised versions was 449 and 371 m 3/s respectively. The measured flow at Waitaki at SH1 Bridge which is a site near the coast is 367 m 3/s. Therefore the customised version better fits the measured flow better at the bottom of the Upper and Lower Waitaki catchments.

The fit of the default and customised versions to the measured flows is described using the root mean square error (RMSE) 2 for the Upper and Lower Waitaki (Table 2-5 and Table 2-6). The RMSE is calculated using the following equation (Chai & Draxler 2014):

RMSE = √∑ − / (2.1) The RMSE for the default and customised versions of CLUES equals 62 and 3 respectively. This means that the customised version is a better fit to the measured data than the default version.

2 The RMSE is used as a standard statistical metric to measure model performance in many fields, including meteorology, air quality, climate research and agriculture. It assumes the errors (= modelled – measured) are unbiased and follow a normal distribution (Chai & Draxler 2014).

Waitaki Water Quality Catchment Modelling 19

Table 2-5: Measured, default CLUES and customised CLUES flows for the water quality calibration sites in the Upper Waitaki. . The water quality site used the flow site(s) in the same row to calculate the nutrient loads. Water quality sites without superscripts 1 or 2 have coincident flow data.

Flow site name Flow site ID Mean measured Flow from Flow from Water quality site Water quality site Approximate flow (m3/s), date default CLUES customised CLUES name ID distance range (m3/s) (m3/s) between flow site and water quality site (m)

– 19.72 16.73 Ahuriri River M1 0 Headwater (upper Ahuriri) 1

Ahuriri at South 71116 23.27, 25.24 23.89 Ahuriri River M3 32500 3 2 Diadem 1/1/1990– mouth 31/12/2013

Grays River at 1875 2.94, 5.09 3.60 Grays River - SQ35117 0 Above Tekapo 1/1/1990– Lower Above Ford Confluence 31/12/2013

Hen Burn @ U/S 2146 0.78, 0.54 0.63 Henburn Rd 2 SQ10824 600 Quail Burn Downs 15/10/1986– Intake 17/12/2013

Mary Burn at SH8 71130 0.57, 0.93 0.74 Maryburn SH8 SQ10275 0 12/6/2003– Bridge 14/1/2014

Mary Burn @ 1872 2.75, 2.89 3.02 Maryburn Stream SQ35115 0 Above Tekapo 1/1/1990– - Lower u/s Confluence 31/12/2013 Tekapo River

20 Waitaki Water Quality Catchment Modelling

Flow site name Flow site ID Mean measured Flow from Flow from Water quality site Water quality site Approximate flow (m3/s), date default CLUES customised CLUES name ID distance range (m3/s) (m3/s) between flow site and water quality site (m)

Ohau C Power 38748 258.15, 0.0007 266.00 Ohau C M10 0 Station at 22/9/1985– Machine Output 6/1/2014

Omarama at 71136 2.09, 1.99 2.54 Omarama Stream SQ10005 400 2 Wardells Bridge 3/8/1988– Omarama (SH8) 31/1/2014

Quailburn 71118 0.89, 0.88 0.55 Quailburn Road SQ35792 0 15/10/1986– Recorder 17/12/2013

– 0.24 0.18 Sutherlands Creek SQ10037 0 Ben Omar Road 1

Forks at Balmoral 71129 3.17, 3.82 3.16 9600 1/1/1990– 31/12/2013

Tekapo at Spillway 71131 14.95, 110.21 5.00 Tekapo at Forks 2 M11 7700 6/6/1968– 7/11/2013

Total = 18.12 Total = 114.03 Total = 8.16

Waitaki Water Quality Catchment Modelling 21

Flow site name Flow site ID Mean measured Flow from Flow from Water quality site Water quality site Approximate flow (m3/s), date default CLUES customised CLUES name ID distance range (m3/s) (m3/s) between flow site and water quality site (m)

Mary Burn @ 1872 2.75, 2.89 3.02 1600 Above Tekapo 1/1/1990– Confluence 31/12/2013

Tekapo Grays River at 1875 2.94, 5.09 3.60 downstream M4 3500 Above Tekapo 1/1/1990– Greys Hill 2 Confluence 31/12/2013

Total = 5.69 Total = 7.98 Total = 6.62

– 300.39 31.05 Tekapo Pukaki M12 0 Mouth 1

Waitaki at Kurow 71104 360.32, 435.49 354.10 Waitaki at Kurow 71104 0 7/9/1964– 18/11/2013

Willow Burn at 1071110 0.92, 0.33 0.66 Willowburn SQ10012 0 Quailburn Rd 14/6/2013– Quailburn Road 14/1/2014 Bridge

Mary Burn at Mt 71122 0.55, 0.72 0.55 Mary Burn Mary SQ26372 1200 2 MacDonald 1/1/1990– Burn Fill 31/12/2013

– 1.14 0.60 Wairepo Creek SQ10804 0 Arm Inlet 1

22 Waitaki Water Quality Catchment Modelling

1 No measured flows so CLUES flow used. 2 Flow and water quality sites not coincident but data from the closest flow monitoring site were adjusted and used. 3 See Footnote in Section 2.2.1.

Table 2-6: Measured, default CLUES and customised CLUES flows for the water quality calibration sites in the Lower Waitaki. The water quality site used the flow site(s) in the same row to calculate the nutrient loads. Water quality sites without superscripts 1 or 2 have coincident flow data.

Flow site name Flow site ID Mean measured Flow from Flow from Water quality site Water quality site Approximate flow (m3/s), date default CLUES customised CLUES name ID distance range (m3/s) (m3/s) between flow site and water quality site (m)

Cattle Creek at 73 0.47, 0.09 0.21 Cattle Creek SQ10814 4500 Cattle Yards 1/1/1990– Morland 2 31/12/2013 Settlement Road

– 0.01 0.12 Rocky Point SQ10816 0 Stream Hakataramea Valley Road 1

– 0.02 0.03 Deadman Stream SQ10818 0 Hakataramea Valley Road 1

– 0.34 0.17 Penticotico Stream SQ10174 0 SH83 1

– 0.32 1.05 Otiake River Mt SQ10167 0 Bell Station 1

Waitaki Water Quality Catchment Modelling 23

Flow site name Flow site ID Mean measured Flow from Flow from Water quality site Water quality site Approximate flow (m3/s), date default CLUES customised CLUES name ID distance range (m3/s) (m3/s) between flow site and water quality site (m)

Maerewhenua 2006 2.49, 2.18 2.57 Maerewhenua SQ10160 0 River at SH83 1/1/1990– River Duntroon Bridge (Duntroon) 31/12/2013

– 0.16 0.15 Waikakahi Stream SQ21160 0 Cock & Hen Road 1

Hakataramea at 71103 5.63, 6.78 5.67 Hakataramea at 71103 0 Above MHBr 1/1/1990– Above MHBr 22/10/2014

– 0.003 0.003 Waikakahi Stream SQ21255 0 Old Ferry Road 1

Otekaieke River at 71102 2.48, 1.05 1.14 Otekaieke River SQ35871 1800 Weir 30/1/2013– D/S of Bushy 2 16/7/2014 Creek

Waikakahi at Te 71195 0.59, 0.56 0.48 Waikakahi Stream SQ21162 0 Maiharoa Rd 23/5/2001– Glenavy Tawai Rd 4/8/2014

Waikakahi at Te 71195 0.59, 0.56 0.48 Waikakahi Stream SQ21254 0 Maiharoa Rd 23/5/2001– Te Maiharoa Road 4/8/2014

1 No measured flows so CLUES flow used. 2 Flow and water quality sites not coincident but data from the closest flow monitoring site were adjusted and used

24 Waitaki Water Quality Catchment Modelling

Flow diversions

The flow diversions due to the hydro scheme were introduced in the customised version of CLUES, with Table 2-7 giving a summary. There are four hydro dams in the catchment and flow regulation from these has changed the natural flow paths requiring flow diversion to be taken into account in the model (which doesn’t happen in the default). Diversions range from 85% of the annual flow from Lake Ohau to 99% of the annual flow from (see Table 2-7).

The diversions were modelled by splitting the flow and nutrient loads from the outlet reach of each lake between the next REC reach in the REC flow sequence and a second reach representing the divert. Details on how the flow distributions were estimated are given in Appendix D.

Table 2-7: Estimated flow diversions in the Waitaki River catchment.

Lake or river Outlet Flow distribution (%)

Tekapo Canal 95 Lake Tekapo Tekapo River 5

Ohau C Canal 99 Lake Ruataniwha (including Wairepo Arm) Lower Ohau River 1

Ohau A Canal 85 Lake Ohau Upper Ohau River 15

Pukaki Canal 95 Lake Pukaki Pukaki River 5

Ahuriri River 70 Wairepo Creek Wairepo Arm 30

2.2 Measured water quality and flow data The flow and TN and TP concentration data that were used to calculate the measured mean annual loads came from three sources: ECan, NIWA and Meridian Energy. These data are described below.

2.2.1 Flows

Load for a point in space and time is calculated as the product of flow and concentration, and for this reason flow data were needed for the calculation of loads. However, as seen in Table 2-5 and Table 2-6, not all the water quality monitoring sites have coincident flow data recorded at the site. For some of these water quality sites, flow data from the closest flow monitoring site were adjusted and used instead, see below for details. For the rest of these water quality sites, i.e., ones that had no

Waitaki Water Quality Catchment Modelling 25

flow monitoring site close by, CLUES estimated flows (Woods et al. 2006) were used for the load calculations.

For those water quality sites that have no measured flow data but have a flow monitoring site close by, an estimate of the measured flow at the water quality site was calculated as follows:

° the modelled mean flow at the water quality site (Q(mean wq )mod ) from Woods et al. 3 (2006) was divided by the modelled mean flow (Q(mean flow )mod ) at the closest flow site (i.e., has measured flow data) – this quotient called ‘multiplier’.

° The ‘multiplier’ was then applied to the measured flow at the flow site (Q(flow) meas ) to

obtain an estimate of the measured flow at the water quality site (Q(wq) meas ).

= × (2.2)

From this equation, a multiplier >1 signals that the closest flow site is upstream from the water quality site, conversely, a multiplier <1 signals that the closest flow site is downstream from a water quality sit. If the water quality site is also a flow site then the multiplier = 1.

2.2.2 Median concentrations Median TN and TP concentrations were calculated for those water quality sites which had at least 10 data points and the percentage of censored data (i.e., concentration

The concentration data for each site and nutrient were first examined for the influence of temporal trends by fitting a linear trend line to the log-transformed concentration data. Temporal trends could be due to factors like land use change (e.g., conversion to dairy farming) or mitigation measures (e.g., fencing of streams). The median concentrations adopted depend on this fitted trend as follows:

° If no trend was visible, then an ‘unadjusted’ or ‘not-detrended’ median concentration was calculated using the measured concentration data with censored data being set to both the detection limit (DL) and half the detection limit (DL/2), giving two median concentrations – one for DL and one for DL/2. Mostly the median concentrations calculated using the two different detection limits for the censored data were similar.

3 The only exception to this was the water quality site at the Ahuriri River mouth (M3), where instead of using the closest flow site (the SH8 Recorder at the Ahuriri River), the more upstream site at South Diadem was used and a multiplier of 1.206 was applied as per Dr Bob Spigel’s advice (Dr Bob Spigel, NIWA, pers. comm. ).

26 Waitaki Water Quality Catchment Modelling

° However, if a temporal trend was evident, then an ‘adjusted’ or a ‘detrended’ median concentration was calculated based on adjusted values for the measured concentration data to avoid the present day median being dragged either up or down depending on the direction of the trend. Details of the adjustments are in Appendix E.

The selection of which site’s concentration data in Table 2-8 and Table 2-9 were suitable for calibration is discussed in Section 2.3.

Waitaki Water Quality Catchment Modelling 27

Table 2-8: Measured median concentrations for TN and TP in the Upper Waitaki catchment.

