Smart Models for Aquifer Management (SAM)
The Smart Models for Aquifer Management (SAM) research programme identified optimal groundwater-surface water flow and transport models to address large-scale, real-time, specific environmental management problems.
The SAM programme focused on three integrated groundwater–surface water catchments – Hauraki, Ruamahanga and Southland. Covering a diverse range of test catchments ensured that the research outputs would be relevant, workable, and transferable to other catchments across Aotearoa New Zealand.
Freshwater management required a new approach
In 2015, at the outset of the SAM research programme, then-current modelling approaches presented a real risk to adaptive management of New Zealand's aquifers under the National Policy Statement on Freshwater 2014 (NPS-FM 2014).
Interactions between groundwater and surface water systems such as rivers, lakes, wetlands and estuaries are complex. Pre-2015 models simulating these interactions were either too complex and slow to be practical, or lacked necessary integration, or were too simple to be accurate.
Despite these deficiencies, the NPS-FM 2014 required holistic freshwater management that satisfies community aspirations. Such an approach called for integrated groundwater-surface water modelling over larger areas and at finer spatial and temporal scales than ever before.
World Water Day 2020: everything you need to know about groundwater – On World Water Day, GNS Science expert Rogier Westerhoff explains why groundwater is so important to Aotearoa New Zealand. transcript
You don't know what groundwater is? Let me explain.
You're currently looking at surface water. Groundwater is the water under the ground in between the rocks and the sediment.
Rainfall seeps through the soils and flows underground to a lower part in the catchment.
This river water also flows as surface water through the catchment and there's a larger body of water under this in the same direction, only a lot slower.
It's hard to visualise it, but look at the cliff. At the top you see all kinds of sediment that are sandy and gravelly and pretty permeable. So water can flow through there. Underneath there's hard rock and it's pretty impermeable. So water is not able to flow through there, and it has to go in a different direction mostly forced by gravity.
So this place appears to be quite dry but you would be mistaken because groundwater can still be quite close. Where the soil becomes saturated we call this the water table.
It's pretty close to the surface here. At other places in New Zealand it might be a bit deeper.
So this body of water under the river is a huge water resource for people. People can extract it. That's called an aquifer. We're just building our mini aquifer here to show the difference in aquifer types.
This is gravel, it fills up pretty quickly, and also pretty easy to extract. Imagine this straw being a bore hole and a pump - you can easily extract it.
So now we'll do the same thing, only with finer sand. This looks a lot finer and it almost would appear no water would fit but still a heap of water will fit in this little aquifer. Although it's slightly harder to extract.
So now we brought some clay. Clay is supposed to be a lot less permeable. If one could put a pump in this it would be very hard to pull out and I'm not going to prove my point because I'm not going to drink this. I definitely won't this time.
So all aquifers in New Zealand are different and that's why it's so important to understand the complexity of the geology in those aquifers, so we can better understand where to draw water from.
Drinking water heavily relies on groundwater irrigation and agriculture heavily rely on it, groundwater feeds our rivers and streams, and that gives it deep cultural meaning. I don't only mean recreations such as fishing and kayaking but also the deeper cultural meanings such as te mauri o te wai, te mana o te wai, and the ability to gather food such as mahinga kai.
GNS helps develop techniques to look at where the groundwater is and how it can be better managed, so that future generations enjoy the same benefits from it as we do.
World Water Day 2020: everything you need to know about groundwater
On World Water Day, GNS Science expert Rogier Westerhoff explains why groundwater is so important to Aotearoa New Zealand.
World Water Day 2019: what is groundwater? – Our GNS Science experts explain all you need to know - and tell you what they're doing to safeguard groundwater's future. transcript
It's quite hard to visualize groundwater, but here at Taniwha Springs you can actually see water discharging from the ground to the surface. It eventually flows into Lake Rotorua.
Rainfall recharges into the ground, and through different flow paths eventually finds its way to the surface.
Although you cannot physically see them New Zealand has a vast labyrinth of groundwater layers, they're what we call our aquifers and they provide drinking water for around 4 in 10 New Zealanders. So we couldn't manage without them.
