TWOTW - Te Whakaheke o Te Wai

Project updates 

The Te Whakaheke o Te Wai (TWOTW) research programme aims to better support water management based on the understanding of flow sources, pathways and lags. Methods are currently being developed to produce the world’s first nationally continuous maps of groundwater age, origin and flow paths, useable for all institutions involved in water management. We are testing and developing methods to represent the origin of flows in groundwater and baseflows in New Zealand’s major aquifer systems and the rivers that drain them.

This work involves a national groundwater and surface water age tracer sampling campaign, and stable isotope sampling programme.  We use complementary hydrogeological, chemical and isotope data to understand origin of recharge and flow pathways, effects of geology, seasonality and stream order. To date this work has already provided important insights into the groundwater and surface water flow systems throughout the country.

Groundwater and surface water mātauranga Māori and mōhiotanga Māori is being explored and approaches for combining this knowledge, with other sources of information, are being developed to improve groundwater management. This represents a unique combination of western science and indigenous knowledge and is demonstrating the importance of combining the two knowledge systems. A technical foundation of the research project is the provision of unique information about the water flow through groundwater systems based on isotopic signatures of the water and knowledge of the historic behaviour of these systems, and development of new modelling technologies that are capable of assimilating this information in order to validate these models

Our ambitious modelling programme is now well underway and has a number of components designed to support decision making, from the national scale through to the very local drinking water source protection zone scale.  New modelling approaches have under development to better integrate tracers, mōhiotanga Māori, and other data across this range of scales.

The programme is led by GNS Science in collaboration with multiple national and international organisations and stakeholders, including major partners NIWA and ESR. Funding was granted through the MBIE Contestable Fund and the programme runs from2018–2023. Project Leaders Catherine Moore and Uwe Morgenstern.

horizontal line red longest
horizontal line red longest

Funder: 2018 Endeavour Fund - Research Programmes, Ministry of Business, Innovation & Employment [New Zealand]

Primary collaborators: National Institute of Water and Atmospheric Research (NIWA, NZ), Institute of Environmental and Scientific Research (ESR, NZ), Te Tai Whenua O Heretaunga (Hawke’s Bay, NZ), Victoria University of Wellington (VUW, NZ), Watermark Numerical Computing (AUS),

Additional collaborators: Hawke's Bay Regional Council (HBRC, NZ), Environment Canterbury Regional Council (ECan, NZ), Monash University (AUS), University of Saskatchewan (CAN), Luxembourg Institute of Science and Technology (LUX), Hastings District Council (HDC, NZ), Hawke’s Bay District Health Board (HBDHB, NZ), Ministry for the Environment (MfE, NZ), Ngāti Kahungunu Iwi Incorporated (NZ),

Co-funding:  Institute of Geological and Nuclear Sciences (GNS Science), National Institute of Water and Atmospheric Research (NIWA, NZ), Institute of Environmental and Scientific Research (ESR, NZ), Hawke's Bay Regional Council (HBRC, NZ),  Environment Canterbury Regional Council (ECan, NZ), Monash University (AUS), University of Saskatchewan (CAN), Luxembourg Institute of Science and Technology (LUX), Hastings District Council (HDC, NZ), Hawke’s Bay District Health Board (HBDHB, NZ), Ministry for the Environment (MfE, NZ), Ngāti Kahungunu Iwi Incorporated (NZ).

Programme Duration: 1 October 2018 – 30 September 2023

Research Programme Leaders: Catherine Moore and Uwe Morgenstern

horizontal line red longest

Keywords

Metamodels, Groundwater, Data assimilation, Water dating, Downscaling, Upscaling, Artificial Intelligence, Isotope hydrology, Decision support, Model transferability, Data worth, Model scalability, Machine learning, Risk assessment, Web-based

