Modelling Graphics accompanying supplementary evidence of Andrew - - PowerPoint PPT Presentation

Te Waikoropupū Supplementary Evidence on Modelling

Graphics accompanying supplementary evidence

• f Andrew Fenemor for Tasman District Council

27 June 2018

Supplementary Presentation

2

(a) Relevance of modelling evidence (b) Summary of the implications of modelling evidence (c) Critiques of the modelling and why the modelling is fit for purpose (d) The validity of the soil parameters used in the various models (f) Model calibration, reliability and uncertainty and the implications (g) Comments on supplementary evidence of Dr Gamlen and Prof Williams

Relevance of modelling evidence

• Setting of limits in evidence of Dr Hickey and Dr Young
• Modelling future scenarios informs you about

– Risks of limits being exceeded – Limitations that chosen limits would impose on community, primary and secondary industry

3

Summary of the implications of modelling evidence

• Current development contributes 97% of the nitrate-nitrogen

in the TWS spring flow.

• All modelled scenarios maintain TWS average nitrate-nitrogen

below the upper range trigger of 0.55 g/m3

4

Modelled nitrate from Te Waikoropupū Springs for 7 land use scenarios

5

No development (AMA recharge zone) 100% dryland dairy Current Dairy Current+ Proposed 766 l/sec allocation limit from FLAG Double current irrigation Current +Proposed+ Plausible Current+Proposed+ Plausible+Unlikely 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 500 1,000 1,500 2,000 2,500 3,000 3,500 Nitrate-nitrogen concentratin (g/m3) Irrigated area in AMA recharge zone (ha)

Nitrate-Nitrogen Concentrations

Main Spring Fish Creek AMA aquifer

Scenario Calculated AMA Fully Mixed Nitrate Concentrations (g/m3) Current Dairy 0.41 Current+Proposed 0.44 Current +Proposed+Plausible 0.47 Current+Proposed+Plausible+Unlikely 0.54 766 l/sec upper allocation limit from FLAG deliberations 0.44 Double current irrigation 0.47 100% dryland dairy 0.35 No Development - AMA recharge zone 0.002

Critiques of the modelling and why the modelling is fit for purpose

• The modelling presents the long-term upper limit of effects,

and can be seen as conservative

• Good model calibration through the good match between

‘Measured’ and ‘Current state (Calibration)’ values

• Modelling is described in detail in Weir & Fenemor 2017 peer

reviewed by Dr John Bright and Joseph Thomas

6

The validity of the soil parameters used in the various models

• Loss rates of nitrogen from land uses vary with soil water

holding capacity (PAW), rainfall and irrigation, plant cover and land use practices such as stocking rates and fertilizer applications

7

Land cover or land use class Average NO3-N loss (kg/ha/year) Reference or Source of Chosen N Loss Rates Forestry 0.65 Aqualinc (2014) Table 9, and further calibrated to match measured N concentrations Intensive pasture/dairying (irrigated) 106 OVERSEER estimated average for Takaka valley, 2014-15 Dryland/low intensity pasture 68 OVERSEER estimated average for Takaka valley, 2014-15 Native grassland / hill scrubland 2.5 Hanson (2010) Tables 1-4, and further calibrated to match measured N concentrations

Plant Available Water (PAW) values used in eigen-modelling for valley floor soils

8

Mean PAW of three predominant soil series from S-Map superimposed

• n my
• riginal Fig 6:

Hamama 66-101mm Takaka 99-105mm Karamea 102-118mm

9

Validity of the soil parameters

• More detailed S-Map data analysis produces soil PAWs in the

AMA recharge zone a little lower than those used in our modelling, but not nearly as low as Mr Thoma has estimated.

• Mean PAW values for the Hamama, Takaka and Karamea soils

derived from S-Map data were found to be in the range 84- 110mm, compared with 20-39mm calculated by Mr Thoma for a 0.6m plant rooting depth for pasture

• Eigen-models were adjusted during calibration to match

measured water levels and springflows, so the effects of using slightly higher PAWs are likely to be small.

• Variability of annual rainfall is a more major driver of N

leaching than soil PAW in the Takaka Valley

1 0

Use of OVERSEER in regulatory processes

(from Freeman et al 2016 Using OVERSEER in regulation)

1 1

Model calibration, reliability and uncertainty and the implications

• Simple eigen-modelling for flow and mass load modelling for

nitrogen were used because of the complexity and uncertainties around hydrology and water quality within the Takaka AMA.

• Appendix 1 addressed model reliability and uncertainty for

FLAG

• Assumes

– Fully mixed at TWS : stability of FOGB data supports this – No attenuation (reduction in N by denitrification, biological and chemical degradation) : conservative assumption – Calculated N losses in recharge zone create current N concentrations at TWS : valid if losses 8 years ago were same or more than currently, assuming 8 year average lag time – Dairy farming in valley floor : conservative assumption – 97% valley floor N contribution may be an over-estimate

1 2

Comments on supplementary evidence of Dr Gamlen and Prof Williams

• See handout

1 3