Latest research outputs

GNS Science’s hydrogeology-related research outputs for the period July 2015 to October 2016

The list of our outputs is organised according the following five research questions (click on a question to jump directly to the relevant outputs):

1. What are the hydrogeological and structural characteristics of New Zealand's aquifer systems?

2. What are the fluxes of water into, out of, and through New Zealand's aquifers?

3. What are the fluxes of key substances into, out of, and through New Zealand's aquifers?

4. How have/will human activities, climate change and other pressures affect New Zealand's groundwater resources?

5. How can we ensure that stakeholders will use our research results appropriately and efficiently?

Each question acts as a building block for us to answer the high-level question: How can we improve the sustainable management of, and economic returns from, New Zealand’s groundwater resources?

horizontal line red longest

1. What are the hydrogeological and structural characteristics of New Zealand's aquifer systems?

Output 1.1      Classification of the national 1:250,000 geological map into hydrogeological units (GWR funded)

Why it's important 

First step toward a seamless, nationally consistent, map and classification of New Zealand aquifers.

In this study, QMAP lithological and chrono-stratigraphic information (i.e., main rock type; geological age; and secondary rock type) were used to develop a nationwide assessment of surficial hydrogeological units and their properties. A number of subsequent map products were delivered. One of these, the Aquifer Potential class (Figure 1.1), shows a good match with the aquifer boundaries of Moreau and Bekele (2015). Additionally, the map provides a quick and simple way to communicate hydrogeological information. The map is capable of refining aquifer boundaries on the regional scale where these boundaries have not been updated since 2001. Future work will include categorising geological system knowledge (e.g., depositional environment) to allow for 3D mapping and characterisation; compilation and incorporation of nation-wide measured hydraulic conductivity values, including uncertainty; and linking with other national data sets, such as the NGMP.

Tschritter, C.; Westerhoff, R.; Rawlinson, Z.; White, P. 2016. Aquifer classification and mapping at the national scale – Phase 1: Identification of hydrogeological units. GNS Science Report 2016/51. (in prep).

Westerhoff, R.S.; Tschritter, C.; Rawlinson, Z.J.; White, P.A. 2016. Classification of the New Zealand geological map into hydrogeological and aquifer properties. Water Infrastructure & the Environment 28 Nov–2 Dec 2016, Queenstown. (accepted).

Moreau, M.; Bekele, M. 2015. Groundwater component of the water physical stock account (WPSA). GNS Science Consultancy Report 2014/290. 24 p.

Figure 1.1: Aquifer potential class for New Zealand

Output 1.2      Hydrogeological interpretation of Airborne Electromagnetic (AEM) data in Otago (SAC funded)

Why it's important 

Methods of interrogating large data sets to obtain hydrogeological characteristics are essential to understand and predict groundwater flows robustly over large areas.

A robust data interpretation using machine-learning (ML) algorithms is currently being trialled to map 3D hydrogeological properties over large areas utilising existing AEM data in Otago. The aim of this interpretation is to map hydrostratigraphic units (HSUs) down to ~100 m at a starting resolution of 1 m close to the surface (increasing with depth) by utilising AEM data in conjunction with all available ground-truthing data: QMAP surface geology; lithological logs; pumping test derived hydraulic properties; water level measurements; water chemistry analyses; and other geophysical measurements. This information will be crucial for informing on aquifer hydrogeological and structural characteristics and will provide data sets useful for scenario modelling using numerical groundwater flow models. We have built a collaborative relationship with Geoscience Australia through this project.

Friedel, M.J. 2016. Smart aquifer characterization and mapping with machine-learning and evolutionary techniques. Australian Earth Sciences Convention, Adelaide, Australia, 26-30 Jun.

Friedel, M.J.; Esfahani, A.; Iwashita, F. 2016. Toward real-time three-dimensional mapping of surficial aquifers using a hybrid modeling approach. Hydrogeology Journal, 24(1): 211-229; doi: 10.1007/s10040-015-1318-2.

Friedel, M. J.; Rawlinson, Z.; Westerhoff, R. 2015. Intelligent mapping of an alluvial aquifer in the Otago region, New Zealand, European Geoscience Union General Assembly 2015, Vienna, 12-17 Apr.

Output 1.3      Improved groundwater system characterisation and mapping using hydrogeophysical data and machine-learning (ML workflows) (SAC, TVH, Geoscience Australia funded)

Why it's important 

Big data techniques, such as ML, enable maximal use of existing data and fill data gaps.

Numerical models provide a way to evaluate groundwater systems, but determining the HSUs used in constructing these models remains subjective, non-unique, and uncertain. GNS Science demonstrated the efficacy of using an ML workflow to arrive at continuous distribution of HSUs from sparse, spatially-limited and scale-dependent hydrogeophysical data (Figure 1.2). In addition to other water-science topics, GNS Science is using ML workflows to arrive at transdisciplinary solutions in climate and land-use change, ecosystems, energy and minerals, and solid-earth studies.

Friedel, M.J. 2016. Estimation and scaling of hydrostratigraphic units: application of unsupervised machine learning and multivariate statistical techniques to hydrogeophysical data. Hydrogeology Journal, doi: 10.1007/s10040-016-1452-5.