80 percent of annual River flow comes from groundwater and it provides billions of dollars to our economy through either irrigation or tourism.
Groundwater is critical to our surface aquatic ecosystems and for Mahina Kai, freshwater food and the places it comes from.
As important as it is the quality and abundance of this vital resource are under threat. We know that 40% of catchments are vulnerable to shortage or contamination. What we don't know well enough are the impacts of climate change in economic growth. Also in many catchments we don't have enough groundwater information.
National and local methods for testing and measuring impact are advancing but are not yet where they need to be. Water use, volumes, flow paths, and fluxes are still too poorly understood.
So we're at a crossroads, we face the challenge of sustaining the social environmental and cultural values of our groundwater resources, yet we are too poorly equipped to resist pressures from economic growth and climate change. The road we need to take has to be built with our best scientific information and expertise, that's where GNS science comes in.
GNS science is a world leader in earth science research.
The work we do is geared towards creating a cleaner safer more prosperous New Zealand.
Our scientists have the skills and experience to produce ground water maps for the whole of New Zealand.
We have the expertise to produce maps that are 2d, the view from the surface, 3d, what the water looks like underground, and 4d, how this changes through time.
We'll develop national data sets that are consistent across regions. These data sets will be spatially detailed dynamic through time and applicable for a multitude of hydrogeological applications.
We'll use the same consistent underlying data to produce both local as well as national scale maps.
We'll start with our own high-resolution geological map of New Zealand, and improve it with sophisticated geophysical data.
We'll invest in the national framework for the modelling of groundwater, it will apply the most advanced the miracle models, use 3d geological models and continuously feed in the most updated data.
We'll apply those data and models and studies that aim to optimize water management under deep uncertainty.
Our research aims to have the strongest impact and benefit to New Zealand. Overall we want a program structure where our scientists are thriving and are constantly positively triggered to perform at their best while having fun.
World Water Day 2019: what is groundwater?
Our GNS Science experts explain all you need to know - and tell you what they're doing to safeguard groundwater's future.
More complexity isn’t always better
Modern environmental decision-making is largely based on numerical models. While it is recognised that “uncertainty analysis” should accompany model outputs, models are often far too complex for this to be done. Increasing complexity can increase numerical instability, drain modelling finances and time, and detract from assessment of model output uncertainty.
The SAM programme attempted to treat uncertainty as the fundamental context for environmental modelling, rather than an afterthought, using three key ideas:
- Often, only one side of a predictive uncertainty distribution is of interest: the side that assesses the possibility of unwanted, rather than wanted, events.
- A model that is tuned to testing, and maybe rejecting, the hypothesis that such an event will occur, may not need to be complex, provided it is constructed specifically to explore that particular hypothesis.
- A simple model may contribute to predictive uncertainty through its very simplicity.
Quantifying uncertainty in data availability and integrity can allow decision-makers to become better aware of what models can deliver. Predictions accompanied by a large amount of irreducible uncertainty can be distinguished from those that are not. In the New Zealand land-use management context, such an approach may prompt decision-makers to base policy and/or legislation on model outcomes that have relatively high predictive integrity.
Modelling complex interactions for environmental decision-making
The modelling within the SAM programme followed two pathways:
- best ways to train simple models from complex groundwater models in each catchment
- development of simple model designs that did not need training on a complex model, and that could be built easily in any catchment – and for which an estimate is available of their simplification error through some complex/simple studies in selected catchments
To balance between these two paths, the project was designed to answer questions such as:
- To what extent can parameters employed by a simplified model be informed by measurable characteristics of catchment geological and soil components?
- To what extent must they be informed by local calibration?
- Can parameters of a simplified model inferred through calibration in one catchment be “regionalised” for the use of simplified models in neighbouring catchments?
- What simplification strategies are appropriate for the type of model outcomes being considered? Appropriateness must take account of:
- ability to quantify predictive uncertainty while increasing it as little as possible
- the ability of simplified parameters to be informed by expert knowledge at a variety of scales as much as this as possible
- reduction of calibration-induced bias incurred through inappropriate simplification
- How can calibration-induced predictive bias of a simple model be reduced through adoption of a “simplification-smart” history matching strategy?