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Ahuiri River M1 20/12/2011–26/2/2013 18, 0% 0.0520, ND 20/12/2011–26/2/2013 18, 0% 0.0062, D Headwater (upper Ahuriri)

Ahuriri River M3 9/12/2008–26/2/2013 26, 0% 0.1297, D 9/12/2008–26/2/2013 26, 0% 0.0107, D mouth

Grays River - SQ35117 22/7/2010–12/6/2012 23, 57% 0.0713, ND 22/7/2010–12/6/2012 23, 4% 0.0160, ND Lower Above Ford

Henburn Rd SQ10824 15/12/2004–13/6/2012 28, 7% 0.3100, ND 15/12/2004–13/6/2012 28, 0% 0.0170, ND

Maryburn SH8 SQ10275 11/7/2005–16/12/2013 55, 44% 0.0990, D 11/7/2005–16/12/2013 54, 6% 0.0141, D Bridge

Maryburn Stream - SQ35115 22/7/2010–12/6/2012 23, 52% 0.0900, ND 22/7/2010–12/6/2012 23, 39% 0.0080, ND Lower u/s Tekapo River

Ohau C M10 9/12/2008–26/2/2013 26, 0% 0.0557, D 9/12/2008–26/2/2013 27, 0% 0.0056, D

28 Waitaki Water Quality Catchment Modelling

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Omarama Stream SQ10005 16/11/2005–7/12/2013 42, 2% 0.2805, D 16/11/2005–17/12/2013 42, 0% 0.0161, D Omarama (SH8)

Quailburn Road SQ35792 20/7/2011–17/12/2013 21, 33% 0.1798, D 20/7/2011–17/12/2013 21, 5% 0.0156, D Recorder

Sutherlands Creek SQ10037 15/12/2004–17/12/2013 46, 9% 0.3188, D 15/12/2004–17/12/2013 46, 0% 0.0287, D Ben Omar Road

Tekapo at Forks M11 20/12/2011–26/2/2013 18, 0% 0.0400, ND 8/12/2008–26/2/2013 18, 0% 0.0102, D

Tekapo M4 20/12/2011–26/2/2013 18, 0% 0.0934, D 20/12/2011–26/2/2013 18, 0% 0.0149, D downstream Greys Hill

Tekapo Pukaki M12 8/12/2008–26/2/2013 27, 0% 0.1408, D 8/12/2008–26/2/2013 27, 0% 0.0045, D Mouth

Waitaki at Kurow 71104 17/2/1994–12/12/2013 226, 0% 0.0567, D 17/2/1994–12/12/2013 239, 0% 0.0034, D

Waitaki Water Quality Catchment Modelling 29

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Willowburn SQ10012 15/4/2004–17/12/2013 58, 0% 1.0251, D 15/12/2004–17/12/2013 58, 0% 0.0301, D Quailburn Road Bridge

Mary Burn Mary SQ26372 30/10/2002–12/6/2012 29, 52% 0.0055, D Burn Fill

Wairepo Creek SQ10804 29/10/2003–22/10/2013 48, 46% 0.0988, D 29/10/2003–22/10/2013 49, 31% 0.0076, D Arm Inlet

Spring Creek SQ10851 29/8/2005–22/1/2014 37, 43% 0.1100, ND 29/8/2005–22/1/2014 36, 50% 0.0068, ND Glenbrook Boundary

Spring Creek SH8 SQ10206 29/8/2005–22/1/2014 37, 30% 0.1366, D

Wairepo Creek M14 20/12/2011–26/2/2013 17, 0% 0.1100, D 20/12/2011–26/2/2013 17, 0% 0.0092, D

Willowburn Above SQ35657 20/1/2010–22/1/2014 17, 0% 1.9131, D 20/1/2010–22/1/2014 17, 0% 0.0236, D Willowburn Boundary

Willowburn SQ10853 29/8/2005–22/1/2014 36, 31% 0.1611, D 29/8/2005–22/1/2014 37, 3% 0.0330, D Benmore Boundary

30 Waitaki Water Quality Catchment Modelling

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Willowburn Glens SQ10854 29/8/2005–22/1/2014 37, 3% 0.2933, D 29/8/2005–22/1/2014 37, 0% 0.0655, D Boundary

Willowburn Trib SQ10856 29/8/2005–22/1/2014 37, 5% 2.0833, D 29/8/2005–22/1/2014 36, 3% 0.0148, D Buscot Boundary

Willowburn SQ10855 29/8/2005–22/1/2014 37, 0% 1.5646, D 29/8/2005–22/1/2014 36, 8% 0.0129, D Boundary

Quailburn SQ10823 15/12/2004–22/1/2013 48, 17% 0.0137, D Henburn Road

Quailburn Road SQ10826 15/12/2004–23/6/2011 16, 0% 0.0152, D

Upper Wairepo SQ10852 29/8/2005–22/1/2014 37, 19% 0.0113, D Creek Pylon Line

Waitaki Water Quality Catchment Modelling 31

Table 2-9: Measured median concentrations for TN and TP in the Lower Waitaki catchment.

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Cattle Creek SQ10814 25/11/2004–12/8/2014 53, 38% 0.1992, D Morland Settlement Road

Rocky Point Stream SQ10816 25/11/2004–23/1/2013 36, 14% 0.1973, D 25/11/2004–23/1/2013 n = 36, 0% 0.0308, D Hakataramea Valley Road

Deadman Stream SQ10818 25/11/2004–12/8/2014 52, 4% 0.6368, D 25/11/2004–12/8/2014 52, 0% 0.0760, ND Hakataramea Valley Road

Penticotico Stream SQ10174 16/11/2005–12/8/2014 50, 0% 1.2050, ND 16/11/2005–12/8/2014 49, 39% 0.0070, D SH83

Otiake River Mt SQ10167 20/12/2005–12/8/2014 42, 55% 0.0593, D Bell Station

Maerewhenua SQ10160 20/12/2005–12/8/2014 50, 24% 0.1700, ND River Duntroon

Waikakahi Stream SQ21160 19/8/1996–5/8/2014 138, 0% 2.8000, ND 7/9/1995–5/8/2014 147, 0% 0.0609, D Cock & Hen Road

32 Waitaki Water Quality Catchment Modelling

TN TP

WQ site Median Median WQ site name ID concentration concentration n, (mg/L), n, (mg/L), Date range Date range % censored detrended (D), % censored detrended (D), not-detrended not-detrended (ND) (ND)

Hakataramea at 71103 19/1/1995–12/12/2013 222, 0% 0.1151, ND 17/2/1994–12/12/2013 233, 0% 0.0067, ND Above MHBr

Waikakahi Stream SQ21255 19/8/1996–5/8/2014 86, 0% 2.8000, ND 7/9/1995–5/8/2014 95, 0% 0.0719, D Old Ferry Road

Otekaieke River SQ35871 16/4/2013–12/8/2014 17, 0% 0.1733, D 16/4/2013–12/8/2014 17, 53% 0.0052, ND D/S of Bushy Creek

Waikakahi Stream SQ21162 9/4/2008–16/6/2011 20, 0% 2.2705, D 9/4/2008–16/6/2011 20, 0% 0.0655, ND Glenavy Tawai Rd

Waikakahi Stream SQ21254 19/8/1996–5/8/2014 158, 0% 2.3000, ND 7/9/1995–5/8/2014 167, 0% 0.0763, D Te Maiharoa Road

Waitaki at SH1 71101 19/1/1995–12/12/2013 227, 0% 0.1250, ND 17/2/1994–12/12/2013 237, 0% 0.0091, ND Bridge

Kirkliston Stream SQ10819 25/11/2004–23/1/2013 36, 0% 0.0295, ND

Waitaki Water Quality Catchment Modelling 33

2.2.3 Loads Nutrient loads were calculated using two methods. The rating curve method based on Generalised Additive Modelling (GAM) methods was used preferentially. However if no rating curve was available for a site, then the less accurate ratio method was used. The methods are described below.

Rating curve method Flows ( Q) at the various water quality sites were calculated as described in Section 2.2.1. Rating curves were then generated using GAM modelling methods (Hastie & Tibshirani 1990). The rating curve formula used was of the form:

(2.3) ̂ = + + sin + cos where represents the log-transformed concentrations, represents the log-transformed ̂ flow values (interpolated to match the water quality sampling times), and the terms denote GAM smoothers. The rating curves were used to predict concentration values matching the times in the original flow record, which were then aggregated to produce mean annual loads. Note that the rating curves were generated using flow values (interpolated if necessary) at the same times as the water quality data, but after that the mean annual loads were calculated based on the times in the flow record. Thus, the rating curves were generated using only the portion of the flow record and the water quality record that overlapped, but the mean annual load calculations used the full length of the flow record.

The GAM statistical routines used to generate the rating curves do not allow for censored data, therefore a number of different techniques were explored to examine the loads calculated for sites with more than a few censored data points. First, the rating curve method described above was carried out with censored data set equal to the detection limit (DL). To try to assess the impact of substitution, rating curves were then recalculated with the censored data set equal to half the detection limit (DL/2). Generally, the mean annual loads calculated were similar (<20% difference), though some sites showed a substantial difference. Second, repeated iterations of the rating curve method were employed with the censored data replaced by random permutations of the modelled set of values from the lognormal distribution (see discussion of the distributed method below).

Furthermore, two sets of mean annual loads were calculated based on whether the median concentrations adopted were detrended (adjusted) or not-detrended (unadjusted), as decided by the existence of temporal trends discussed in Section 2.2.2. For sites where the detrended median concentration was adopted, mean annual loads were calculated based on the rating curve formula in Equation 2.3. Alternatively, for sites where the unadjusted median concentration was adopted, mean annual loads were calculated based on the rating curve formula in Equation 2.3 with the term omitted.

Estimates of the uncertainty in the two sets of mean annual loads using rating curves were calculated by bootstrapping and distributed techniques as follows.

Bootstrap technique Confidence intervals for the mean annual loads were produced by repeating the rating curve procedure using bootstrap resampling. For those sites with no censored data, the bootstrap results were used preferentially to calculate the mean annual load as they give an assessment of the uncertainty. For those sites with censored data, substitution of DL followed by DL/2 was used. For some sites, the bootstrap results were not considered suitable for calibration, either because:

34 Waitaki Water Quality Catchment Modelling

° the confidence intervals were extremely large (which could be due to the limited number of data points and the degree of censoring);

° most of the flows used in Equation 2.3 were high flows; or

° the standard deviation of the results was too big.

Distributed technique The distributed method was only employed for those sites with censored data. The concentration results from the distributed method were used for calculating the mean annual loads, provided the bootstrap results were acceptable for the same site. For the distributed method, the censored data were replaced by the set of modelled values from the Regression on Order Statistics (ROS) calculations outlined in Appendix E for calculating the median concentrations and rating curves were then applied. The set of modelled values for the censored data were drawn from the quantiles of the lognormal distribution and should not be assigned to any particular observations. As such, the method was employed using 100 iterations, with a random permutation of the modelled set assigned to the censored observations in each iteration.

Ratio method This method is not as accurate as that using rating curves as described above, but given that only some of the water quality sites were deemed suitable as calibration sites using the rating curve method, the ‘ratio’ method was invoked for those sites that did not. The total mean annual load L is given by:

× cmedian Qmean L = (2.4) ratio

where is the measured median concentration (see Section 2.2.2), is the mean flow, and is the median concentration divided by the flow-weighted mean concentration. The concentrations used for the were measured if available, or if unavailable then the modelled ratio from CLUES was used.

Waitaki Water Quality Catchment Modelling 35

2.3 Determination of suitable calibration sites This section describes the method used to identify sites suitable for model calibration. Besides excluding any water quality site with <10 samples or ≥ 60% of its data being censored from being considered in the measured loads calculations (see Section 2.2.2), sites whose measured loads had high uncertainty were also not included. High uncertainty in this context means either the natural logarithm of the standard deviation of the mean annual load (as determined using the rating curve method described above) exceeded 1, or the fraction of flows (associated with the loads) below the 99 percentile for flows was ≤ 0.3, or the fraction of the loads that came from flows below the highest sample flow was ≤ 0.3.

Of the 40 and 32 water quality stations in in the Upper and Lower Waitaki catchments respectively, taking into account censoring (see Section 2.2.2) and uncertainty, the number of sites suitable for calibration are 15 for TN and 17 for TP in the Upper Waitaki and 12 for TN in the lower Waitaki (Figure 2-4).

A ‘reliability score’ was given to each site selected for calibration. This reliability was used to guide calibration and was assigned in the following way:

° Sites with loads obtained using the ‘rating curve’ method were assigned a ‘reliability score’ of either 1 (the highest) or 2 4.

° For the loads calculated using the ‘ratio’ method, if the mean flow and the ratio were measured, then the load was given a ‘reliability score’ of 3, and if they were estimated using the CLUES model, the load was given a ‘reliability score’ of 4 (the lowest).

Reliability scores are discussed further in the results section below.

2.3.1 Other excluded data In addition to the sites excluded on the basis of censored data and uncertainty, two further sites were excluded from the study.

Site at SH1 Bridge in Lower Waitaki, in the lower reaches of the Waitaki catchment (near site C15 in Figure 2-4) was excluded from the calibration sites for the following reasons: ° The (mostly measured) load between the SH1 Bridge and Kurow (site C29 in Figure 2-4) sites originating from the tributaries between the sites was much less than indicated by the difference in measured loads between the two sites. The model considerably under-estimated the load at SH1 Bridge site (modelled 1076 t/y compared with the measured of 1858 t/y).

° ECan groundwater scientists advised that the SH1 Bridge site is located on the Waitaki fan where they expect quite a lot of groundwater/surface water interaction which cannot be modelled using CLUES. Mr Ned Norton (ECan, pers. comm. ) also commented that water quality sampling at this site is undertaken on the edge of a side braid which may receive more nitrogen from riverside spring and groundwater inputs – compared to the central braids – the implication being that measured loads at that site could be higher than would be expected from the upstream surface flows. If this is the case,

4 Those sites allocated a reliability score of 2 either had a small number of samples (10 ≤ n ≤ 20), or 0.3 < the fraction of flows (associated with the loads) below the 99 percentile for flows ≤ 0.5, or 0.3 < the fraction of the loads that came from flows below the highest sample flow ≤ 0.5.

36 Waitaki Water Quality Catchment Modelling

calibrating to this site could result in over-estimation of modelled nitrogen concentrations.

The Awamoko Stream SH83 monitoring site in Lower Waitaki was also excluded as there were no concurrent measured flow or water quality data available for the site such that a rating curve could not be generated. Flow monitoring at the site ceased in 1995, whereas the water quality record did not start until 2003 (except for a single data point in 1983).

Table 2-10: Map key for calibration sites. The Upper Waitaki sites C8 and C17 were calibrated just for TP, since they were found to be unsuitable as calibration sites for TN (see Section 2.3). The remainder of the Upper Waitaki sites were calibrated for both TN and TP. The Lower Waitaki sites were just calibrated for TN as instructed by ECan.

Label Name, site ID Label Name, site ID Label Name, site ID

C1 Omarama Stream C2 Willowburn Quailburn C3 Sutherlands Creek Ben Omarama (SH8), SQ10005 Road Bridge, SQ10012 Omar Road, SQ10037

C4 C5 Otiake River Mt Bell C6 Penticotico Stream Duntroon, SQ10160 Station, SQ10167 SH83, SQ10174

C7 Maryburn SH8 Bridge, C8 Wairepo Creek Arm Inlet, C9 Cattle Creek Morland SQ10275 SQ10804 Settlement Road, SQ10814

C10 Rocky Point Stream C11 Deadman Stream C12 Henburn Rd, SQ10824 Hakataramea Valley Road, Hakataramea Valley Road, SQ10816 SQ10818

C13 Waikakahi Stream Cock & C14 Waikakahi Stream Te C15 Waikakahi Stream Hen Road, SQ21160 Maiharoa Road, SQ21254 Glenavy Tawai Rd, SQ21162

C16 Waikakahi Stream Old C17 Mary Burn Mary Burn Fill, C18 Maryburn Stream - Ferry Road, SQ21255 SQ26372 Lower u/s Tekapo River, SQ35115

C19 Grays River - Lower Above C20 Quailburn Road Recorder, C21 Otekaieke River D/S of Ford, SQ35117 SQ35792 Bushy Creek, SQ35871

C22 Ahuriri River mouth, M3 C23 Ohau C, M10 C24 Tekapo Pukaki Mouth, M12

C25 Tekapo downstream Greys C26 Tekapo at Forks, M11 C27 Ahuiri River Headwater Hill, M4 (upper Ahuriri), M1

C28 Hakataramea at Above C29 Waitaki at Kurow, 71104 MHBr, 71103

Waitaki Water Quality Catchment Modelling 37

Figure 2-4: Waitaki River sub-catchments showing streams of order ≥ 3, lakes and calibration sites. See Table 2-10 for a list of sites (denoted by C).