Friedel, M.J. 2016. Improved groundwater system mapping and characterization workflows using machine-learning and evolutionary techniques. GNS Science International Limited Consultancy Report 2016/13, 26 pp.

Figure 1.2: Regional-scale HSU estimates at airborne sounding locations (24 depths at 35,735 locations). HSU estimates reflect groupings of lithologies, hydraulic properties, aqueous chemistry and isotopes, and field parameters. Only the predominant lithology is presented in the legend.

Output 1.4      3D geological modelling (GWR, BOPRC, Environment Southland (ES), Waikato Regional Council (WRC) funded)

Why it's important 

Fundamental to understanding the 3D distribution of our aquifers and supporting groundwater flow model development.

New Zealand’s largest 3D geological model, for the Southland region, was finalised. Outputs from this model form the structural hydrogeological foundation for the development for the first of four Freshwater Management Units’ groundwater flow models.

Between 2006 and 2015, GNS Science developed seven sub-regional three-dimensional (3D) geological models for BOPRC: Western Bay of Plenty (Tauranga), Paengaroa-Matata (Matata), Rangitaiki Plains, Opotiki-Ohope, Tarawera, and Upper Rangitaiki. In the past year, two of these models, Tauranga and Matata, were updated and merged and are now being used as the fundament of groundwater flow models.

Tschritter, C.; Rawlinson, Z.J.; Barrell, D.J.A.; Alcaraz, S.A. 2016. Three-dimensional geological model of Environment Southland's area of interest for freshwater management. GNS Science Consultancy Report 2015/123, 67 p.

Tschritter, C.; Rawlinson, Z.; White, P.A.; Schaller, K. 2016. Update of the 3D geological models for the Western Bay of Plenty and Paengaroa-Matata area. GNS Science Consultancy Report 2015/196.  67 p.

White, P.A.; Tschritter, C. 2016. Geological model of the Upper Waikato catchment. GNS Science Consultancy Report 2015/199.

Output 1.5      Assessing the worth of using AEM data vs traditional methods for aquifer parameter characterisation (SAC funded)

Why it's important 

Data worth analyses enables optimisation of new data acquisition to minimise cost while reducing the uncertainty of groundwater model predictions.

Do the more detailed spatial patterns of aquifer parameters from AEM surveys improve the reliability of groundwater impact assessments? A data worth case study analysis has been conducted on a small-scale flow model. AEM data has been used to refine hydraulic conductivity distribution. Parameter identifiability is used to quantify the improvement on the model structure of this added data (Figure 1.3). Parameter identifiability which ranges from 0 (“unidentifiable”) to 1 (“identifiable”) indicates the uncertainty associated with a parameter estimate. If the parameter has an identifiability of 0, it is informed only by expert knowledge or prior information, whereas a parameter with an identifiability of 1 has been estimated with perfect certainty. Identifiability values between 0 and 1 reflect uncertainties linked to measurement, model structural error and noise. The worth of this AEM data in terms of improving parameter identifiability is then compared to adding information to the model through traditionally acquired data (e.g., more head measurements, pump tests).

Moore, C.; Rawlinson, Z.; Moreau, M. 2016. Predictive uncertainty and data worth analysisto determine cost effectiveness of airborne EM data for defining hydraulic properties in a groundwater flow model used for predicting long term groundwater level drawdowns. International Association of Hydrogeologist Annual Conference 2016, 25-29 Sep.

Figure 1.3: Spatial distribution of parameter identifiability (draft) in the modified Cromwell Modflow model (layer1 right; layer 2 left) after inclusion of information from AEM data. Warmer colours indicate parameter that are more identifiable.

2. What are the fluxes of water into, out of, and through New Zealand's aquifers?

Output 2.1      Precipitation isoscapes for New Zealand (GWR funded)

Why it's important 

Provides isotopic input that is required for modelling or interpretation.

Precipitation isotope ratios (δ2H and δ18O) have been mapped across New Zealand (Figure 2.1) using three years of integrated monthly precipitation samples from more than 50 stations and precipitation-weighted monthly climate parameters to give statistically-valid predictions of the isotope composition of precipitation. This approach has substantial benefits for studies that require the isotope composition of precipitation during specific time intervals.

Baisden, W.T.; Keller, E.D.; Van Hale, R.; Frew, R.D.; Wassenaar, L.I. 2016. Precipitation isoscapes for New Zealand: enhanced temporal detail using precipitation-weighted daily climatology. Isotopes Environmental Health Studies, 52(4-5): 343-52. doi: 10.1080/10256016.2016.1153472.

Figure 2.1: Mean annual δ18O in rainfall, and similar maps for δ2H and d-excess or underlying data available for specific time periods can be used for many purposes in hydrology, hydrogeology, paleoclimate and other environmental studies.

Output 2.2      A new paleo-temperature archive for New Zealand (GWR, Lamont Climate Center, Earth Institute Internship Programme, Columbia University funded)

Why it's important 

Provides input to terrestrial paleoclimate reconstructions and paleoceanographic temperature studies.