Te Whakaheke o Te Wai (2021) – This project focuses on the groundwater systems in Hawke’s Bay’s Heretaunga catchment, and the nature of the water and how it flows through these lands. transcript
The name of this research program is Te Whakaheke o Te Wai or loosely translated the meandering waters, and that is a reflection of our groundwater systems and also our focus that we're working in Hawke's Bay on, on the Hirotanga catchment, and the nature of the water and how it flows through our lands.
The aim of the project is to better understand groundwater flow, the pathways of groundwater flow and the origins of groundwater beneath the surface with the idea that if we can understand it better we can manage it better.
The project is a sort of sum of three parts isotope science mataranga knowledge and modelling and I lead the modelling stream. And what we're doing is to build a framework that allows models to be built very very quickly as soon as there's a concern and you tailor the model to answer the question or the concern that it's meant to be addressing, and they'll answer the question a lot more accurately.
So my role in this project is to ensure that mataranga maori or maori knowledge is incorporated into this project.
It's important that we increasingly and incorporate indigenous knowledge with western science when we're trying to and better understand things particularly in this case around the water and how we can better look after this resource that we have.
For Hawke's Bay and one outcome from this project is that we are definitely going to be able to manage our water resources much more effectively, but as a nation I think having these tools that can be applied nationally will help address the some of the big problems we have with water resource management that are fairly consistent across the country.
Over many years I have been developing techniques to read the isotopic signature of groundwater which can tell us about what where the water is being recharged so the recharge source, but also how long it has been underground which then will give us a better understanding of how to manage groundwater resources.
Apart from just the tools that this project's going to provide, I think it's one of the best things about it is working with a really really super talented bunch of people at GNS science, and getting into some novel research and practical science that is pushing the limits of what's being done before so in that respect it's really quite exciting.
Te Whakaheke o Te Wai (2021)
This project focuses on the groundwater systems in Hawke’s Bay’s Heretaunga catchment, and the nature of the water and how it flows through these lands.
The Most Accurate Water Dating Lab in the World – GNS Science in New Zealand has the world's most accurate groundwater dating facility. It is used to find the age of groundwater and glacier ice. transcript
I'm here at Buick Street in Petone where there is an artesian well. Lots of people come here to get their fresh drinking water.
But where did this water come from? And how long did it take to get here?
Our team at GNS science can date water to find out the answers to these questions.
It turns out that the water at Petone came from here, ten kilometres up the valley at Taita.
You can see that most of the water travels on the surface of the river, but some of it percolates through the gravel and enters the groundwater system.
Now I'm going to take you to the water dating laboratory at GNS Science to show you how we measure how long the water travels underground to get from here to Petone.
Here we are at the lab. This is where we enrich and isolate a natural isotope in the water called tritium. What is tritium and where does it come from?
Cosmic rays from outer space are bombarding nitrogen which is in the atmosphere. This converts a very tiny amount of the nitrogen into tritium. This is a radioactive isotope of hydrogen, only found in very tiny quantities.
When tritium combines with oxygen atoms, like normal hydrogen it forms a water molecule. This water molecule is included in the rain that falls. Some of this water goes down into the groundwater system.
Because tritium is an unstable isotope, it breaks down over time.
How do we measure the age of water?
The groundwater travels from point A to point B. We can take a sample of the water and measure the amount of tritium atoms in that sample.
Tritium has a half-life of 12.3 years. It means after 12.3 years, only half of the tritium atoms which were once in that sample are left. We can figure out how many years it has taken the water to travel in the ground between the two sites by looking at the number of tritium atoms in the two samples.
So imagine that one business card represents a hydrogen atom. We could stack business cards one on top of the other, all the way to the Sun 100,000 times and in those stacks, if one of those business cards was a tritium atom, we could detect it.
At GNS Science, we have the most accurate water dating lab in the world.
So we know how long the water travelled to get here underground. It's about 18 years.