38 Waitaki Water Quality Catchment Modelling

2.4 Calibration The default CLUES model was calibrated nationally which has led to model parameters in some regions being unable to represent local conditions. The customised version of CLUES has been calibrated here using only data from the Waitaki Catchment.

For the Upper Waitaki, the measured loads from 15 (TN) and 17 (TP) of the 40 monitoring sites (refer to Figure 2-4) were used to adjust some of the key coefficients of the SPARROW component within CLUES in order to obtain the best fit of the modelled to the measured loads. Similarly, 12 of the 32 monitoring sites in the Lower Waitaki were used to calibrate for TN (Figure 2-4). No calibration was undertaken in the Lower Waitaki for TP. Calibration was done manually by adjusting selected parameters – attenuation coefficients, lake residence time of TN and TP and some leaching rates (see below) – to improve the fit between measured and modelled loads. The reliability of measured loads was taken into account during this calibration process, giving more weight to sites with higher reliability (see above for a description of reliability scores). That is, more effort was put into getting a better fit between measured and modelled loads for those sites with higher reliability scores than those with lower scores. Formal calibration methods were not used, due to the small amount of reliable data.

For the Upper Waitaki the following leaching rates supplied by ECan were modified (see Table 2-11):

° ‘Sheep & beef on the flat, no irrigation’ and ‘Sheep & beef on the flat, no irrigation, developed land’ – both TN and TP leaching rates halved from the original values. This was done to improve the model fit to sites with low-intensity pasture, which were otherwise over-estimated. The adjustment was limited to a factor of two, to avoid too great a departure from the Overseer-based values provided by ECan.

° TN leaching for ‘Dairy support, no irrigation’ was lowered to that for ‘Sheep & beef on the flat, no irrigation’ to avoid over-prediction.

° Similarly TP leaching for ‘Dairy support, no irrigation’ was lowered to that for ‘Sheep & beef on the flat, no irrigation’ X 2 to avoid over-prediction. Originally ECan had set the TN and TP leaching rates for ‘Dairy support, no irrigation’ which is equal to that for ‘Dairy support, spray irrigation’ which seemed unlikely. Furthermore, the Wairepo Creek sub-catchment has the largest area of ‘Dairy support, no irrigation’ (see Figure 2-3 and Table A-1 in Appendix A), and the relatively low concentrations measured there helped to justify the downward adjustments.

° Exotic and Native forest, Tussock, Scrub and Other (primarily mountainous) leaching rates were increased to improve the fit at sites such as the upper Ahuriri monitoring site, which has a large proportion of these land uses upstream.

° Urban leaching rates were also increased to help with the calibration.

For the Lower Waitaki the leaching rates for Exotic and Native forest, Other and Urban were adjusted so that they aligned with the values used for the Upper Waitaki (see Table 2-12).

Waitaki Water Quality Catchment Modelling 39

Table 2-11: Adjusted TN and TP leaching rates for the Upper Waitaki. Values in the brackets are the original rates supplied by ECAN.

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on the flat, Dairy support, Light UW800 1.5 (3.0) 0.03 (0.06) Light UW800 1.5 (26.4) 0.06 (0.60) no irrigation no irrigation

Sheep & beef on the flat, 0.006 Dairy support, Light UW650 2.1 (4.3) Light UW650 2.1 (23.7) 0.013 (0.50) no irrigation (0.013) no irrigation

Sheep & beef on the flat, 0.006 Dairy support, Light UW475 2.1 (4.2) Light UW475 2.1 (18.4) 0.012 (0.50) no irrigation (0.012) no irrigation

Sheep & beef on the flat, Medium- Dairy support, UW800 3.8 (7.6) 0.03 (0.06) Medium-deep UW800 3.8 (26.4) 0.06 (0.50) no irrigation deep no irrigation

Sheep & beef on the flat, Medium- Dairy support, UW650 2.6 (5.1) 0.01 (0.03) Medium-deep UW650 2.6 (23.7) 0.03 (0.50) no irrigation deep no irrigation

Sheep & beef on the flat, Medium- Dairy support, UW475 2.8 (5.5) 0.01 (0.02) Medium-deep UW475 2.8 (18.4) 0.02 (0.40) no irrigation deep no irrigation

Sheep & beef on the flat, Poorly Dairy support, UW800 1.7 (3.4) 0.05 (0.10) Poorly drained UW800 1.7 (26.4) 0.10 (0.80) no irrigation drained no irrigation

Sheep & beef on the flat, Poorly Dairy support, UW650 1.8 (3.6) 0.02 (0.03) Poorly drained UW650 1.8 (23.7) 0.03 (0.70) no irrigation drained no irrigation

Sheep & beef on the flat, Poorly Dairy support, UW475 1.8 (3.5) 0.02 (0.03) Poorly drained UW475 1.8 (18.4) 0.03 (0.50) no irrigation drained no irrigation

Sheep & beef on the flat, Light, medium- UW800, no irrigation, developed Light UW800 1.5 (3.0) 0.03 (0.07) Exotic forest deep, poorly UW650, 2.0 (1.0) 0.05 land drained UW475

40 Waitaki Water Quality Catchment Modelling

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on the flat, Light, medium- UW800, no irrigation, developed Light UW650 4.6 (9.2) 0.02 (0.03) Other deep, poorly UW650, 2.0 (0.0) 0.01 (0.00) land drained UW475

Sheep & beef on the flat, Light, medium- UW800, no irrigation, developed Light UW475 3.9 (7.9) 0.01 (0.02) Tussock deep, poorly UW650, 1.0 (0.5) 0.01 land drained UW475

Sheep & beef on the flat, Light, medium- UW800, Medium- no irrigation, developed UW800 12.1 (24.1) 0.03 (0.06) Native forest deep, poorly UW650, 1.5 (0.5) 0.01 deep land drained UW475

Sheep & beef on the flat, Light, medium- UW800, Medium- no irrigation, developed UW650 6.7 (13.3) 0.02 (0.03) Scrub deep, poorly UW650, 1.5 (0.5) 0.01 deep land drained UW475

Sheep & beef on the flat, Light, medium- UW800, Medium- no irrigation, developed UW475 4.5 (9.1) 0.01 (0.02) Urban deep, poorly UW650, 4.0 (0.0) 0.40 (0.00) deep land drained UW475

Sheep & beef on the flat, Poorly no irrigation, developed UW800 2.5 (5.0) 0.05 (0.10) drained land

Sheep & beef on the flat, Poorly no irrigation, developed UW650 1.5 (3.0) 0.02 (0.03) drained land

Sheep & beef on the flat, Poorly no irrigation, developed UW475 1.5 (3.0) 0.02 (0.03) drained land

Waitaki Water Quality Catchment Modelling 41

Table 2-12: Adjusted TN and TP leaching rates for the Lower Waitaki. Values in the bracket are the original rates supplied by ECAN.

Land use Soil Climate TN yield (kg/ha/y)

Exotic forest All, Blank LW800, LW650, LW500 2.0 (1.0)

Native forest All, Blank LW800, LW650, LW500 1.5 (0.5)

Other All, Blank LW800, LW650, LW500 2.0 (0.5)

Urban All, Blank LW800, LW650, LW500 4.0 (1.5)

42 Waitaki Water Quality Catchment Modelling

3 Results

The measured and modelled loads and yields for TN and TP for the Upper Waitaki are given in Table 3-1 and Table 3-2, while the measured and modelled TN loads and yields for Lower Waitaki are in Table 3-3. Table 3-4 shows the measured loads and yields for TP for the Lower Waitaki. There is a wide range in the percentage load errors 5: -50 to 155 for TN Upper Waitaki, 6 to 280 for TN Lower Waitaki, and 1 to 552 for TP Upper Waitaki.

As can be seen from Table 3-1, Table 3-3 and Table B-1, where there is land use involving dairy or irrigation, each of which with its associated higher leaching rates (see Appendix C), the cumulative yields 6 (kg/ha/y) of TN increase. For example, Willowburn Quailburn Road Bridge and Waikakahi Stream Old Ferry Road. This is not the case with TP cumulative yield, see Table 3-2, Table 3-4 and Table B-1. 3.1 Model fit and uncertainty

Figure 3-1 and Figure 3-2 display the fit between measured and modelled loads for the Upper and Lower Waitaki respectively. The RMSE was calculated from modelled and measured loads from the Upper and Lower Waitaki separately using Equation 2-1 (see Section 2.1.2).

The RMSEs for TN and TP in Upper Waitaki were 34.35 t/y and 2.00 t/y respectively. In Lower Waitaki, the RMSE TN was 10.38 t/y.

A possible source of uncertainty is that the CLUES model assumes groundwater lags are zero (i.e., that stream concentrations reflect current land use) and the calibration adjusted key coefficients (e.g., TN and TP yields and stream attenuation coefficients) to match current observations. If there are significant groundwater lags in the region, then the CLUES’ results are likely to under-predict stream loads – the effects of the recent growth in dairying will not yet be fully shown in the loads.

Wairepo Creek Arm Inlet for TP is an outlier in Figure 3-1. Not only does its measured load have the lowest reliability score of 4, but 31% of its concentration data are censored unlike the other reliability 4 sites which have no censored data (Table 3-2).

TN loads from pasture with low development in the upper catchment are still over-estimated at some sites (e.g., Grays River – Lower Above Ford, Maryburn SH8 Bridge, Tekapo at Forks (just TN)).

Uncertainty in the modelled loads arises from a number of sources including limited data availability for nitrogen and phosphorus concentrations, both in terms of the number of data points and the percentage of readings that were censored, albeit we used techniques to try and use these data. Also there is uncertainty around the flows connected to the hydro schemes. Therefore it is considered wisest to focus on modelled percentage changes when comparing the results from various scenarios.

5 Equals (Modelled – Measured)/Measured. 6 Total in-stream yield for the reach on which the water quality site is located. This represents the total yield for the reach and its upstream tributaries.

Waitaki Water Quality Catchment Modelling 43

Table 3-1: TN measured and modelled loads and yields at the calibration sites in the Upper Waitaki catchment.

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured Modelled cumulative

ID of measured load (t/y) load (t/y) load error cumulative yield yield (kg/ha/y) loads 1 (kg/ha/y)

Ahuiri River Headwater (upper M1 4 35.92 31.82 -11 1.28 1.14 Ahuriri)

Ahuriri River mouth M3 3 127.03 167.64 32 0.93 1.22

Grays River – Lower Above SQ35117 3 13.05 32.01 145 0.24 0.58 Ford

Henburn Rd SQ10824 3 8.36 6.66 -20 1.18 0.94

Maryburn SH8 Bridge SQ10275 2 5.20 11.89 129 0.45 1.03

Maryburn Stream – Lower u/s SQ35115 1 6.83 17.42 155 0.17 0.43 Tekapo River

Ohau C M10 1 445.74 372.54 -16 1.15 0.96

Omarama Stream Omarama SQ10005 3 25.72 39.90 55 0.85 1.32 (SH8)

Quailburn Road Recorder SQ35792 2 10.50 5.25 -50 1.22 0.61

Sutherlands Creek Ben Omar SQ10037 4 2.46 4.65 89 0.85 1.61 Road

Tekapo at Forks 2 M11 3 17.76 23.65 33 0.84 1.11

Tekapo downstream Greys M4 3 40.53 78.41 93 0.32 0.61 Hill 2

44 Waitaki Water Quality Catchment Modelling

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured Modelled cumulative

ID of measured load (t/y) load (t/y) load error cumulative yield yield (kg/ha/y) loads 1 (kg/ha/y)

Tekapo Pukaki Mouth 2 M12 4 174.49 98.93 -43 1.07 0.61

Waitaki at Kurow 71104 1 704.97 757.27 7 0.73 0.78

Willowburn Quailburn Road SQ10012 3 26.63 24.93 -6 2.19 2.05 Bridge 1 See Section 2.3. 2 Flows for these sites uncertain because of the influence of the Tekapo Canal, so loads and yields also uncertain.

Table 3-2: TP measured and modelled loads and yields at the calibration sites in the Upper Waitaki catchment.