Noble gas paleothermometry in paleogroundwater from the Deep Moutere, Deep Wairau and Taranaki aquifers enabled the development of a new paleo-temperature archive for New Zealand. This archive was used to reconstruct the mean annual surface temperature over the last glacial period. The ∼4.6°C cooling from the last inter-glacial period to present day estimate agrees with a number of terrestrial paleoclimate reconstructions and nearby paleoceanographic temperature studies. This demonstrates that paleoclimate information is archived in New Zealand’s groundwater, providing an additional tool for understanding past climatic conditions for the validation of climate models.

Seltzer, A.M.; Stute, M.; Morgenstern, U.; Stewart, M.K.; Schaefer, J.M. 2015. Mean annual temperature in New Zealand during the last glacial maximum derived from dissolved noble gases in groundwater. Earth and Planetary Science Letters, 431: 206-216.

Output 2.3      A rainfall-recharge lysimeter for installation in a cropping situation is now available (GWR, Hawke’s Bay Regional Council (HBRC), NIWA funded)

Why it's important 

Lysimeters provide in-situ groundwater recharge measurements (in conjunction with ground-level rain gauge).

Lysimeter measurements are used to calibrate and validate modelled estimates of rainfall recharge over the larger scale (e.g., catchment, sub-regional and national), which are critical water budget parameters for effective water management. The traditional lysimeter design, which was used to date by GNS Science to install lysimeters in five regions, cannot be used in cropping situations as it does not allow for cultivation of the soil within and around the lysimeter. The new design allows for cultivation over the soil monolith. It was developed by GNS Science, HBRC and Lincoln Agritech, with input from NIWA. The first cropping lysimeter has been installed on a private farm in Otane, Hawkes Bay. 

Output 2.4      Novel age tracers: radon and halon (SAC, GWR, Horizons Regional Council funded)

Why it's important 

New tracer techniques have been developed to: solve contamination and sensitivity issues for CFCs and SF6; and locate surface water connections at a low cost.

It was previously reported that a new age tracer, Halon-1301, was developed as a suitable replacement for chlorofluorocarbons, which has contamination and sensitivity issues. The use of Halon-1301 is extremely efficient as it can be analysed simultaneously with SF6 from the same sample. This causes little additional cost for the large benefit gained from taking a third age tracer sample. The use of multiple tracers, including Halon, increases the robustness of water age interpretation.

Radon surveys for groundwater-surface water interaction characterisation in gravel river bed systems were conducted in both the Hutt and Mangatainoka rivers. The groundwater discharge flux and pattern identified by radon were not always consistent with previous concurrent flow gauging surveys, highlighting the need for a multi-faceted approach. In some sections of the studied rivers, the concurrent flow gauging data indicated areas of groundwater recharge or discharge where the radon data showed the opposite process to be occurring. Evidence pointed to underflow beneath the gravels and other parafluvial exchange processes causing the concurrent flow gauging results to be misleading. Flow gauging combined with radon sampling provides a better understanding of the groundwater and river water interaction processes in the gravel-bed rivers.

Martindale, H. 2015. The use of radon and complementary hydrochemistry tracers for the identification of groundwater - surface water interaction in New Zealand. Master of Environmental Management, Massey University, Palmerston North. MSc Thesis. 146 p. 

Output 2.5      Distributed Temperature Sensing (DTS) studies in New Zealand (Waterscape, BOPRC, HBRC, ES funded)

Why it's important 

High-precision techniques to locate and measure groundwater inflow to rivers, streams and lakes.

A better understanding of groundwater–surface water interaction has been obtained in the Bay of Plenty, Hawke’s Bay and Southland regions using fibre optic distributed temperature sensing (FODTS) in conjunction with concurrent gaugings and/or radon surveys. A 5 km length of fibre optic cable was deployed for the first time in the Ngaruroro River, Hawke’s Bay, during low-flow conditions. The FODTS results were consistent with the gauging measurements indicating there was very little, if any, direct groundwater inflow to the river. This indicated an increase in river flow rate in the coastal reach downstream of Fernhill Bridge was likely to be via contribution from spring fed tributaries, e.g., Tutaekuri-Waimate Stream.

A deployment in the Waimea Stream, Southland, included radon sampling, FODTS, stream gauging, and water chemistry sampling, to elucidate the impacts of different physiographic settings on water quality entering the stream (Figure 2.2). FODTS results indicated that inflow to the stream was from small drain tributaries rather than diffuse inflow through the bed of the Waimea Stream. This information will allow ES to better understand and manage the impacts of land use on water quality within the catchment. In addition, FODTS was used to monitor the temperature profile of a geothermal monitoring well in Wairakei Geothermal Field, Taupo; and in a trial to determine suitability for using the technique for identifying inflow of groundwater and infiltration to sub-surface drains.

Figure 2.2: Conjunctive use of radon and DTS to locate and quantify groundwater-surface water interaction in the Waimea Stream, Southland. On the left is a summary of average hourly temperature traces and draft inflows from one of the DTS deployments on the Waimea Stream. On the right is shown the location of sampling sites and results of radon analysis.