The Most Accurate Water Dating Lab in the World
GNS Science in New Zealand has the world's most accurate groundwater dating facility. It is used to find the age of groundwater and glacier ice.
While SAM ran from 2015 - 2018, publications continue to flow from the programme.
GNS Science reports can be downloaded from the GNS Science Shop(external link).
GNS Science Reports
- Allan, M. 2018 Quantifying uncertainty within an ecologically-coupled lake hydrodynamic model : Lake Wairarapa case study. Lower Hutt, N.Z.: GNS Science.GNS Science report 2018/47. 23 p.; doi:10.21420/J482-WF31(external link)
- Elliott, S.; Rajanayaka, C.; Yang, J.; White, J. 2019 CLUES-GW : a simple coupled steady state surface-groundwater model for contaminant transport. Lower Hutt, N.Z.: GNS Science.GNS Science report 2018/44. 51 p.; doi: 10.21420/34XR-AG12(external link)
- Hemmings, B.J.C.; Knowling, M.J.; Moore, C.R. 2019 Assessing the uncertainty of water quality and lake influx predictions made using complex regional models : Ruamahanga South case study. Lower Hutt, N.Z.: GNS Science.GNS Science report 2019/30. 72 p.; doi: 10.21420/9N73-QE35(external link)
- Hemmings, B.J.C.; Knowling, M.J.; Moore, C.R. 2019 Assessing the uncertainty of water quality and quantity predictions made using complex regional models : Ruamahanga North case study. Lower Hutt, N.Z.: GNS Science.GNS Science report 2019/29. 80 p.; doi: 0.21420/J32Q-W248(external link)
- Howard, S.W.; Griffiths, J.; Zammit, C.; Rouse, H. 2019 Model choice effects on ecological modelling in Mataura River : SAM Programme 2018. Lower Hutt, N.Z.: GNS Science.GNS Science report 2019/05. 55 p.; doi: 10.21420/J32Q-W248(external link)
- Snelder, T. 2018 Nutrient concentration targets to achieve periphyton biomass objectives incorporating uncertainties. Lower Hutt, N.Z.: GNS Science.GNS Science report 2018/38. 41 p.; doi: 10.21420/AJSH-NW16(external link)
- Zammit, C.; Yang, J.; Griffiths, J.; Rajanayaka, C. 2019 Smart models for aquifer management : TopNet modelling suite. Lower Hutt, N.Z.: GNS Science.GNS Science report 2019/27. 118 p.; doi: 10.21420/FBS9-G965(external link)
- Everitt, L.C. 2020 Applications of digital baseflow separation techniques for model validation, Wairarapa valley, New Zealand. Thesis (MSc Physical geography) – Victoria University of Wellington. 165 p; link to full text at: http://researcharchive.vuw.ac.nz/handle/10063/9114(external link)
- Op den Kelder T. 2018 Using predictive uncertainty analysis to optimise data acquisition for stream depletion and land-use change predictions. Thesis (MSc Physical geography and Quaternary geology) – Stockholm University. 78 p; link to full text at: http://www.diva-portal.se/smash/get/diva2:1254304/FULLTEXT01.pdf(external link)
- Rayner, S. 2019 Understanding the potential for nitrate attenuation from paddock to stream using dual nitrate isotopes. Thesis (PhD) – Lincoln University. 185 p; link to full text at: https://researcharchive.lincoln.ac.nz/handle/10182/11449(external link)
Peer reviewed journal articles
- Hemmings, B.; Knowling, M.J.; 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.