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured cumulative Modelled cumulative

ID of measured load (t/y) load (t/y) load error yield (kg/ha/y) yield (kg/ha/y) loads 1

Ahuiri River Headwater M1 4 12.17 13.49 11 0.43 0.48 (upper Ahuriri)

Ahuriri River mouth M3 3 22.24 26.80 21 0.16 0.20

Grays River – Lower SQ35117 3 2.65 4.22 59 0.05 0.08 Above Ford

Henburn Rd SQ10824 3 0.79 0.51 35 0.12 0.07

Maryburn SH8 Bridge SQ10275 1 0.48 0.65 35 0.04 0.06

Maryburn Stream – Lower SQ35115 3 1.34 2.18 63 0.03 0.05 u/s Tekapo River

Waitaki Water Quality Catchment Modelling 45

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured cumulative Modelled cumulative

ID of measured load (t/y) load (t/y) load error yield (kg/ha/y) yield (kg/ha/y) loads 1

Ohau C M10 1 50.62 53.27 6 0.13 0.14

Omarama Stream SQ10005 1 2.49 2.62 5 0.08 0.09 Omarama (SH8)

Quailburn Road Recorder SQ35792 2 0.75 0.74 1 0.09 0.09

Sutherlands Creek Ben SQ10037 4 0.35 0.34 3 0.12 0.12 Omar Road

Tekapo at Forks 2 M11 4 5.28 3.35 37 0.25 0.16

Tekapo downstream M4 3 6.20 10.00 61 0.05 0.08 Greys Hill 2

Tekapo Pukaki Mouth 2 M12 4 11.61 12.66 45 0.07 0.08

Waitaki at Kurow 71104 1 70.03 66.19 5 0.07 0.07

Willowburn Quailburn SQ10012 3 0.90 1.34 49 0.07 0.11 Road Bridge

Mary Burn Mary Burn Fill SQ26372 3 0.12 0.13 8 0.02 0.02

Wairepo Creek Arm Inlet SQ10804 4 0.27 1.76 552 0.02 0.11 1 See Section 2.3. 2 Flows for these sites uncertain because of the influence of the Tekapo Canal, so loads and yields also uncertain.

46 Waitaki Water Quality Catchment Modelling

Table 3-3: TN measured and modelled loads and yields at the calibration sites in the Lower Waitaki catchment.

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured cumulative Modelled cumulative

ID of measured load (t/y) load (t/y) load error yield (kg/ha/y) yield (kg/ha/y) loads 1

Cattle Creek Morland SQ10814 3 9.35 5.90 37 3.92 2.48 Settlement Road

Rocky Point Stream SQ10816 4 1.20 3.87 223 0.58 1.88 Hakataramea Valley Road

Deadman Stream SQ10818 4 1.52 1.01 34 2.53 1.68 Hakataramea Valley Road

Penticotico Stream SH83 SQ10174 4 9.24 3.85 58 1.81 0.75

Otiake River Mt Bell SQ10167 4 2.25 8.55 280 0.46 1.75 Station

Maerewhenua River SQ10160 3 24.33 48.47 99 0.86 1.72 Duntroon

Waikakahi Stream Cock SQ21160 4 21.06 9.75 54 4.83 2.24 and Hen Road

Hakataramea at Above 71103 2 133.90 151.80 13 1.49 1.69 MHBr

Waikakahi Stream Old SQ21255 4 0.34 0.39 15 4.03 4.58 Ferry Road

Otekaieke River D/S of SQ35871 3 19.99 14.21 29 2.25 1.60 Bushy Creek

Waitaki Water Quality Catchment Modelling 47

WQ site name WQ site Reliability Measured mean Modelled mean Percentage Measured cumulative Modelled cumulative

ID of measured load (t/y) load (t/y) load error yield (kg/ha/y) yield (kg/ha/y) loads 1

Waikakahi Stream Glenavy SQ21162 2 45.99 57.30 25 3.37 4.20 Tawai Rd

Waikakahi Stream Te SQ21254 1 60.72 57.30 6 4.45 4.20 Maiharoa Road 1 See Section 2.3.

48 Waitaki Water Quality Catchment Modelling

Table 3-4: TP measured loads and yields at the calibration sites in the Lower Waitaki catchment (see Figure 2-4).

WQ site name WQ site ID Reliability Measured mean Measured cumulative

of measured load (t/y) yield (kg/ha/y) loads 1

Rocky Point Stream SQ10816 4 0.20 0.10 Hakataramea Valley Road

Deadman Stream SQ10818 4 0.13 0.21 Hakataramea Valley Road

Kirkliston Stream SQ10819 4 0.04 0.04 Hakataramea Valley Road

Penticotico Stream SH83 SQ10174 4 0.10 0.02

Waikakahi Stream Cock & SQ21160 4 0.67 0.15 Hen Road

Hakataramea at Above 71103 2 4.23 0.05 MHBr

Waikakahi Stream Old SQ21255 4 0.01 0.11 Ferry Road

Otekaieke River D/S of SQ35871 1 1.69 0.19 Bushy Creek

Waikakahi Stream Glenavy SQ21162 1 0.96 0.07 Tawai Rd

Waikakahi Stream Te SQ21254 1 3.73 0.27 Maiharoa Road 1 See Section 2.3.

Waitaki Water Quality Catchment Modelling 49

1000

100

10 Modelled TNload (t/y) 1 1 10 100 1000

100

10

Wairepo Creek Arm Inlet 1 0.1 1.0 10.0 100.0 Modelled TPload (t/y)

0.1 Measured load (t/y)

Figure 3-1: Measured vs modelled TN and TP loads for the calibration sites in the Upper Waitaki. Dotted red line is 1-1 line. Green dots represent reliability 1 (the best), blue dots reliability 2, orange dots reliability 3, and black dots reliability 4.

50 Waitaki Water Quality Catchment Modelling

1000

100

10

1

Modelled TNload (t/y) 0.1 1.0 10.0 100.0 1000.0 0.1 Measured TN Load (t/y)

Figure 3-2: Measured vs modelled TN loads for the calibration sites in the Lower Waitaki. Dotted red line is 1-1 line. Green dots represent reliability 1 (the best), blue dots reliability 2, orange dots reliability 3, and black dots reliability 4.

Waitaki Water Quality Catchment Modelling 51

4 Summary and conclusions This report has documented the development and calibration of a customised version of the CLUES model specifically for the Waitaki River Catchment.

In the catchment, the impact of land use intensification in terms of dairying and/or irrigation can be seen in the modelling (and measured) cumulative yields of TN in particular. This intensification does not appear to affect the TP yields in a similar way.

The customised CLUES used here to model the loads and yields has been refined from the default CLUES, i.e., the REC unit is divided into unique combinations of land use class, soil and rainfall not just land use independent of soil and rainfall as in the default version.

Furthermore these combinations each have their own sets of source yields meaning that the customised version has many more classes than the default. The refinement enables the customised CLUES to better represent the Waitaki catchment in terms of land use/soil/rainfall and with an eye to improving the model’s capabilities upon recalibration. Moreover, the customised version of CLUES allows users to access the leaching rate tables and therefore adjust the leaching rates as desired.

Because of the greater level of sophistication in representing land use and the ability to access and alter the leaching rate tables, the customised CLUES is better suited to ECan’s purposes. Also the flows as calculated by the customised version were a much better fit to the measured flows than those calculated by the default version (see Section 2.2.1, Flows).

Even though the customised version of CLUES is more flexible in its usability and better represents catchment characteristics than the default version, the fits to the measured loads covered a wide range of percentage load errors: 50 to 155 for TN Upper Waitaki, 6 to 280 for TN Lower Waitaki, and 1 to 552 for TP Upper Waitaki. The RMSE equalled 34.35 t/y for TN Upper Waitaki, 10.38 t/y for TN Lower Waitaki, and 2.00 t/y for TP Upper Waitaki.

Uncertainty in the modelled loads arises from a number of sources including limited data availability for nitrogen and phosphorus concentrations, both in terms of the number of sites for which water quality is recorded and the percentage of readings from these sites that were censored. With respect to the latter, techniques were used to use these concentration data. There is also uncertainty around the flows connected to the hydro schemes that were used to calculate loads. Therefore it is considered wisest to focus on modelled percentage changes when comparing the results from various scenarios.

52 Waitaki Water Quality Catchment Modelling

5 References Alexander, R.B., Elliott, A.H., Shankar, U., McBride, G.B. (2002) Estimating the sources and transport of nutrients in the Waikato River Basin, New Zealand. Water Resources Research , 38(12): 1–23.

Chai, T., Draxler, R.R. (2014) Root mean square error (RMSE) or mean absolute error (MAE)? – Arguments against avoiding RMSE in the literature. Geosci. Model Dev. , 7: 1247–1250.

Close, M.E., Pang, L., Magesan, G.N., Lee, R., Green, S.R. (2003) Field study of pesticide leaching in an allophanic soil in New Zealand. 2: Comparison of simulations from four leaching models. Australian Journal of Soil Research , 41(5): 825–846. Doi 10.1071/Sr02081

Elliott, S., McBride, G., Shankar, U., Semadeni-Davies, A., Quinn, J., Wheeler, D., Wedderburn, L., Small, B., Hewitt, A., Gibb, R., Parfitt, R., Clothier, B., Green, S., Harris, S., Rys, G. (2008) CLUES Spatial DSS: From Farm-Scale Leaching Models to Regional Decision Support, iEMSs 2008: International Congress on Environmental Modelling and Software , Barcelona, Spain, 7 July–10 July.

Elliott, A.H., Shankar, U., Hicks, D.M., Woods, R.A., Dymond, J.R. (2008) SPARROW Regional Regression for Sediment Yields in New Zealand Rivers. Sediment Dynamics in Changing Environments, Christchurch, New Zealand, December 2008.

Elliott, A.H., Alexander, R.B., Schwarz, G.E., Shankar, U., Sukias, J.P.S., McBride, G.B. (2005) Estimation of nutrient sources and transport for New Zealand using the hybrid mechanistic-statistical model SPARROW. Journal of Hydrology (New Zealand) , 44(1): 1– 27.

Fish & Game New Zealand. (1997) Crown Pastoral Land Tenure Review Omaha Downs Pt 092. Fish & Game Report.

Gabites, S., Horrell, G. (2005) Seven day mean annual flow mapping of the tributaries of the Waitaki River. Report prepared for Environment Canterbury.

GHD. (2009) Cumulative Water Quality Effects of Nutrients from Agricultural Intensification in the Upper Waitaki Catchment. Prepared for Russell McVeagh on behalf of Mackenzie Water Research Limited.

Helsel, D.R. (2012) Statistics for Censored Environmental Data Using Minitab and R. Second Edition. John Wiley & Sons.

Hastie, T., Tibshirani, R. (1990) Generalized additive models. First Edition. Chapman and Hall.

Long, D., Chesterton, J. (2005) Waitaki Catchment Hydrological Information. Prepared for Ministry for the Environment by Tonkin & Taylor Limited.

Newsome, P.F.J., Wilde, R.H., Willoughby, E.J. (2008) Land Resource Information System spatial data layers: Data Dictionary, Landcare Research New Zealand Ltd.

Waitaki Water Quality Catchment Modelling 53

Rosen, M.R., Reeves, R.R., Green, S., Clothier, B., Ironside, N. (2004) Prediction of groundwater nitrate contamination after closure of an unlined sheep feedlot. Vadose Zone Journal , 3(3): 990–1006.

Sarmah, A.K., Close, M.E., Pang, L.P., Lee, R., Green, S.R. (2005) Field study of pesticide leaching in a Himatangi sand (Manawatu) and a Kiripaka bouldery clay loam (Northland). 2. Simulation using LEACHM, HYDRUS-1D, GLEAMS, and SPASMO models. Australian Journal of Soil Research , 43(4): 471–489. Doi 10.1071/Sr04040

Semadeni-Davies, A., Elliott, S., Shankar, U. (2011) The CLUES Project: Tutorial manual for CLUES 3.0. Prepared for Ministry of Agriculture and Forestry, NIWA Client Report: HAM2011-003.

Smith, R.A., Schwarz, G.E., Alexander, R.B. (1997) Regional interpretation of water-quality monitoring data. Water Resources Research , 33(12): 2781–2798.

Snelder, T., Biggs, B., Weatherhead, M. (2010) New Zealand River Environment Classification User Guide. March 2004 (Updated June 2010), ME Number 499.

Wheeler, D.M., Ledgard, S.F., Monaghan, R.M., McDowell, R., DeKlein, C.A.M. (2006) OVERSEER® nutrient budget model – what it is, what it does. Implementing sustainable nutrient management strategies in agriculture. Occasional Report No. 19, Fertiliser and Lime Research Centre, Massey University, Palmerston North, New Zealand

Woods, R., Elliott, S., Shankar, U., Bidwell, V., Harris, S., Wheeler, D., Clothier, B., Green, S., Hewitt, A., Gibb, R., Parfitt, R. (2006) The CLUES Project: Predicting the Effects of Land- use on Water Quality – Stage II. NIWA Client Report HAM2006-096.

54 Waitaki Water Quality Catchment Modelling

Appendix A Waitaki sub-catchments

Table A-1: Figure 1-1 key for sub-catchments.