Output 2.6      Bi-modal hydrodynamics observed at an NGMP coastal well near Blenheim (GWR funded)

Why it's important 

Understanding seasonal variation in groundwater sources in monitoring wells is required to interpret changes in groundwater quality.

Repeated age tracer sampling was undertaken in a shallow coastal well near Blenheim. This sampling was carried out to distinguish between young and old groundwater from nearby and far sources to understand changes in the redox state of the groundwater and related significant variability in nitrate concentration. We found that the water in the wells has seasonally alternating sources, pulses of fresh oxic locally recharged water high in nitrate during the recharge season, and older anoxic water indicating longer flow pathways from further away. This situation of alternating sources of the water in monitoring wells may not be uncommon at coastal, shallow wells throughout New Zealand.

Morgenstern, U.; Moreau, M.; Davidson, P. 2016. Changes in groundwater age and source during drought conditions, and connections between surface water and groundwater in the Lower Wairau Valley, New Zealand. International Association of Hydrogeologists Annual Conference 2016, 25-29 Sep.

 Output 2.7      Development of binary mixing model approach using multi-tracer time series (TVH funded)

Why it's important 

Robust transient age tracer transport models are required for assessing drinking water security.

In complex geological systems where several aquifers can contribute to the water discharge of a well, the application of simple lumped parameter models can result in misleading groundwater age distributions. Significant fractions of young water from shallow flow pathways might be underestimated which is important for the assessment of drinking water security of groundwater wells. Binary mixing models that can represent such complex age distributions, on the other hand, could not be validated via age tracers in the past due to too many parameters. Using time series of two robust age tracers, tritium and SF6, we have now been able to validate such complex mixing models, enabling for more realistic multi-parameter age distributions.

Morgenstern, U.; Daughney, C.J.; Leonard, G.; Gordon, D.; Donath, F.M.; Reeves, R. 2015. Using groundwater age and hydrochemistry to understand sources and dynamics of nutrient contamination through the catchment into Lake Rotorua, New Zealand. Hydrology and Earth System Sciences, 19: 803-822.

Output 2.8      Framework to reduce tritium and tracer analytical errors (GWR funded) 

Why it's important 

Optimisation of the tritium method, leading to improved and new applications.

A Generalised Likelihood Uncertainty Estimation (GLUE) framework was developed to improve the analysis of uncertainties associated with the use of lumped parameter models to derived mean residence time. This initiative is a collaboration between GNS Science (New Zealand), the Institute of Environment and Water Research (Spain), the Research Center for Environmental Health (Germany), and the Luxembourg Institute of Science and Technology (Luxembourg).

Gallart, F.; Roig-Planasdemunt, M.; Stewart, M.K.; Llorens, P.; Morgenstern, U.; Stichler, W.; Pfister, L.; Latron, J. 2016. Implementing a GLUE-based approach for analysing the uncertainties associated with the modelling of water mean transit times using tritium. In: EGU General Assembly 2016, 17-22 April, 2016, Vienna, Austria, Geophysical Research Abstracts, vol. 18:7641, 2016

Gallart, F.; Roig-Planasdemunt, M.; Stewart, M.K.; Llorens, P.; Morgenstern, U.; Stichler, W.; Pfister, L.; Latron, J. 2016. A GLUE framework to improve the analysis of catchment baseflows: a proof-of concept study relying on tritium and tracer analytical errors. In press Hydrological Processes.  

Output 2.9      Increased capacity for tritium sample processing at GNS Science (GNS Science SIF funded)

Why it's important 

Enhances capability and reduces turnaround time.

Two new Quantulus counters were purchased for the Water Dating Laboratory to increase its capacity and turnaround times to process radon and tritium samples. Quantulus counters are used for measuring alpha and beta-emitting radioactive isotopes such as tritium, radon, 14C, and 32Si.

Output 2.10      High precision dating techniques (GWR funded)

Why it's important 

High precision tritium measurements can be used to define groundwater dynamics parameters (e.g., recharge rate, groundwater storage) that can in turn be used for modelling groundwater and surface water flow and transport.

Tritium, if measured at the precision at which GNS Science’s Water Dating Laboratory provides, is a valuable tracer for improved understanding of hydrological systems, in both the southern and northern hemispheres. When tritium is used in combination with other age tracers (e.g., SF6, CFC, halon) the technique becomes even more powerful as it reduces the uncertainty. The combined age tracer approach provides the transit time of the older water component in streamflow. Documenting these longer timescales is vital for understanding the flow and water quality responses of rivers to changes in land use, diffuse or point-source pollution, ecological degradation, and climate change; responses which have proven to be considerably longer than previously thought. Including the longer timescales in models allows for a more robust water management approach.

Stewart, M.K.; Morgenstern, U. 2016. Importance of tritium‐based transit times in hydrological systems. WIREs Water. doi: 10.1002/wat2.1134.