horizontal line red longest

Overview

Figure 1 TWOTW

Te Whakaheke o Te Wai holistically describes the flow sources, pathways and lags of water moving through a catchment (Figure 1). New Zealand currently lacks this knowledge and as a result, national and sub-regional water management strategies cannot prevent land use degradation of rivers and aquifers, which impacts cultural values, drinking water supplies, agriculture, and tourism. This paradigm-shifting project will derive the whakaheke of groundwater and baseflows in New Zealand’s 200 major aquifer systems and the rivers that drain them. We will measure age tracers, which integrate all flow velocities (of water and contaminants) above any measurement point. We will use complementary hydrogeological, chemical and isotope data to understand origin of recharge and flow pathways, effects of geology, seasonality and stream order. New modelling approaches will integrate the tracer and other data across scales. The metamodel-data assimilation will yield New Zealand’s first national maps of groundwater age, origin and flow paths, useable for all institutions involved in water management. Working with hapū, iwi, and national Māori partners, we will incorporate mātauranga-a-iwi/hapū into our models alongside the tracer and other related data. We will research use of these models across institutional boundaries to promote a collective understanding of the whakaheke of water, its relationship to whakapapa and its role in empowering kaitiakitanga. Applications of the research outputs will include informing: setting of national policies, managing catchment scale contaminant inflows to groundwater-fed rivers, and protecting local potable water supplies (e.g. Havelock North).

Our goal can only be achieved through an integrated programme with fundamental scientific advances in each of four complementary areas: (1) new data to constrain groundwater age and transit time; (2) merging data with expert knowledge to conceptualise groundwater systems; (3) integration of data and conceptual understanding within numerical models; and (4) application of model predictions to address specific water management decisions at relevant scales (Figure 2).

  1. 1.     Environmental tracer survey with world-leading methods and unprecedented coverage

For over 300 years, rainfall-runoff and groundwater models have been calibrated to measurements of hydraulic head and stream flow. These data provide little insight into groundwater age or transit time as needed for improved water management. Environmental tracers, notably isotopes, give the most direct data informing flow models.

  1. 2.     Merging mātauranga-a-iwi/hapū with western science for better conceptual understanding

Historically, conceptual understanding of groundwater systems has been based on western science. It is increasingly recognised that indigenous knowledge also provides expert insight into freshwater systems. In NZ, mātauranga-a-iwi/hapū, as it pertains to landscape, springs and rivers, derives from cultural perspectives and close association with the environment. This knowledge is preserved by long oral traditions and thus can provide insights that extend into the past beyond the modern instrumental record. Increasingly, this mātauranga is applied in cultural health monitoring of rivers and supporting improved water policy and planning. The challenge is that mātauranga-a-iwi/hapū pertaining specifically to groundwater presently has low visibility, so its influence remains limited.

  1. 3.     Advanced workflows for rapid, accurate modelling at relevant scales

Only by bringing data and expert knowledge together in a numerical model is it possible to make predictions for locations without data, or to explore ‘what if’ scenarios (e.g. increased abstraction, land use change). Industry-standard groundwater software is now routinely used for these purposes, in NZ and globally. The challenge is that current methods for model calibration and uncertainty analysis are not suitable for depicting TWOTW because: they do not readily integrate tracer and other data alongside expert knowledge and mātauranga-a-iwi/hapū; they do not effectively represent highly connected flow pathways; and national, regional and local scales cannot be examined within a single model due to prohibitively slow run times.

  1. 4.     Better use of data, knowledge and models in the decisions that matter

Historically, many scientific investigations were conducted to advance the state of knowledge in general, without a specific application in mind. It is increasingly recognised that, to deliver improved water management for NZ, these investigations must be tailored from the outset to address the specific water management issues NZ is currently facing. The challenge is that different water management organisations have different information needs depending on their various mandates and scales of operation.

Figure 2 TWOTW

Figure 2. Proposed approach, showing fusion of data, expert knowledge and modelling within the context of decision-making. Existing data and methods are shown in italics; bold text indicates the scientific advances to be made in this programme.

horizontal line red longest

Recent publications and documents

Journal Publications:

Hemmings, B, Knowling, M.J., and Moore C.R., 2020. Early uncertainty quantification for an improved decision support modelling workflow: A streamflow reliability and water quality example. Frontiers in Earth Science. doi: 10.3389/feart.2020.565613.