Label Name Label Name Label Name Label Name Label Name Label Name

1 Birch Hill Stream 2 Hopkins River 3 Freds Stream 4 Washdyke Stream 5 Landslip Creek 6 Boundary Stream

7 Bush Stream 8 Irishman Creek 9 Maryburn 10 Twin Stream 11 Edwards 12 Lake Pukaki Stream

13 Whale Stream 14 Camp Stream 15 Huxley River 16 Jacks Stream 17 Tekapo River 18 Sawdon Stream

19 20 Snowy Gorge 21 Ohau River 22 Sawyers Creek 23 Freehold 24 Stony River Creek Creek

25 East Branch Ahuriri 26 Wairepo Creek 27 Dalgety 28 Maori Creek 29 Six Mile Creek 30 Serpentine River Stream Creek

31 32 Grampian 33 Quailburn 34 Birch Creek 35 Dunstan 36 Scrubby Creek Stream Stream

37 East Diadem Creek 38 Falstone Creek 39 Henburn 40 West Diadem Creek 41 Sutherlands 42 Two Mile Creek Stream

43 Avonburn 44 Ribbonwood 45 Totara Creek 46 Silver Creek 47 Peters Stream 48 Shepards Creek Creek

49 Gorman Stream 50 Coal Creek 51 Deep Stream 52 Cattle Creek 53 Lion Creek 54 Lake Aviemore

55 Hemphills Creek 56 Stony Stream 57 Rocky Point 58 McKays Stream 59 Omarama 60 Longslip Creek Stream Stream

61 Old Man Creek 62 Cattle Creek 63 Glen Creek 64 Manuka Creek 65 Station Stream 66 Stony Creek

Waitaki Water Quality Catchment Modelling 55

Label Name Label Name Label Name Label Name Label Name Label Name

67 Clarks Creek 68 Otamatapaio 69 McLays Creek 70 Potato Creek 71 Bluestone 72 Farm Stream River Creek

73 Little Omarama 74 Padkins Creek 75 Station Creek 76 Macauley River 77 78 Murchison Stream River

79 Hooker River 80 Cass River 81 Parsons Rock 82 Otematata River 83 Sisters Creek 84 Coal Stream Creek

85 Rugged Ridges 86 87 Banks Stream 88 McRae Stream 89 Dry Stream 90 Brothers Creek Creek

91 Fern Gully Creek 92 Corbies Creek 93 Wharekuri 94 Awahokomo Creek 95 Waitaki River 96 Clear Stream Creek

97 Little Awakino 98 Awakino River 99 Camp Creek 100 Penticotico Stream 101 Cattle Gully 102 Shaw Creek River

103 Chimney Creek 104 Diggers Gully 105 Kurow River 106 Millers Creek 107 Yeomans 108 Elephant Hill Creek Creek Stream

109 Malcolms Creek 110 Otiake River 111 Waikakahi 112 Doctors Creek 113 Waihuna 114 Otekaieke Stream Stream River

115 Tewatapoki 116 Rambling Gorge 117 Waikaura 118 Maerewhenua River 119 Lone Creek 120 Whitneys Stream Creek Creek

121 Waipati Creek 122 Guffies Creek 123 Boundary 124 Green Gully 125 Home Creek 126 Browns Creek Creek

127 Waimihi Creek 128 Blue Duck Creek 129 Jacksons Creek 130 Sheepwash 131 Robertsons 132 Pringles Creek Creek/Bushy Creek Gully

133 Danseys Creek 134 Godley River 135 Pukaki River 136 Boundary Stream 137 Grays River 138 Temple Creek

56 Waitaki Water Quality Catchment Modelling

Label Name Label Name Label Name Label Name Label Name Label Name

139 Lake Ohau 140 Ahuriri River 141 Harris Creek 142 Fraser River 143 Maitland 144 Snow River Creek

145 Parsons Creek 146 Jollie River 147 Lake Tekapo 148 Mistake River 149 Coal River 150 Black Birch Stream

151 Mt Gerald Creek 152 Fork Stream 153 Sawyers 154 Hoophorn Stream 155 Dobson River Stream

Waitaki Water Quality Catchment Modelling 57

Appendix B Percentage of land use by area for each of the watersheds of the calibrated sites

Table B-1: Percentage of land use by area for each of the watersheds of the calibrated water quality monitoring sites. 0 <0a <0.5%

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Ahuiri River Headwater (upper 0 0 0 0 0 0 0a 8 32 0 1 0 1 0 0 4 54 0a Ahuriri), M1, C27

Quailburn Road 0 0 0 0 9 0 5 0a 3 0 35 0a 25 0 0 2 21 0 Recorder, SQ35792, C20

Henburn Rd, 0a 0 0 0 0 0 0a 0a 0 0 51 1 44 0 0 0a 4 0 SQ10824, C12

Willowburn Quailburn 0 0 3 0 36 0a 1 1 2 0 26 6 15 0 0 0a 10 0 Road Bridge, SQ10012, C2

58 Waitaki Water Quality Catchment Modelling

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Omarama Stream Omarama 0a 0 0a 0 7 2 0a 0a 2 0a 16 3 54 0 0 1 15 0a (SH8), SQ10005, C1

Ahuriri River mouth, M3, 0a 0 1 0 7 1 1 2 9 0a 18 2 32 0 0 2 27 0a C22

Sutherlands Creek Ben 0 0 5 0 0 0 0a 0 4 0 28 5 56 0 0 2 0a 0 Omar Road, SQ10037, C3

Wairepo Creek Arm 5 0 2 0 31 0a 0a 1 1 0a 22 2 18 0 0 0a 17 0 Inlet, SQ10804, C8

Ohau C, M10, 0a 0a 1 0 2 0a 1 3 36 0a 13 0a 24 0 0 3 18 0a C23

Mary Burn Mary Burn Fill, 0 0 0a 0 0a 0 5 0a 0a 0 49 0a 4 0 0 0 41 0 SQ26372, C17

Waitaki Water Quality Catchment Modelling 59

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Maryburn SH8 Bridge, 0 0a 0a 0 0a 0 3 0a 0a 3 65 2 8 0 0 0a 19 0 SQ10275, C7

Maryburn Stream - Lower u/s 0a 0a 0a 0 0a 0 3 0a 1 3 57 2 6 0 0 0a 28 0a Tekapo River, SQ35115, C18

Grays River - Lower Above 0a 0 0 0 0 0 0a 0a 0a 0a 51 0a 48 0 0 0a 0a 0a Ford, SQ35117, C19

Tekapo at Forks, M11, 0a 0 0a 0 0 0 1 0a 40 0 15 0a 22 0 0 2 20 0a C26

Tekapo downstream 0a 0a 0a 0 0a 0 1 0a 24 0a 32 0a 25 0 0 1 17 0a Greys Hill, M4, C25

Tekapo Pukaki Mouth, M12, 0a 0a 0a 0 0a 0 2 0a 32 0a 27 0a 19 0 0 2 17 0a C24

60 Waitaki Water Quality Catchment Modelling

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Waitaki at Kurow, 71104, 0a 0a 0a 0 1 0a 1 2 20 0a 20 1 34 0 0 2 18 0a C29

Cattle Creek Morland Settlement 7 0 0 0 0 0 0a 0a 0a 0 37 0 9 0 0 6 42 0 Road, SQ10814, C9

Rocky Point Stream Hakataramea 7 0 0 0 0 0 0a 0a 0a 0 73 2 13 0 0 3 3 0 Valley Road, SQ10816, C10

Deadman Stream Hakataramea 0 0 0 0 0 0 0a 0a 1 0 5 0 8 85 0 0a 1 0 Valley Road, SQ10818, C11

Hakataramea at Above 3 0 0a 0 1 1 0a 0a 1 0 29 2 46 7 0a 1 8 0 MHBr, 71103, C28

Waitaki Water Quality Catchment Modelling 61

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Otiake River Mt Bell 0 0 0 0a 1 2 0a 0a 8 0a 1 1 62 0 0 1 23 0 Station, SQ10167, C5

Otekaieke River D/S of 0 0 0 0 0 0 0a 0a 5 0 0a 1 86 0a 0 0a 8 0 Bushy Creek, SQ35871, C21

Penticotico Stream SH83, 0 0 1 0 0 0 0a 0 0a 0 4 1 93 0a 0 0a 0a 0 SQ10174, C6

Maerewhenua River 0a 0a 9 0 0a 0 2 0a 1 0a 9 2 55 4 0 1 15 0a Duntroon, SQ10160, C4

Waikakahi Stream Cock 0a 4 27 1 3 0a 1 0a 0 10 32 20 0 3 0 0 0 0 & Hen Road, SQ21160, C13

62 Waitaki Water Quality Catchment Modelling

Percentage of land use by area

Water quality calibration site Name, ID, Site key from Figure 1-1 Arable Dairy_BDyk Dairy_Spra DairySuppB DairySuppD DairySuppS ExoticFor NativeFor Other SBF_BDyke SBF_Dry SBF_Spray SBH_Dry Deer_Dry Deer_Spray Scrub Tussock Urban

Waikakahi Stream Old 0 12 61 9 4 12 0a 0 0 0 1 0a 0 0 0 0 0 0a Ferry Road, SQ21255, C16

Waikakahi Stream Glenavy Tawai 0a 9 25 1 1 0a 1 0a 0a 5 20 10 27 1 0 0a 0a 0a Rd, SQ21162, C15

Waikakahi Stream Te Maiharoa 0a 9 25 1 1 0a 1 0a 0a 5 20 10 27 1 0 0a 0a 0a Road, SQ21254, C14

Waitaki Water Quality Catchment Modelling 63

Appendix C TN and TP leaching rates

Table C-1: TN and TP leaching rates for the Upper Waitaki (ECan, pers. comm.). See Table 2-3 for a description of the soils.

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Dairy, borderdyke Arable Light UW800 19.2 0.50 Light UW800 50.6 2.30 irrigation

Dairy, borderdyke Arable Light UW650 17.6 0.30 Light UW650 45.5 2.20 irrigation

Dairy, borderdyke Arable Light UW475 15.9 0.20 Light UW475 35.4 2.20 irrigation

Dairy, borderdyke Arable Medium-deep UW800 16.4 0.50 Medium-deep UW800 50.6 2.00 irrigation

Dairy, borderdyke Arable Medium-deep UW650 14.4 0.30 Medium-deep UW650 45.5 2.00 irrigation

Dairy, borderdyke Arable Medium-deep UW475 13.0 0.20 Medium-deep UW475 35.4 1.90 irrigation

Dairy, borderdyke Arable Poorly drained UW800 5.5 0.60 Poorly drained UW800 50.6 2.10 irrigation

Dairy, borderdyke Arable Poorly drained UW650 4.8 0.50 Poorly drained UW650 45.5 2.00 irrigation

Dairy, borderdyke Arable Poorly drained UW475 4.3 0.40 Poorly drained UW475 35.4 1.90 irrigation

Dairy, spray Dairy support, Light UW800 38.9 1.00 Light UW800 34.3 1.90 irrigation borderdyke irrigation

64 Waitaki Water Quality Catchment Modelling

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Dairy, spray Dairy support, Light UW650 35.0 0.90 Light UW650 30.8 1.80 irrigation borderdyke irrigation

Dairy, spray Dairy support, Light UW475 27.2 0.90 Light UW475 24.0 1.80 irrigation borderdyke irrigation

Dairy, spray Dairy support, Medium-deep UW800 38.9 0.90 Medium-deep UW800 34.3 1.60 irrigation borderdyke irrigation

Dairy, spray Dairy support, Medium-deep UW650 35.0 0.90 Medium-deep UW650 30.8 1.60 irrigation borderdyke irrigation

Dairy, spray Dairy support, Medium-deep UW475 27.2 0.80 Medium-deep UW475 24.0 1.50 irrigation borderdyke irrigation

Dairy, spray Dairy support, Poorly drained UW800 38.9 1.20 Poorly drained UW800 34.3 1.80 irrigation borderdyke irrigation

Dairy, spray Dairy support, Poorly drained UW650 35.0 1.00 Poorly drained UW650 30.8 1.70 irrigation borderdyke irrigation

Dairy, spray Dairy support, Poorly drained UW475 27.2 0.90 Poorly drained UW475 24.0 1.50 irrigation borderdyke irrigation

Dairy support, spray Dairy support, no Light UW800 26.4 0.60 Light UW800 1.5 (26.4) 0.06 (0.60) irrigation irrigation

Dairy support, spray Dairy support, no Light UW650 23.7 0.50 Light UW650 2.1 (23.7) 0.01 (0.50) irrigation irrigation

Dairy support, spray Dairy support, no Light UW475 18.4 0.50 Light UW475 2.1 (18.4) 0.01 (0.50) irrigation irrigation

Waitaki Water Quality Catchment Modelling 65

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Dairy support, spray Dairy support, no Medium-deep UW800 26.4 0.50 Medium-deep UW800 3.8 (26.4) 0.06 (0.50) irrigation irrigation

Dairy support, spray Dairy support, no Medium-deep UW650 23.7 0.50 Medium-deep UW650 2.6 (23.7) 0.03 (0.50) irrigation irrigation

Dairy support, spray Dairy support, no Medium-deep UW475 18.4 0.40 Medium-deep UW475 2.8 (18.4) 0.02 (0.40) irrigation irrigation

Dairy support, spray Dairy support, no Poorly drained UW800 26.4 0.80 Poorly drained UW800 1.7 (26.4) 0.10 (0.80) irrigation irrigation

Dairy support, spray Dairy support, no Poorly drained UW650 23.7 0.70 Poorly drained UW650 1.8 (23.7) 0.03 (0.70) irrigation irrigation

Dairy support, spray Dairy support, no Poorly drained UW475 18.4 0.50 Poorly drained UW475 1.8 (18.4) 0.03 (0.50) irrigation irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Light UW800 40.0 1.85 Light UW800 1.5 (3.0) 0.03 (0.06) flat, no irrigation irrigation

Sheep & beef on the Sheep & beef on the 0.006 flat, borderdyke Light UW650 36.0 1.23 Light UW650 2.1 (4.3) flat, no irrigation (0.013) irrigation

Sheep & beef on the Sheep & beef on the 0.006 flat, borderdyke Light UW475 28.0 1.23 Light UW475 2.1 (4.2) flat, no irrigation (0.012) irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Medium-deep UW800 36.7 1.60 Medium-deep UW800 3.8 (7.6) 0.03 (0.06) flat, no irrigation irrigation

66 Waitaki Water Quality Catchment Modelling

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on the Sheep & beef on the flat, borderdyke Medium-deep UW650 33.0 1.07 Medium-deep UW650 2.6 (5.1) 0.01 (0.03) flat, no irrigation irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Medium-deep UW475 25.7 1.07 Medium-deep UW475 2.8 (5.5) 0.01 (0.02) flat, no irrigation irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Poorly drained UW800 30.0 2.30 Poorly drained UW800 1.7 (3.4) 0.05 (0.10) flat, no irrigation irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Poorly drained UW650 27.0 1.53 Poorly drained UW650 1.8 (3.6) 0.02 (0.03) flat, no irrigation irrigation

Sheep & beef on the Sheep & beef on the flat, borderdyke Poorly drained UW475 21.0 1.53 Poorly drained UW475 1.8 (3.5) 0.02 (0.03) flat, no irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Light UW800 2.0 0.06 Light UW800 30.0 0.30 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Light UW650 2.0 0.03 Light UW650 27.0 0.30 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Light UW475 2.0 0.02 Light UW475 21.0 0.20 flat, spray irrigation irrigation