The decrease of the bomb tritium in the environment allows for new applications for the tritium method, for example: estimating parameters of groundwater dynamics such as recharge rate and groundwater storage; linking between water dynamics through catchments and geology; and identification of methane sources. Because of the limited occurrence of instrumented catchments within New Zealand, some of our GWR funded research is dedicated to overseas collaborations, in order to demonstrate robustness of the new methods, via the peer review process of international journals. Recent examples of such applications were:

  • Transit times from rainfall to baseflow (GWR, Flinders University funded): Tritium was used to define transit times of water contributing to streams from the upper reaches of the Ovens River in south-east Australia. Mean transit times of years to decades imply that these streams are buffered against rainfall variations on timescales of years. Impacts of any changes to land use in these catchments may take years to decades to manifest themselves in changes to streamflow or water quality. The dynamics of the water through headwater catchments, despite their importance as significant contributors to total flow to many river systems, are still poorly understood. The development of new techniques, subsidised by the Australian research grant, in this work has direct relevance to New Zealand catchment studies and brings international validation and credibility of the approach.

    Cartwright, I.; Morgenstern, U. 2015. Transit times from rainfall to baseflow in headwater catchments estimated using tritium: the Ovens River, Australia. Hydrology and Earth System Sciences, 19: 3771-3785, doi:10.5194/hess-19-3771-2015.

  • Groundwater transit times and volumes in Hokkaido, Japan (GWR funded): Groundwater and vadose zone transit times and groundwater storage volumes through entire river catchments in Japan were estimated using single tritium measurements collected from the catchment discharge (river) during baseflow conditions. This is the first application of New Zealand’s high-precision tritium technique on Northern Hemisphere waters recharged after the atmospheric nuclear bomb testing effects had decayed. This work was undertaken in response to the Japanese “Water Cycle-Policy” Act of March 2014 requiring quantification of the subsurface groundwater volume. To date, the GNS Science tritium technique is the only robust and cost effective approach available. The approach has useful application in New Zealand for cost effective groundwater storage assessment.

    Gusyev, M.A.; Morgenstern, U.; Stewart, M.K.; Yamazaki, Y.; Kashiwaya, K.; Nishihara, T.; Kuribayashi, D.; Sawano, H.; Iwami, Y. 2016. Application of tritium in precipitation and baseflow in Japan: a case study of groundwater transit times and storage in Hokkaido watersheds. Hydrology and Earth System Sciences, 20: 3043-3058, doi:10.5194/hess-20-3043-2016.

  • Water dating techniques applied to quantify the exchange process between rivers and groundwater (GWR, Commonwealth Scientific and Industrial Research Organisation funded): Characterisation of infiltration rates for the Lower Namoi River, New South Wales, Australia. The exchange of water between the river and alluvial aquifers are known to be a significant component of the water balance, but poorly understood in terms of location and flux. This study showed that GNS Science water dating techniques can be applied to quantify the exchange process. The application of this technique in Australia and publication of the results bring international validation and credibility of the approach.

    Lamontagne, S.; Taylor, A.R.; Batlle-Aguilar, J.; Suckow, A.; Cook, P.G.; Smith, S.D.; Morgenstern, U.; Stewart, M.K. 2015. River infiltration to a subtropical alluvial aquifer inferred using multiple environmental tracers. Water Resources Research,51: 4532-4549, doi:10.1002/2014WR015663.

  • Dating of two ice cores from the southern and central Tibetan Plateau region (external collaboration technical advisory group funded): The results show that the two sites had not received net ice accumulation since at least the 1950s and 1980s, implying annual ice loss rate of more than several hundred millimetre water equivalent over the past 30-60 years. This mass loss raises concerns over the rapid rate of glacier ice loss and associated changes in surface glacier runoff and water availability in large areas of Asia. The project and publication brings valuable international exposure of the Water Dating Laboratory and credibility of the approach.

    Grigholm, B.; Mayewski, P.A.; Kang, S.; Zhang, Y.; Morgenstern, U.; Schwikowski, M.; Kaspari, S.; Aizen, V.; Aizen, E.; Takeuchi, N.; Maasch, K.A.; Birkel, S.; Handley, M.; Sneed, S. 2015. Twentieth century dust lows and the weakening of the westerly winds over the Tibetan Plateau. Geophysical Research Letters, 42: 2434-2441, doi:10.1002/2015GL063217.
    Kang, S.; Wang, F.; Morgenstern, U.; Zhang, Y.; Grigholm, B.; Kaspari, S.; Schwikowski, M.; Ren, J.; Yao, T.; Qin, D.; Mayewski, P.A. 2015. Dramatic loss of glacier accumulation area on the Tibetan Plateau revealed by ice core tritium and mercury records. The Cryosphere, 9: 1213-1222, doi:10.5194/tc-9-1213-2015.

3. What are the fluxes of key substances into, out of, and through New Zealand's aquifers?

Output 3.1      Origin of methane in groundwater (GWR, Queensland University of Technology funded)

Why it's important 

Effective management of groundwater resources relies on discriminating natural from anthropogenic processes.

A comprehensive Australian data set including tritium and radiocarbon was used to uncover the origin of methane in groundwater in a coal seam gas and overlying alluvial aquifer. Results indicate that methane in the alluvial aquifer, which is an important water resource for agriculture, is most likely derived from in-situ processes, rather than migration from a deep gas reservoir that is exploited for coal seam gas reserves. The nature publication brings invaluable exposure and credibility for attracting future research funding.