Knowling, M.J.; White, J.T.; Moore, C.R.; Rakowski, P.; Hayley, K. 2020 On the assimilation of environmental tracer observations for model-based decision support. Hydrology and Earth System Sciences, 24(4): 1677-1689; doi: 10.5194/hess-24-1677-2020

White, J.T., Hemmings, B., Fienen, M.N., Knowling, M.J., 2021. Towards improved environmental modeling outcomes: Enabling low-cost access to high-dimensional, geostatistical-based decision-support analyses. Environmental Modelling and Software. Volume 139, May 2021, 105022 https://www.sciencedirect.com/science/article/pii/S1364815221000657?dgcid=coauthor 

Reports:

Dudley, B., Yang, J., Shankar, U., 2020. Estimated young water fractions at NIWA river water quality network sites. Prepared for GNS Science September 2020. NIWA CLIENT REPORT No: 2020287CH.

Morgenstern U., Davidson P., Townsend D.B., White P.A., van der Raaij R.W., Stewart M.K., Moreau M., and Daughney C. 2019. From rain through river catchment to aquifer: the flow of water through the Wairau hydrologic system. Lower Hutt (NZ): GNS Science. 83 p. (GNS Science report; 2019/63). doi:10.21420/7125-ST46

Aranui, A. 2020. Te Whakaheke o te Wai: Mātauranga a Iwi/Hapū Research Report. Victoria University of Wellington.

Conference Presentations:

Daughney C.J., and Morgenstern U., 2020. Use of symbolic regression to estimate groundwater age distributions from hydrochemistry, Heretaunga Plains. In:  NZ Hydrological Society Conference, from 1 – 4 December 2020, Invercargill Waihōpai, NZ: oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Dudley, B.D., Shankar, U., Yang, J., and Montgomery, K. 2019. Development of a surface water isotope layer for New Zealand. In: NZ Hydrological Society Conference 3-6 December 2019, Rotorua, NZ :oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Johnson, P.J.; Morgenstern, U.; Moore, C.R.; Cameron, S.G. 2019 Te Whakaheke o te Wai : improving understanding of groundwater flow pathways in New Zealand. p. 90 In: NZ Hydrological Society Conference 3-6 December 2019, Rotorua, NZ: oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Moore, C., 2019 Keynote Speaker, NZHS. Model design for decision support: uncertainty quantification, data assimilation and model complexity entanglement. In: NZ Hydrological Society Conference 3-6 December 2019, Rotorua, NZ : oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Moore, C. and Doherty J. 2020. How upscaling hydraulic properties undermines the reliability of our decision-support predictions. In: NZ Hydrological Society Conference, from 1 – 4 December 2020, Invercargill Waihōpai, NZ : oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Morgenstern, U., Davidson, P., Townsend, D., Stewart, M.K., 2019. Groundwater Storage that feeds the Wairau River and Streams in the Wairau Plain. In: NZ Hydrological Society Conference 3-6 December 2019, Rotorua, NZ : oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Sarris T., Scott, D.M., Close, M.E., and Moore C., 2020. Transition Probability Analysis of Lithology data: Implications for numerical delineation of well capture zones. In: NZ Hydrological Society Conference, from 1 – 4 December 2020, Invercargill Waihōpai, NZ : oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Toews, M.W.; Hemmings, B.J.C. 2019 A surface water network method for generalising streams and rapid groundwater model development. p. 166-167 In: NZ Hydrological Society Conference 3-6 December 2019, Rotorua, NZ : oral abstracts. [Wellington, N.Z.]: New Zealand Hydrological Society.

Yang, J., Jakeman A., and Rajanayaka, C., 2019. Emulation methods for sensitivity and uncertainty analyses in hydrologic modelling. MODSIM Conference. 2019, Canberra, Australia.

Outreach and community:

Aranui, A., 2019. 'Mapping Whenua: Te Whakaheke o te Wai: Mātauranga a iwi me ngā hapū' at Victoria University of Wellington for Te Kawa a Māui course MAR203 Mapping Whenua.

Morgenstern U., 2019. Wairau Aquifer Envirolink project, presented to the Environment Committee at Marlborough Regional Council. 

Presentation and discussion at Fish hook summit https://www.kahungunu.iwi.nz/fish-hook-programme