Waitaki Water Quality Catchment Modelling 67

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on hill Sheep & beef on the country, no Medium-deep UW800 2.0 0.06 Medium-deep UW800 24.0 0.30 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Medium-deep UW650 2.0 0.03 Medium-deep UW650 24.0 0.20 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Medium-deep UW475 2.0 0.02 Medium-deep UW475 19.0 0.20 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Poorly drained UW800 2.0 0.06 Poorly drained UW800 20.0 0.80 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Poorly drained UW650 2.0 0.03 Poorly drained UW650 19.0 0.70 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no Poorly drained UW475 2.0 0.02 Poorly drained UW475 18.0 0.50 flat, spray irrigation irrigation

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Light UW800 1.5 (3.0) 0.03 (0.07) Light UW800 3.0 0.27 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Light UW650 4.6 (9.2) 0.02 (0.03) Light UW650 3.0 0.13 irrigation, developed developed land land

68 Waitaki Water Quality Catchment Modelling

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Light UW475 3.9 (7.9) 0.01 (0.02) Light UW475 3.0 0.07 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Medium-deep UW800 12.1 (24.1) 0.03 (0.06) Medium-deep UW800 3.0 0.27 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Medium-deep UW650 6.7 (13.3) 0.02 (0.03) Medium-deep UW650 3.0 0.13 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Medium-deep UW475 4.5 (9.1) 0.01 (0.02) Medium-deep UW475 3.0 0.07 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Poorly drained UW800 2.5 (5.0) 0.05 (0.10) Poorly drained UW800 3.0 0.27 irrigation, developed developed land land

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Poorly drained UW650 1.5 (3.0) 0.02 (0.03) Poorly drained UW650 3.0 0.13 irrigation, developed developed land land

Waitaki Water Quality Catchment Modelling 69

TN yield TP yield TN yield TP yield Land use Soil Climate Land use Soil Climate (kg/ha/y) (kg/ha/y) (kg/ha/y) (kg/ha/y)

Sheep & beef on hill Sheep & beef on the country, no flat, no irrigation, Poorly drained UW475 1.5 (3.0) 0.02 (0.03) Poorly drained UW475 3.0 0.07 irrigation, developed developed land land

Light, UW800, Light, UW800, Exotic forest medium-deep, UW650, 2.0 (1.0) 0.05 Native forest medium-deep, UW650, 1.5 (0.5) 0.01 poorly drained UW475 poorly drained UW475

Light, UW800, Light, UW800, Other medium-deep, UW650, 2.0 (0.0) 0.01 (0.00) Scrub medium-deep, UW650, 1.5 (0.5) 0.01 poorly drained UW475 poorly drained UW475

Light, UW800, Light, UW800, Tussock medium-deep, UW650, 1.0 (0.5) 0.01 Urban medium-deep, UW650, 4.0 (0.0) 0.40 (0.00) poorly drained UW475 poorly drained UW475

70 Waitaki Water Quality Catchment Modelling

Table C-2: TN leaching rates for the Lower Waitaki ( ECan, pers. comm .). Values in the brackets were originally ECan’s but were adjusted as a result of calibration and further investigation. LW800, LW650 and LW500 correspond to mm of annual rainfall in the Lower Waitaki (LW). See Table 2-4 for a description of the soils.

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Arable, borderdyke irrigation Blank LW800 1.5 Arable, no irrigation Blank LW800 1.5

Arable, borderdyke irrigation Blank LW650 1.5 Arable, no irrigation Blank LW650 1.5

Arable, borderdyke irrigation Blank LW500 1.5 Arable, no irrigation Blank LW500 1.5

Arable, borderdyke irrigation Class 4 to 6 LW800 9.0 Arable, no irrigation Class 4 to 6 LW800 9.0

Arable, borderdyke irrigation Class 4 to 6 LW650 6.0 Arable, no irrigation Class 4 to 6 LW650 6.0

Arable, borderdyke irrigation Class 4 to 6 LW500 4.0 Arable, no irrigation Class 4 to 6 LW500 4.0

Arable, borderdyke irrigation Class 6 LW800 5.0 Arable, no irrigation Class 6 LW800 5.0

Arable, borderdyke irrigation Class 6 LW650 4.0 Arable, no irrigation Class 6 LW650 4.0

Arable, borderdyke irrigation Class 6 LW500 3.0 Arable, no irrigation Class 6 LW500 3.0

Arable, borderdyke irrigation Class 7 LW800 1.5 Arable, no irrigation Class 7 LW800 1.5

Arable, borderdyke irrigation Class 7 LW650 1.5 Arable, no irrigation Class 7 LW650 1.5

Arable, borderdyke irrigation Class 7 LW500 1.5 Arable, no irrigation Class 7 LW500 1.5

Arable, borderdyke irrigation H LW800 14.2 Arable, no irrigation H LW800 5.5

Arable, borderdyke irrigation H LW650 10.3 Arable, no irrigation H LW650 4.0

Arable, borderdyke irrigation H LW500 10.3 Arable, no irrigation H LW500 4.0

Arable, borderdyke irrigation L LW800 28.1 Arable, no irrigation L LW800 19.3

Waitaki Water Quality Catchment Modelling 71

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Arable, borderdyke irrigation L LW650 24.1 Arable, no irrigation L LW650 18.0

Arable, borderdyke irrigation L LW500 24.1 Arable, no irrigation L LW500 18.0

Arable, borderdyke irrigation M LW800 20.4 Arable, no irrigation M LW800 11.6

Arable, borderdyke irrigation M LW650 17.9 Arable, no irrigation M LW650 10.8

Arable, borderdyke irrigation M LW500 17.9 Arable, no irrigation M LW500 10.8

Arable, borderdyke irrigation Pd LW800 7.1 Arable, no irrigation Pd LW800 2.7

Arable, borderdyke irrigation Pd LW650 5.1 Arable, no irrigation Pd LW650 2.0

Arable, borderdyke irrigation Pd LW500 5.1 Arable, no irrigation Pd LW500 2.0

Arable, borderdyke irrigation PdL LW800 14.0 Arable, no irrigation PdL LW800 9.7

Arable, borderdyke irrigation PdL LW650 12.0 Arable, no irrigation PdL LW650 9.0

Arable, borderdyke irrigation PdL LW500 12.0 Arable, no irrigation PdL LW500 9.0

Arable, borderdyke irrigation VL LW800 25.7 Arable, no irrigation VL LW800 17.4

Arable, borderdyke irrigation VL LW650 21.9 Arable, no irrigation VL LW650 17.3

Arable, borderdyke irrigation VL LW500 21.9 Arable, no irrigation VL LW500 17.3

Arable, borderdyke irrigation XL LW800 30.7 Arable, no irrigation XL LW800 25.8

Arable, borderdyke irrigation XL LW650 29.0 Arable, no irrigation XL LW650 26.0

Arable, borderdyke irrigation XL LW500 29.0 Arable, no irrigation XL LW500 26.0

72 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Arable, spray irrigation Blank LW800 1.5 Dairy, borderdyke irrigation Blank LW800 1.5

Arable, spray irrigation Blank LW650 1.5 Dairy, borderdyke irrigation Blank LW650 1.5

Arable, spray irrigation Blank LW500 1.5 Dairy, borderdyke irrigation Blank LW500 1.5

Arable, spray irrigation Class 4 to 6 LW800 9.0 Dairy, borderdyke irrigation Class 4 to 6 LW800 13.2

Arable, spray irrigation Class 4 to 6 LW650 6.0 Dairy, borderdyke irrigation Class 4 to 6 LW650 9.1

Arable, spray irrigation Class 4 to 6 LW500 4.0 Dairy, borderdyke irrigation Class 4 to 6 LW500 5.7

Arable, spray irrigation Class 6 LW800 5.0 Dairy, borderdyke irrigation Class 6 LW800 5.0

Arable, spray irrigation Class 6 LW650 4.0 Dairy, borderdyke irrigation Class 6 LW650 4.0

Arable, spray irrigation Class 6 LW500 3.0 Dairy, borderdyke irrigation Class 6 LW500 3.0

Arable, spray irrigation Class 7 LW800 1.5 Dairy, borderdyke irrigation Class 7 LW800 1.5

Arable, spray irrigation Class 7 LW650 1.5 Dairy, borderdyke irrigation Class 7 LW650 1.5

Arable, spray irrigation Class 7 LW500 1.5 Dairy, borderdyke irrigation Class 7 LW500 1.5

Arable, spray irrigation H LW800 14.2 Dairy, borderdyke irrigation H LW800 38.1

Arable, spray irrigation H LW650 10.3 Dairy, borderdyke irrigation H LW650 34.5

Arable, spray irrigation H LW500 10.3 Dairy, borderdyke irrigation H LW500 27.1

Arable, spray irrigation L LW800 28.1 Dairy, borderdyke irrigation L LW800 50.6

Arable, spray irrigation L LW650 24.1 Dairy, borderdyke irrigation L LW650 45.5

Waitaki Water Quality Catchment Modelling 73

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Arable, spray irrigation L LW500 24.1 Dairy, borderdyke irrigation L LW500 35.4

Arable, spray irrigation M LW800 20.4 Dairy, borderdyke irrigation M LW800 44.1

Arable, spray irrigation M LW650 17.9 Dairy, borderdyke irrigation M LW650 40.0

Arable, spray irrigation M LW500 17.9 Dairy, borderdyke irrigation M LW500 31.4

Arable, spray irrigation Pd LW800 7.1 Dairy, borderdyke irrigation Pd LW800 19.1

Arable, spray irrigation Pd LW650 5.1 Dairy, borderdyke irrigation Pd LW650 17.3

Arable, spray irrigation Pd LW500 5.1 Dairy, borderdyke irrigation Pd LW500 13.5

Arable, spray irrigation PdL LW800 14.0 Dairy, borderdyke irrigation PdL LW800 25.3

Arable, spray irrigation PdL LW650 12.0 Dairy, borderdyke irrigation PdL LW650 22.8

Arable, spray irrigation PdL LW500 12.0 Dairy, borderdyke irrigation PdL LW500 17.7

Arable, spray irrigation VL LW800 25.7 Dairy, borderdyke irrigation VL LW800 81.9

Arable, spray irrigation VL LW650 21.9 Dairy, borderdyke irrigation VL LW650 69.8

Arable, spray irrigation VL LW500 21.9 Dairy, borderdyke irrigation VL LW500 51.4

Arable, spray irrigation XL LW800 30.7 Dairy, borderdyke irrigation XL LW800 148.5

Arable, spray irrigation XL LW650 29.0 Dairy, borderdyke irrigation XL LW650 126.5

Arable, spray irrigation XL LW500 29.0 Dairy, borderdyke irrigation XL LW500 93.1

Dairy, no irrigation Blank LW800 1.5 Dairy, spray irrigation Blank LW800 1.5

74 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy, no irrigation Blank LW650 1.5 Dairy, spray irrigation Blank LW650 1.5

Dairy, no irrigation Blank LW500 1.5 Dairy, spray irrigation Blank LW500 1.5

Dairy, no irrigation Class 4 to 6 LW800 13.2 Dairy, spray irrigation Class 4 to 6 LW800 13.2

Dairy, no irrigation Class 4 to 6 LW650 9.1 Dairy, spray irrigation Class 4 to 6 LW650 9.1

Dairy, no irrigation Class 4 to 6 LW500 5.7 Dairy, spray irrigation Class 4 to 6 LW500 5.7

Dairy, no irrigation Class 6 LW800 5.0 Dairy, spray irrigation Class 6 LW800 5.0

Dairy, no irrigation Class 6 LW650 4.0 Dairy, spray irrigation Class 6 LW650 4.0

Dairy, no irrigation Class 6 LW500 3.0 Dairy, spray irrigation Class 6 LW500 3.0

Dairy, no irrigation Class 7 LW800 1.5 Dairy, spray irrigation Class 7 LW800 1.5

Dairy, no irrigation Class 7 LW650 1.5 Dairy, spray irrigation Class 7 LW650 1.5

Dairy, no irrigation Class 7 LW500 1.5 Dairy, spray irrigation Class 7 LW500 1.5

Dairy, no irrigation H LW800 16.7 Dairy, spray irrigation H LW800 29.1

Dairy, no irrigation H LW650 13.8 Dairy, spray irrigation H LW650 26.7

Dairy, no irrigation H LW500 11.4 Dairy, spray irrigation H LW500 21.0

Dairy, no irrigation L LW800 22.8 Dairy, spray irrigation L LW800 39.0

Dairy, no irrigation L LW650 18.4 Dairy, spray irrigation L LW650 35.0

Dairy, no irrigation L LW500 14.9 Dairy, spray irrigation L LW500 27.0

Waitaki Water Quality Catchment Modelling 75

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy, no irrigation M LW800 19.7 Dairy, spray irrigation M LW800 34.0

Dairy, no irrigation M LW650 16.1 Dairy, spray irrigation M LW650 30.8

Dairy, no irrigation M LW500 13.2 Dairy, spray irrigation M LW500 24.1

Dairy, no irrigation Pd LW800 8.4 Dairy, spray irrigation Pd LW800 14.6

Dairy, no irrigation Pd LW650 6.9 Dairy, spray irrigation Pd LW650 13.3

Dairy, no irrigation Pd LW500 5.7 Dairy, spray irrigation Pd LW500 10.5

Dairy, no irrigation PdL LW800 11.4 Dairy, spray irrigation PdL LW800 19.5

Dairy, no irrigation PdL LW650 9.2 Dairy, spray irrigation PdL LW650 17.5

Dairy, no irrigation PdL LW500 7.4 Dairy, spray irrigation PdL LW500 13.5

Dairy, no irrigation VL LW800 35.2 Dairy, spray irrigation VL LW800 61.8

Dairy, no irrigation VL LW650 26.4 Dairy, spray irrigation VL LW650 51.3

Dairy, no irrigation VL LW500 19.8 Dairy, spray irrigation VL LW500 36.6

Dairy, no irrigation XL LW800 49.4 Dairy, spray irrigation XL LW800 88.0

Dairy, no irrigation XL LW650 35.2 Dairy, spray irrigation XL LW650 69.2

Dairy, no irrigation XL LW500 25.1 Dairy, spray irrigation XL LW500 46.8

Dairy support, borderdyke irrigation Blank LW800 1.5 Dairy support, no irrigation Blank LW800 1.5