Owen, D.D.R.; Shouakar-Stash, O.; Morgenstern, U.; Aravena, R. 2016. Thermodynamic and hydrochemical controls on CH4 in a coal seam gas and overlying alluvial aquifer: new insights into CH4 origins. Nature Scientific Reports, 6: 32407, doi: 10.1038/srep32407, http://www.nature.com/ articles/Srep32407.

Output 3.2      New method for mapping groundwater transit time through the vadose zone (TVH funded)

Why it's important 

Transit time through the vadose zone has implications for nitrate transport.

A new method for mapping the groundwater transit time through the vadose zone using depth to groundwater and recharge rate was applied to the Ashley-Waimakariri area at the request of Environment Canterbury (ECAN). This work has shown that the transit time of groundwater through the vadose zone ranges from near instantaneous in the vicinity of large rivers to over 50 years in areas of deep groundwater and low permeable materials (Figure 3.1). This information has important ramifications for nitrate transport through the aquifer system, given that significant transit time can occur for the vadose zone alone.

Figure 3.1: Transit time through the vadose zone, Ashley-Waimakariri area.

Output 3.3      National groundwater quality monitoring (GWR funded)

Why it's important 

Supports characterisation of the state and trends, through time, of New Zealand groundwaters.

The NGMP network presently consists of 110 active sites across New Zealand (Figure 3.2). In the last year, two sites were replaced in the Canterbury region. Since June, a 100 mL bottle has been added to the NGMP sampling kit, following an operational review against the 2012 Standards for the examination of water and wastewater. The current sampling kit for NGMP consists of: one 250 mL, unfiltered, unpreserved bottle for the analysis of bicarbonate and electric conductivity; two 100 mL filtered, acid-preserved bottles for the analysis of cations (HNO3 preserved) and ammonia-nitrogen (H2SO4 preserved); and one 100 mL filtered, unpreserved bottle for the analysis of the remaining anions. Parallel ammonia-nitrogen analysis for selected samples has been undertaken for the June and September 2016 sampling events. Groundwater quality data collected as part of NGMP can be accessed through the Geothermal and Groundwater database.

Figure 3.2: NGMP site locations.

Output 3.4      Geochemical baseline for southern New Zealand (GNS Science core funded)

Why it's important 

Determining chemical concentration baseline (soil and groundwater) provides context to define anomalous concentrations.

A multi-element geochemical soil survey of southern New Zealand has been undertaken to define national baseline soil compositions. Sample collection, preparation and analytical methodologies were tested to develop an appropriate survey design for a national-scale baseline survey. Soil samples were collected from 348 sites, spaced approximately 8 km apart, from two depths: 0-30 cm (A-horizon); and 50-70 cm (generally B-horizon, where minimal soil development, C-horizon). Splits of samples were analysed using X-Ray Fluorescence for the following major and trace elements: Si; Al; Fe; Ca; Mg; Na; K; Mn; Ti; P; Cr; and Ba. Total C and S were also measured, and a suite of 65 trace elements including rare earth elements were analysed using Induced Coupled Plasma - Mass Spectrometry on Aqua Regia digested dilutions. A further subset of samples were analysed for Sr, C, N and S isotopes. Preliminary analysis indicates that regional variations in chemical concentration across southern New Zealand soil chemistry is strongly influenced by underlying rock type, particularly for the deeper soil samples. Some variation of some element concentrations (e.g., S, P, Pb, Hg, Cd) is attributed to anthropogenic input, for instance from fertilizers, paints, vehicle emissions and industrial emissions, particularly for the A-depth samples.

Martin, A.P.; Turnbull, R.E.; Rattenbury, M.S.; Baisden, W.T.; Christie, A.B.; Cohen, D.R.; Hoogewerff, J.A.; Rogers, K.M. 2015. Geochemical atlas of southern New Zealand. GNS Science Report 2015/26, 221 p.

Martin, A.P.; Turnbull, R.E.; Rattenbury, M.S.; Cohen, D.R.; Hoogewerff, J.; Rogers, K.M.; Baisden, W.T.; Christie, A.B. 2016. The regional geochemical baseline soil survey of southern New Zealand: design and initial interpretation. Journal of Geochemical Exploration, 167: 70-82.

Output 3.5      Dual-isotope nitrate reveals spatial and temporal variations in nitrate sources in a spring-fed stream entering Lake Ellesmere (GWR funded)

Why it's important 

Stable isotope data to provide integrative measures of catchment nitrate loss pathways.