- Doherty, J.; Moore, C.R. 2020 Decision support modeling : data assimilation, uncertainty quantification, and strategic abstraction. Ground water, 58(3): 327-337; doi: 10.1111/gwat.12969(external link); open access available to download at: https://ngwa.onlinelibrary.wiley.com/doi/full/10.1111/gwat.12969(external link)
- Knowling, M.J.; White, J.T.; McDonald, G.W.; Kim, J.-H.; Moore, C.R.; Hemmings, B.J.C. 2020 Disentangling environmental and economic contributions to hydro-economic model output uncertainty: an example in the context of land-use change impact assessment. Environmental Modelling & Software, 127: 104653; doi: 10.1016/j.envsoft.2020.104653(external link); open access available to download at: https://www.sciencedirect.com/science/article/pii/S1364815219305031?via%3Dihub(external link)
- Knowling, M.J.; White, J.T.; Moore, C.R. 2019 Role of model parameterization in risk-based decision support: an empirical exploration. Advances in Water Resources, 128: 59-73; doi: 10.1016/j.advwatres.2019.04.010(external link); open access available to download at: https://www.sciencedirect.com/science/article/pii/S0309170819300909?via%3Dihub(external link)
- 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(external link); open access available to download at:
- Rayner, S.; Clough, T.J.; Baisden, T.; Moir, J. 2020 Can ruminant urine-N rate and plants affect nitrate leaching and its isotopic composition?. New Zealand journal of agricultural research, 63(1): 87-105; doi: 10.1080/00288233.2019.1648302(external link)
Not open access so would need to be inter-loaned or purchased from the publisher at: https://www.tandfonline.com/doi/full/10.1080/00288233.2019.1648302(external link)
- Sarris, T.S.; Close, M.E.; Moore, C.R. 2019 Uncertainty assessment of nitrate reduction in heterogeneous aquifers under uncertain redox conditions. Stochastic Environmental Research and Risk Assessment, 33(8-9): 1609-1627; doi: 10.1007/s00477-019-01715-w(external link); open access full text available at: https://link.springer.com/article/10.1007/s00477-019-01715-w(external link)
- Sarris, T.S.; Scott, D.M.; Close, M.E.; Humphries, B.; Moore, C.R.; Burbery, L.F.; Rajanayaka, C.; Barkle, G.; Hadfield, J. 2019 The effects of denitrification parameterization and potential benefits of spatially targeted regulation for the reduction of N-discharges from agriculture. Journal of environmental management, 247: 299-312; doi: 10.1016/j.jenvman.2019.06.074(external link); not open access so would need to be inter-loaned or purchased through the publisher at: https://www.sciencedirect.com/science/article/pii/S0301479719308825?via%3Dihub(external link)
- Snelder, T.H.; Moore, C.R.; Kilroy, C. 2019 Nutrient concentration targets to achieve periphyton biomass objectives incorporating uncertainties. Journal of the American Water Resources Association, 55(6): 1443-1463; doi: 10.1111/1752-1688.12794(external link); not open access so would need to be inter-loaned or purchased through the publisher at:
- White, J.T.; Knowling, M.J.; Fienen, M.N.; Feinstein, D.T.; McDonald, G.W.; Moore, C.R. 2020 A non-intrusive approach for efficient stochastic emulation and optimization of model-based nitrate-loading management decision support. Environmental Modelling & Software, 126: article 104657; doi: 10.1016/j.envsoft.2020.104657(external link)
Not open access so would need to be inter-loaned or purchased through the publisher at: https://www.sciencedirect.com/science/article/pii/S1364815219309934?via%3Dihub(external link)
- White, J.T.; Knowling, M.J.; Moore, C.R. 2019 Consequences of groundwater-model vertical discretization in risk-based decision making. Ground water, Online first: doi: 10.1111/gwat.12957(external link); not open access so would need to be inter-loaned or purchased through the publisher at: https://ngwa.onlinelibrary.wiley.com/doi/full/10.1111/gwat.12957(external link)
Research project details
Primary collaborators: Victoria University of Wellington (VUW), National Institute of Water and Atmospheric Research (NIWA), Market Economics, Institute of Environmental and Scientific Research (ESR)
Additional collaborators: Beef and Lamb, CSIRO, Department of Conservation, Earth in Mind Ltd, Flinders University, Kitson Associates, Landwaterpeople Ltd, Ministry for the Environment, Ravensdown, Tubingen University, University of Waikato, Watermark Numerical Computing
Cath Moore, GNS Science
Ministry of Business, Innovation & Employment (MBIE)