Dairy support, borderdyke irrigation Blank LW650 1.5 Dairy support, no irrigation Blank LW650 1.5

76 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy support, borderdyke irrigation Blank LW500 1.5 Dairy support, no irrigation Blank LW500 1.5

Dairy support, borderdyke irrigation Class 4 to 6 LW800 13.2 Dairy support, no irrigation Class 4 to 6 LW800 13.2

Dairy support, borderdyke irrigation Class 4 to 6 LW650 9.1 Dairy support, no irrigation Class 4 to 6 LW650 9.1

Dairy support, borderdyke irrigation Class 4 to 6 LW500 5.7 Dairy support, no irrigation Class 4 to 6 LW500 5.7

Dairy support, borderdyke irrigation Class 6 LW800 5.0 Dairy support, no irrigation Class 6 LW800 5.0

Dairy support, borderdyke irrigation Class 6 LW650 4.0 Dairy support, no irrigation Class 6 LW650 4.0

Dairy support, borderdyke irrigation Class 6 LW500 3.0 Dairy support, no irrigation Class 6 LW500 3.0

Dairy support, borderdyke irrigation Class 7 LW800 1.5 Dairy support, no irrigation Class 7 LW800 1.5

Dairy support, borderdyke irrigation Class 7 LW650 1.5 Dairy support, no irrigation Class 7 LW650 1.5

Dairy support, borderdyke irrigation Class 7 LW500 1.5 Dairy support, no irrigation Class 7 LW500 1.5

Dairy support, borderdyke irrigation H LW800 37.6 Dairy support, no irrigation H LW800 20.2

Dairy support, borderdyke irrigation H LW650 27.4 Dairy support, no irrigation H LW650 12.4

Dairy support, borderdyke irrigation H LW500 20.1 Dairy support, no irrigation H LW500 7.6

Dairy support, borderdyke irrigation L LW800 49.8 Dairy support, no irrigation L LW800 29.6

Dairy support, borderdyke irrigation L LW650 36.1 Dairy support, no irrigation L LW650 18.5

Dairy support, borderdyke irrigation L LW500 26.3 Dairy support, no irrigation L LW500 11.7

Dairy support, borderdyke irrigation M LW800 43.4 Dairy support, no irrigation M LW800 24.7

Waitaki Water Quality Catchment Modelling 77

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy support, borderdyke irrigation M LW650 31.8 Dairy support, no irrigation M LW650 15.3

Dairy support, borderdyke irrigation M LW500 23.3 Dairy support, no irrigation M LW500 9.5

Dairy support, borderdyke irrigation Pd LW800 18.8 Dairy support, no irrigation Pd LW800 10.1

Dairy support, borderdyke irrigation Pd LW650 13.7 Dairy support, no irrigation Pd LW650 6.2

Dairy support, borderdyke irrigation Pd LW500 10.0 Dairy support, no irrigation Pd LW500 3.8

Dairy support, borderdyke irrigation PdL LW800 24.9 Dairy support, no irrigation PdL LW800 14.8

Dairy support, borderdyke irrigation PdL LW650 18.1 Dairy support, no irrigation PdL LW650 9.3

Dairy support, borderdyke irrigation PdL LW500 13.1 Dairy support, no irrigation PdL LW500 5.8

Dairy support, borderdyke irrigation VL LW800 80.7 Dairy support, no irrigation VL LW800 49.1

Dairy support, borderdyke irrigation VL LW650 55.4 Dairy support, no irrigation VL LW650 29.4

Dairy support, borderdyke irrigation VL LW500 38.2 Dairy support, no irrigation VL LW500 17.6

Dairy support, borderdyke irrigation XL LW800 146.3 Dairy support, no irrigation XL LW800 72.8

Dairy support, borderdyke irrigation XL LW650 100.4 Dairy support, no irrigation XL LW650 42.4

Dairy support, borderdyke irrigation XL LW500 69.1 Dairy support, no irrigation XL LW500 24.7

Dairy support, spray irrigation Blank LW800 1.5 Deer, borderdyke irrigation Blank LW800 1.5

Dairy support, spray irrigation Blank LW650 1.5 Deer, borderdyke irrigation Blank LW650 1.5

Dairy support, spray irrigation Blank LW500 1.5 Deer, borderdyke irrigation Blank LW500 1.5

78 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy support, spray irrigation Class 4 to 6 LW800 13.2 Deer, borderdyke irrigation Class 4 to 6 LW800 9.0

Dairy support, spray irrigation Class 4 to 6 LW650 9.1 Deer, borderdyke irrigation Class 4 to 6 LW650 6.0

Dairy support, spray irrigation Class 4 to 6 LW500 5.7 Deer, borderdyke irrigation Class 4 to 6 LW500 4.0

Dairy support, spray irrigation Class 6 LW800 5.0 Deer, borderdyke irrigation Class 6 LW800 5.0

Dairy support, spray irrigation Class 6 LW650 4.0 Deer, borderdyke irrigation Class 6 LW650 4.0

Dairy support, spray irrigation Class 6 LW500 3.0 Deer, borderdyke irrigation Class 6 LW500 3.0

Dairy support, spray irrigation Class 7 LW800 1.5 Deer, borderdyke irrigation Class 7 LW800 1.5

Dairy support, spray irrigation Class 7 LW650 1.5 Deer, borderdyke irrigation Class 7 LW650 1.5

Dairy support, spray irrigation Class 7 LW500 1.5 Deer, borderdyke irrigation Class 7 LW500 1.5

Dairy support, spray irrigation H LW800 29.4 Deer, borderdyke irrigation H LW800 15.0

Dairy support, spray irrigation H LW650 21.8 Deer, borderdyke irrigation H LW650 13.5

Dairy support, spray irrigation H LW500 16.2 Deer, borderdyke irrigation H LW500 11.4

Dairy support, spray irrigation L LW800 39.4 Deer, borderdyke irrigation L LW800 25.0

Dairy support, spray irrigation L LW650 28.6 Deer, borderdyke irrigation L LW650 22.0

Dairy support, spray irrigation L LW500 20.8 Deer, borderdyke irrigation L LW500 18.0

Dairy support, spray irrigation M LW800 34.4 Deer, borderdyke irrigation M LW800 19.5

Dairy support, spray irrigation M LW650 25.2 Deer, borderdyke irrigation M LW650 17.5

Waitaki Water Quality Catchment Modelling 79

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Dairy support, spray irrigation M LW500 18.5 Deer, borderdyke irrigation M LW500 14.6

Dairy support, spray irrigation Pd LW800 14.7 Deer, borderdyke irrigation Pd LW800 3.8

Dairy support, spray irrigation Pd LW650 10.9 Deer, borderdyke irrigation Pd LW650 3.4

Dairy support, spray irrigation Pd LW500 8.1 Deer, borderdyke irrigation Pd LW500 2.8

Dairy support, spray irrigation PdL LW800 19.7 Deer, borderdyke irrigation PdL LW800 6.2

Dairy support, spray irrigation PdL LW650 14.3 Deer, borderdyke irrigation PdL LW650 5.5

Dairy support, spray irrigation PdL LW500 10.4 Deer, borderdyke irrigation PdL LW500 4.5

Dairy support, spray irrigation VL LW800 62.5 Deer, borderdyke irrigation VL LW800 58.8

Dairy support, spray irrigation VL LW650 41.9 Deer, borderdyke irrigation VL LW650 47.0

Dairy support, spray irrigation VL LW500 28.2 Deer, borderdyke irrigation VL LW500 34.9

Dairy support, spray irrigation XL LW800 89.0 Deer, borderdyke irrigation XL LW800 144.5

Dairy support, spray irrigation XL LW650 56.6 Deer, borderdyke irrigation XL LW650 111.6

Dairy support, spray irrigation XL LW500 36.0 Deer, borderdyke irrigation XL LW500 80.1

Deer, no irrigation Blank LW800 1.5 Deer, spray irrigation Blank LW800 1.5

Deer, no irrigation Blank LW650 1.5 Deer, spray irrigation Blank LW650 1.5

Deer, no irrigation Blank LW500 1.5 Deer, spray irrigation Blank LW500 1.5

Deer, no irrigation Class 4 to 6 LW800 9.0 Deer, spray irrigation Class 4 to 6 LW800 9.0

80 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Deer, no irrigation Class 4 to 6 LW650 6.0 Deer, spray irrigation Class 4 to 6 LW650 6.0

Deer, no irrigation Class 4 to 6 LW500 4.0 Deer, spray irrigation Class 4 to 6 LW500 4.0

Deer, no irrigation Class 6 LW800 5.0 Deer, spray irrigation Class 6 LW800 5.0

Deer, no irrigation Class 6 LW650 4.0 Deer, spray irrigation Class 6 LW650 4.0

Deer, no irrigation Class 6 LW500 3.0 Deer, spray irrigation Class 6 LW500 3.0

Deer, no irrigation Class 7 LW800 1.5 Deer, spray irrigation Class 7 LW800 1.5

Deer, no irrigation Class 7 LW650 1.5 Deer, spray irrigation Class 7 LW650 1.5

Deer, no irrigation Class 7 LW500 1.5 Deer, spray irrigation Class 7 LW500 1.5

Deer, no irrigation H LW800 18.6 Deer, spray irrigation H LW800 19.9

Deer, no irrigation H LW650 10.2 Deer, spray irrigation H LW650 17.8

Deer, no irrigation H LW500 5.5 Deer, spray irrigation H LW500 14.9

Deer, no irrigation L LW800 17.6 Deer, spray irrigation L LW800 25.0

Deer, no irrigation L LW650 10.2 Deer, spray irrigation L LW650 22.0

Deer, no irrigation L LW500 5.8 Deer, spray irrigation L LW500 18.0

Deer, no irrigation M LW800 18.1 Deer, spray irrigation M LW800 22.4

Deer, no irrigation M LW650 10.2 Deer, spray irrigation M LW650 19.9

Deer, no irrigation M LW500 5.6 Deer, spray irrigation M LW500 16.5

Waitaki Water Quality Catchment Modelling 81

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Deer, no irrigation Pd LW800 9.3 Deer, spray irrigation Pd LW800 9.9

Deer, no irrigation Pd LW650 5.1 Deer, spray irrigation Pd LW650 8.9

Deer, no irrigation Pd LW500 2.7 Deer, spray irrigation Pd LW500 7.4

Deer, no irrigation PdL LW800 8.8 Deer, spray irrigation PdL LW800 12.5

Deer, no irrigation PdL LW650 5.1 Deer, spray irrigation PdL LW650 11.0

Deer, no irrigation PdL LW500 2.9 Deer, spray irrigation PdL LW500 9.0

Deer, no irrigation VL LW800 19.6 Deer, spray irrigation VL LW800 36.3

Deer, no irrigation VL LW650 11.4 Deer, spray irrigation VL LW650 30.6

Deer, no irrigation VL LW500 6.5 Deer, spray irrigation VL LW500 24.0

Deer, no irrigation XL LW800 21.4 Deer, spray irrigation XL LW800 49.2

Deer, no irrigation XL LW650 12.4 Deer, spray irrigation XL LW650 40.1

Deer, no irrigation XL LW500 7.1 Deer, spray irrigation XL LW500 30.4

Sheep & beef, borderdyke irrigation Blank LW800 1.5 Sheep & beef, no irrigation Blank LW800 1.5

Sheep & beef, borderdyke irrigation Blank LW650 1.5 Sheep & beef, no irrigation Blank LW650 1.5

Sheep & beef, borderdyke irrigation Blank LW500 1.5 Sheep & beef, no irrigation Blank LW500 1.5

Sheep & beef, borderdyke irrigation Class 4 to 6 LW800 9.0 Sheep & beef, no irrigation Class 4 to 6 LW800 9.0

Sheep & beef, borderdyke irrigation Class 4 to 6 LW650 6.0 Sheep & beef, no irrigation Class 4 to 6 LW650 6.0

82 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Sheep & beef, borderdyke irrigation Class 4 to 6 LW500 4.0 Sheep & beef, no irrigation Class 4 to 6 LW500 4.0

Sheep & beef, borderdyke irrigation Class 6 LW800 5.0 Sheep & beef, no irrigation Class 6 LW800 5.0

Sheep & beef, borderdyke irrigation Class 6 LW650 4.0 Sheep & beef, no irrigation Class 6 LW650 4.0

Sheep & beef, borderdyke irrigation Class 6 LW500 3.0 Sheep & beef, no irrigation Class 6 LW500 3.0

Sheep & beef, borderdyke irrigation Class 7 LW800 1.5 Sheep & beef, no irrigation Class 7 LW800 1.5

Sheep & beef, borderdyke irrigation Class 7 LW650 1.5 Sheep & beef, no irrigation Class 7 LW650 1.5

Sheep & beef, borderdyke irrigation Class 7 LW500 1.5 Sheep & beef, no irrigation Class 7 LW500 1.5

Sheep & beef, borderdyke irrigation H LW800 26.0 Sheep & beef, no irrigation H LW800 12.8

Sheep & beef, borderdyke irrigation H LW650 23.4 Sheep & beef, no irrigation H LW650 8.2

Sheep & beef, borderdyke irrigation H LW500 18.4 Sheep & beef, no irrigation H LW500 4.8

Sheep & beef, borderdyke irrigation L LW800 33.8 Sheep & beef, no irrigation L LW800 12.1

Sheep & beef, borderdyke irrigation L LW650 30.5 Sheep & beef, no irrigation L LW650 8.2

Sheep & beef, borderdyke irrigation L LW500 23.7 Sheep & beef, no irrigation L LW500 5.1