Viable indicators of nitrogen (N) attenuation at the catchment scale are needed in order to sustainably manage global agricultural intensification. We hypothesised that the dominance of a single land use (pasture production) and strong ground-to-surface water connectivity would combine to create a system in which surface water nitrate isotopes (δ15N and δ18O of NO3-) could be used to monitor variations in catchment-scale attenuation. Nitrate isotopes were measured monthly over a two-year period in four reaches along a spring-fed, gaining stream (mean NO3--N of 6 mg L−1) in Canterbury, New Zealand. The stream water NO3- pool indicated that the highest degree of denitrification occurred in the shallow upper reaches. Moving downstream through increasingly sandy soils, the isotopic signature of denitrification became progressively weaker. The lowest reaches fell into the expected range for NO3- produced from the nitrification of pasture N sources (urine and fertilizers), implying that the attenuation capacity of the groundwater and riparian systems was lower than the rate of N inputs. After excluding months affected by effluent spills or extreme weather (n = 4), variations in the degree of denitrification over stream distance were combined with the measured NO3- discharge to estimate N attenuation over time in the sub-catchment. Attenuation was highly responsive to rainfall: 93% of calculated attenuation (20 kg NO3--N ha−1 yr−1) occurred within 48 hours of rainfall. These findings demonstrate the potential for detailed NO3- stable isotope data to provide integrative measures of catchment NO3- loss pathways.

Wells, N.S.; Baisden, W.T.; Horton, T.; Clough, T.J. 2016. Spatial and temporal variations in nitrogen export from a New Zealand pastoral catchment revealed by streamwater nitrate isotopic composition. Water Resources Research: doi: 10.1002/2015WR017642.

4. How have/will human activities, climate change and other pressures affect New Zealand's groundwater resources?

Output 4.1      Four mapping methods tested to communicate model uncertainty (TVH funded)

Why it's important 

Management decisions must be informed by both model outputs and associated uncertainties to be effective.

Four different mapping methods for displaying the model uncertainty were presented to stakeholders at Greater Wellington Regional Council’s Science Advisory Group and during the Uncertainty Workshop held in Christchurch in April 2016. The mapping methods were illustrated using groundwater LE in the Wairarapa Valley model. Groundwater LE is defined here as the time that will elapse before a parcel of groundwater exits the aquifer system via discharge to surface water. The four representation methods were: paired maps; intrinsic and extrinsic approaches; and hypothesis testing. The extrinsic method (Figure 4.1) was the preferred method of representation from the workshop participants (21 responses).

Daughney, C.J.; Toews, M.W.; Morgenstern, U.; Cornaton, F.J. 2016. Visual presentation of uncertainty in groundwater age and lifetime expectancy, Wairarapa Valley, New Zealand. International Association of Hydrogeologists Annual Congress 2016, 25-29 Sep.

Figure 4.1: Extrinsic approach for displaying modelled groundwater LE in the Middle Wairarapa Valley (colours), with cross-hatching indicating areas where the standard deviation in LE is greater than 10% of the corresponding mean LE.

Output 4.2      Advanced model calibration and uncertainty analysis studies in  New Zealand (SAC, SAM, GWRC, HBRC, ECAN, WRC, Environmental  Science and Research Institute (ESR) funded)

Why it's important 

Management decisions must be informed by both model outputs and associated uncertainties to be effective.

Deployment of existing advanced model calibration methods and uncertainty analysis and development of new methods have been incorporated in a number of groundwater modelling studies throughout the country. Many of these modelling projects have been funded by regional councils who are currently setting water and land use limits as required by the National Policy Statement for Freshwater Management 2014, and uncertainty quantification is required so that risks associated with future management options can be assessed. These are large regional scale models that will be scrutinised by the public and in Environment Court hearing processes. This scrutiny dictates that both the methods and assumptions used in model calibration and the uncertainty of model outputs must be conveyed and quantified. Two of the projects involved mentoring council staff to build in-house capability.

Development of new calibration, uncertainty quantification, and data worth methods have been undertaken within the SAC and SAM programmes. In collaboration with the Environmental Science and Research Institute (ESR), groundwater modelling methods have also been developed to better predict rapid flow and pathogen transport along permeable gravel lenses. This small scale modelling needs to be undertaken in a probabilistic context and has been developed to better assess the risks of pathogen contamination in fast moving heterogeneous groundwater systems (such as occurred in Hawkes Bay this year).

Gosses, M.; Moore, C.; Wöhling, T. 2016. Model reduction in coupled groundwater-surface water systems - potentials and limitations of the applied proper orthogonal decomposition (POD) method. European Geosciences Union General Assembly 2016, 17-22 Apr.

Janardhanan, S.; Moore, C.; Wolf, L. 2015. Pareto-based efficient stochastic simulation–optimization for robust and reliable groundwater management. Journal of Hydrology. doi:10.1016/ j.jhydrol.2015.12.001.

Moore, C. 2016 Smart Aquifer Characterisation validated using Information Theory and Cost benefit analysis. European Geosciences Union General Assembly 2016, 17-22 Apr.

Output 4.3      Large carbon sinks in New Zealand forests (GWR funded)

Why it's important 

May have implications for New Zealand’s future negotiating positions on climate change mitigation.

Following advances in high-resolution atmospheric modelling, recent work has shown that variation in CO2 concentrations appear to show large carbon sinks in New Zealand forests, which may have implications for New Zealand's future negotiating positions on climate change mitigation. GNS Science made important contributions to this NIWA led submission using the Biome-BioGeochemical Cycles model to provide the daily carbon cycle estimates across of all of New Zealand, and knowledge of C land-based C sinks not covered in the model.

Steinkamp, K.; Mikaloff Fletcher, S.E.; Brailsford, G.; Smale, D.; Moore, S.; Keller, E.D.; Baisden, W.T.; Mukai, H.; Stephens, B.B. 2016. Atmospheric CO2 observations and models suggest strong carbon uptake by forests in New Zealand. Atmospheric Chemistry and Physics Discussions 2016: 1-55.

Output 4.4      Changes in pasture production under climate change scenarios (GWR funded)

Why it's important 

Predictive tool for land use input into groundwater flow and transport models.

Projected impact of climate change on New Zealand’s national pasture production was modelled, using the downscaled climate outputs of six General Circulation Models (GCMs) and four different Representative Concentration Pathways (RCPs). Model ensemble mean for two types of pasture systems, sheep/beef and dairy were calculated. Furthermore, the 20-year average total annual production was evaluated for all scenarios at two time slices, 2046-2065 (mid-century) and 2081-2100 (end-of-century) by comparison to a baseline “RCP past” average from 1986-2005. Changes in seasonal pasture production rates were also investigated. Overall, modelled annual total national pasture yields increase under all scenarios in most locations, with increased CO2 atmospheric concentrations providing a strong boost to production and offsetting any adverse effects from climate. Seasonal trends indicate increases in winter and spring production but a sharp decline in summer growth, creating a seasonal shift in production and a summer feed gap likely to require adaptation.

Specific results have been presented at lowland and upland case study workshops representing the Bay of Plenty and the MacKenzie Basin and Upper Waitaki Catchment, respectively. http://ccii.org.nz/workshop-presentations/

5. How can we ensure that stakeholders will use our research results appropriately and efficiently?

Output 5.1      Workshops

Why it's important 

Direct exchange with stakeholders.

The table below summarises the workshops to which we have contributed this year:

Name Lead GNS Science programme Date held Workshop outcome
Vision Matauranga SAM December2015 This provided an important foundation for a number of case study specific meetings with iwi representatives that were subsequently held.
Uncertainty quantification SAC/ESR-core April2016 This identified technical difficulties in computing and presenting uncertainty and the need for effective two-way communication between policy and technical teams.
Groundwater-surface water interaction SAC August2015 Inform and update New Zealand researchers and water managers on methods for groundwater-surface water interaction characterisation.
Reactive transport modelling SAM September2016 This workshop was to train New Zealand modellers in reactive transport modelling for a number of important outputs including physically based assessment of land use impacts on nitrate concentrations in our freshwaters.

Output 5.2      Continued contribution to the development of national and international standards (SAC and GWR funded)

Why it's important 

Supports regular national reporting on the Environment by the New Zealand Ministry for the Environment.

GNS Science has continued to be a strong contributor at workshops on developing National Environmental Monitoring Standards (NEMS) for discrete water quality sampling; data transfer in New Zealand; and the Groundwater Markup Language 2 (GWML2). The NEMS discrete water quality sampling is still in drafting stage. GWML2 represents key entities relevant to hydrogeology such as aquifers, water wells, groundwater flow, groundwater fluid properties, and aquifer test results and provides a standardised framework for the seamless combination of hydro-climate and geo-scientific datasets in order to enable integrated analysis and visualisation workflows on-line. GWML2 was tested by significant groundwater data providers in North America, Europe, and Australasia (GNS Science and Salzburg University as part of SAC) and has now been passed by the Open Geospatial Consortium (OGC) Technical Committee as an internationally approved standard under the title OGC "WaterML 2: Part 4 – GroundWaterML 2 (GWML2)" [16-032r2]. The OGC standards “WaterML 2.0: Part 1- Timeseries” for data aggregation, developed and reported last year, has been adopted by Land and Water Aotearoa and, in the commercial software space, Hilltop Data Tamer; and is recommended by Inspire Europe and World Meteorological Organisations.

Brodaric, B.; Boisvert, E.; Lucido, J.; Simons, B.; Dahlhaus, P.; Wagner, B.; Grellet, S.; Kmoch, A. 2015. OGC GroundWaterML 2 – GW2IE FINAL REPORT. https://portal.opengeospatial.org/ files/ ?artifact_id=64688.

 Output 5.3      Web-based data portal for the Awahou catchment (NRW, Ngati Rangiwewehi and BOPRC)

Why it's important 

Resource to inform and plan for future freshwater development in the catchment.

Scientific information combined with matauranga-a-iwi information, has been incorporated into a web-based portal, using data access and visualisation methods developed through the SAC Programme. Scientific information includes groundwater and surface water datasets (e.g., water quality, flow rates, age), geology, and information from a low-flow gauging programme (Figure 5.1). Matauranga-a-iwi information included archival data and media held by the iwi, that articulated the on-going concerns iwi members have for Taniwha Springs and the Awahou Stream. The web portal is accompanied by a science report summarising existing scientific information for the catchment, and informing Ngati Rangiwewehi on the impact of continued abstraction from Taniwha Springs for municipal supply. See the NRW web portal here.

Lovett, A.P.; White, P.A. 2016. Ka TuTe Taniwha – Ka Ora Te Tangata Scientific Repository, Awahou Catchment. GNS Science Report 2016/13. 83 p.

Figure 5.1:      Screenshot of the Awahou web portal.