Sheep & beef, borderdyke irrigation M LW800 30.1 Sheep & beef, no irrigation M LW800 12.5

Sheep & beef, borderdyke irrigation M LW650 27.1 Sheep & beef, no irrigation M LW650 8.2

Sheep & beef, borderdyke irrigation M LW500 21.4 Sheep & beef, no irrigation M LW500 5.0

Sheep & beef, borderdyke irrigation Pd LW800 13.0 Sheep & beef, no irrigation Pd LW800 6.4

Waitaki Water Quality Catchment Modelling 83

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Sheep & beef, borderdyke irrigation Pd LW650 11.7 Sheep & beef, no irrigation Pd LW650 4.1

Sheep & beef, borderdyke irrigation Pd LW500 9.2 Sheep & beef, no irrigation Pd LW500 2.4

Sheep & beef, borderdyke irrigation PdL LW800 17.3 Sheep & beef, no irrigation PdL LW800 6.1

Sheep & beef, borderdyke irrigation PdL LW650 15.4 Sheep & beef, no irrigation PdL LW650 4.1

Sheep & beef, borderdyke irrigation PdL LW500 12.1 Sheep & beef, no irrigation PdL LW500 2.6

Sheep & beef, borderdyke irrigation VL LW800 54.8 Sheep & beef, no irrigation VL LW800 13.2

Sheep & beef, borderdyke irrigation VL LW650 46.7 Sheep & beef, no irrigation VL LW650 9.1

Sheep & beef, borderdyke irrigation VL LW500 34.4 Sheep & beef, no irrigation VL LW500 5.7

Sheep & beef, borderdyke irrigation XL LW800 99.4 Sheep & beef, no irrigation XL LW800 14.4

Sheep & beef, borderdyke irrigation XL LW650 84.7 Sheep & beef, no irrigation XL LW650 9.9

Sheep & beef, borderdyke irrigation XL LW500 62.3 Sheep & beef, no irrigation XL LW500 6.2

Sheep & beef, spray irrigation Blank LW800 1.5 Exotic forest All LW800 2.0 (1.0)

Sheep & beef, spray irrigation Blank LW650 1.5 Exotic forest All LW650 2.0 (1.0)

Sheep & beef, spray irrigation Blank LW500 1.5 Exotic forest All LW500 2.0 (1.0)

Sheep & beef, spray irrigation Class 4 to 6 LW800 9.0 Exotic forest Blank LW800 2.0 (1.0)

Sheep & beef, spray irrigation Class 4 to 6 LW650 6.0 Exotic forest Blank LW650 2.0 (1.0)

Sheep & beef, spray irrigation Class 4 to 6 LW500 4.0 Exotic forest Blank LW500 2.0 (1.0)

84 Waitaki Water Quality Catchment Modelling

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Sheep & beef, spray irrigation Class 6 LW800 5.0 Native forest All LW800 1.5 (0.5)

Sheep & beef, spray irrigation Class 6 LW650 4.0 Native forest All LW650 1.5 (0.5)

Sheep & beef, spray irrigation Class 6 LW500 3.0 Native forest All LW500 1.5 (0.5)

Sheep & beef, spray irrigation Class 7 LW800 1.5 Native forest Blank LW800 1.5 (0.5)

Sheep & beef, spray irrigation Class 7 LW650 1.5 Native forest Blank LW650 1.5 (0.5)

Sheep & beef, spray irrigation Class 7 LW500 1.5 Native forest Blank LW500 1.5 (0.5)

Sheep & beef, spray irrigation H LW800 18.3 Other All LW800 2.0 (0.5)

Sheep & beef, spray irrigation H LW650 16.6 Other All LW650 2.0 (0.5)

Sheep & beef, spray irrigation H LW500 13.3 Other All LW500 2.0 (0.5)

Sheep & beef, spray irrigation L LW800 24.0 Other Blank LW800 2.0 (0.5)

Sheep & beef, spray irrigation L LW650 21.6 Other Blank LW650 2.0 (0.5)

Sheep & beef, spray irrigation L LW500 16.8 Other Blank LW500 2.0 (0.5)

Sheep & beef, spray irrigation M LW800 21.3 Scrub All LW800 1.5

Sheep & beef, spray irrigation M LW650 19.2 Scrub All LW650 1.5

Sheep & beef, spray irrigation M LW500 15.2 Scrub All LW500 1.5

Sheep & beef, spray irrigation Pd LW800 9.1 Scrub Blank LW800 1.5

Sheep & beef, spray irrigation Pd LW650 8.3 Scrub Blank LW650 1.5

Waitaki Water Quality Catchment Modelling 85

Land use Soil Climate TN yield (kg/ha/y) Land use Soil Climate TN yield (kg/ha/y)

Sheep & beef, spray irrigation Pd LW500 6.6 Scrub Blank LW500 1.5

1.5 Sheep & beef, spray irrigation PdL LW800 12.2 Tussock All LW800

Sheep & beef, spray irrigation PdL LW650 10.9 Tussock All LW650 1.5

Sheep & beef, spray irrigation PdL LW500 8.5 Tussock All LW500 1.5

Sheep & beef, spray irrigation VL LW800 38.0 Tussock Blank LW800 1.5

Sheep & beef, spray irrigation VL LW650 31.6 Tussock Blank LW650 1.5

Sheep & beef, spray irrigation VL LW500 22.7 Tussock Blank LW500 1.5

Sheep & beef, spray irrigation XL LW800 54.2 Urban All LW800 4.0 (1.5)

Sheep & beef, spray irrigation XL LW650 42.7 Urban All LW650 4.0 (1.5)

Sheep & beef, spray irrigation XL LW500 29.1 Urban All LW500 4.0 (1.5)

Urban Blank LW650 4.0 (1.5)

Urban Blank LW800 4.0 (1.5)

Urban Blank LW500 4.0 (1.5)

86 Waitaki Water Quality Catchment Modelling

Appendix D Flow diversions

Lake Tekapo (Figure D-1) Outflow from Lake Tekapo is diverted to Lake Pukaki via the Tekapo Canal. The remaining outflow is carried via the Tekapo River. Two alternative estimates of the flow diversion are given below:

1. Prior to the commissioning of the Tekapo Canal in 1977, the Tekapo River carried the entire Lake Tekapo outflow with a mean flow approximately equal to 80 m 3/s (Fish & Game New Zealand 1997, Gabites & Horrell 2005). According to Long & Chesterton (2005; Appendix A of that report), the mean flow in Lake Tekapo Canal (site 8793) is approximately 76 m 3/s. Assuming that the mean outflow from Lake Tekapo is 80 m 3/s, this suggests that around 95% of the lake outflow is diverted through the canal.

2. From the flow data provided by ECan, the mean flow at Tekapo Spillway (site 71131) is approximately 15 m 3/s. The mean flow in the Tekapo River is likely to be less than this because some of the flow may still be diverted through the canal downstream of the spillway. This was confirmed by Fish & Game New Zealand (1997), who noted that the mean flow in the river is about 7-10 m3/s and sourced primarily from the Forks and Grays Rivers, Maryburn and Irishman’s Creek. Supposing that the mean river flow is 10 m3/s, then this constitutes 13% of the 80 m 3/s coming from Lake Tekapo, i.e., 87% of the lake outflow is diverted through the canal.

The two estimates for the flow diversion, 95% and 87%, are of a similar magnitude and are consistent with the majority of Lake Tekapo outflow being diverted through the canal. 95% was chosen for the model.

71131 Tekapo at Spillway

Tekapo Canal

Figure D-1: Flow diversions from Lake Tekapo.

Waitaki Water Quality Catchment Modelling 87

Lake Ruataniwha (Figure D-2)

Outflow from Lake Ruataniwha (including the Wairepo Arm) is diverted to Lake Benmore via the Ohau B (and subsequently Ohau C) Canal. The remainder of the flow is carried via the lower Ohau River.

Flow data from the Ohau C Power Station at Machine Output (site 38748) indicates a mean flow for the Ohau C Canal of approximately 256 m 3/s, whilst flow data at the Lake Ruataniwha spillgate (site 8750) indicates a mean flow for the lower Ohau River of approximately 4 m 3/s. Thus, the mean flow in the canal is around 98.5% of the total mean outflow from the lake. This estimate is consistent with the water balance for Lake Ruataniwha in Figure 10 of GHD (2009), which indicates that around 99% of the surface outflow volume from Lake Ruataniwha (including the Wairepo Arm) is diverted through the canal.

71119 Ohau at SHBr

Ohau B Canal

Ohau C Canal

Figure D-2: Flow diversions from Lake Ruataniwha.

88 Waitaki Water Quality Catchment Modelling

Lake Ohau (Figure D-3)

Outflow from Lake Ohau is diverted to Lake Ruataniwha via the Ohau A Canal (which joins the Pukaki Canal upstream of the Ohau A power station and the inlet to Lake Ruataniwha). The remainder of the flow is carried by the upper Ohau River.

According to Long & Chesterton (2005) (Appendix A), the mean flow in the Ohau A canal (site 8763) is approximately 73 m 3/s. From the flow data provided by ECan, the mean flow in the upper Ohau River at Ohau below Siphon (site 71194) is approximately 13 m3/s. This suggests that approximately 85% of the outflow from Lake Ohau is diverted through the canal.

Pukaki Canal

Ohau A Canal

71194 Ohau at Below Siphon

Figure D-3: Flow diversions from Lake Ohau.

Waitaki Water Quality Catchment Modelling 89

Lake Pukaki (Figure D-4)

Outflow from Lake Pukaki is diverted to Lake Ruataniwha via the Pukaki Canal. The remainder of the flow is carried by the Pukaki River.

According to Long & Chesterton (2005; Appendix A), the mean flow in the Pukaki Canal (site 8773) is approximately 183 m 3/s. Unfortunately data from the Wardell’s site are unavailable, so data from site 38772 Lake Pukaki Spillway Flow @ Gate 19 (one of the power station sites) are used. There are a lot of zero flows in the record since it is a spillway site, there are also missing days which may (or may not) have been zero flows. Of the flows recorded, the mean flow is approximately 30 m 3/s. This suggests that approximately 85% of the outflow from Lake Pukaki is diverted through the canal. This is probably too low given the missing data from the spillway. Therefore, it is assumed that the diversion to the canal is 95%.

71134 Pukaki at Wardells

Pukaki Canal

Figure D-4: Flow diversions from Lake Pukaki.

90 Waitaki Water Quality Catchment Modelling

Appendix E Detrended concentrations If the temporal plot of a site’s concentrations showed a trend, then the measured concentration at time was adjusted by adding the difference between the modelled concentration on 31/12/2013 (reference ‘present day’ date) and the modelled concentration at time to it, see Equation E.1. = + ( − ) ˆ(tc i )adj ˆ(tc i ) ˆ(tc 2013/12/31 )mod ˆ(tc i )mod (E.1)

where ˆ(tc i )adj is the adjusted concentration at time , ˆ tc i )( is the measured concentration time ,

ˆ(tc 2013/12/31 )mod is the modelled concentration at reference time 31/12/2013 (see below), and

ˆ(tc i )mod is the modelled concentration from the trend line at time (see below).

For sites with more than a few censored data points, excluding the censored data or substituting a value such as half the detection limit could provide a false indication of a trend or mask its existence. Instead, a linear trend line was fitted to the log-transformed concentration data (both censored and uncensored) with the coefficients estimated using Maximum Likelihood (Helsel 2012, Ch. 12). The maximum likelihood regression method requires that the data follow an assumed distribution (in this case lognormal). The assumption that the data followed a lognormal distribution was tested for each site by plotting the quantiles of the uncensored data against the theoretical quantiles of a lognormal distribution (see Figure E-1 for example, where the uncensored data is indicated by the solid black dots). The measured uncensored data for all sites were found to approximately follow a lognormal distribution.

For each site with censored data, the fitted distribution (the straight line as shown for the example in Figure E-1), was used to generate a set of modelled values for the censored data using the Regression on Order Statistics (ROS) method described in Helsel (2012) 7. The modelled set of values are indicated by the open circles in Figure E-1. That is why in the plots, the censored data always fell exactly on the line (because they have been drawn from the line). The ultimate result was a set of values for the censored data that were distributed below the detection limit. For example, if there were 100 data points (60 uncensored and 40 censored) with a detection limit of 0.1, the original data set would contain the 60 uncensored values plus 40 values specified as <0.1, whereas after generating the modelled values the new data set would contain the 60 uncensored values plus 40 values distributed below 0.1.

7 It is important to note that care needs to be exercised using the modelled set of values for the censored data because they are drawn from a probability distribution and cannot be assigned to particular times within the record.

Waitaki Water Quality Catchment Modelling 91

Figure E-1: Distribution plot: quantiles of the measured concentration data versus the theoretical quantiles of a lognormal distribution for water quality site SQ10814, Cattle Creek.

The resulting trend line for the example above is shown below in Figure E-2. The blue line indicates the fitted trend 8, the open circles indicate the uncensored data, and the vertical dashed lines indicate the censored data (illustrating that the true value of the data point may lie anywhere along the line).

Median values were then calculated based on all data, both measured (uncensored) and modelled (censored). This method is capable of handling variable detection limits and is considered robust because it avoids the problem of transformation bias. To check that the adjusted or detrended concentrations, ˆ(tc i )adj , were still lognormally distributed, quantile plots were reproduced as above for the adjusted or detrended data, see Figure E-3.

8 The blue fitted trend line in Figure E-2 is no longer linear because the data is not log-transformed as previously.

92 Waitaki Water Quality Catchment Modelling

Figure E-2: Measured concentration data and the fitted trend line for water quality site SQ10814, Cattle Creek. The open circles are the uncensored data and the vertical dashed lines are the censored data.

Figure E-3: Distribution plot: quantiles of adjusted or detrended concentration data versus the theoretical quantiles of a lognormal distribution for water quality site SQ10814, Cattle Creek.

Waitaki Water Quality Catchment Modelling 93