Oluwunmi, P.A.; Moreau, M.; Cox, S.C. 2026. Extracting hidden information from groundwater-level monitoring: CDF project 25008. Lower Hutt, N.Z.: GNS Science. GNS Science report 2025/19. 44 p.; doi: 10.21420/4MZH-6Q74
Abstract
Major uncertainties make our current groundwater management difficult. Future groundwater availability will be determined by complex climatic conditions that are constantly changing. In contrast, current groundwater allocation is mostly fixed in time, and at times over-subscribed, leading to loss of productive efficiency and resource degradation. Complete understanding of groundwater–surface-water interconnection and effects of abstractive use is rare. A 2025 GNS Science Capability Development Fund (CDF) project aimed to examine innovative solutions for systematically evaluating the context of groundwater levels that will be needed before New Zealand can confidently shift toward more flexible and dynamic resource management required in a changing climate context. This report provides a general overview of the CDF project, an outline of the results, and information supporting future development of a ‘business case’ and/or further work. There is a perceived issue by the authors that groundwater levels and data routinely monitored by councils and the rural sector are under-utilised, based on the growing number of publications of open-source tools in international literature enabling the potential of groundwater-level data to be unlocked. The authors propose that there is an opportunity to use changes in atmospheric pressure, earth and ocean tides, earthquakes, extreme precipitation, climate cycles and sea-level rise, which each affect groundwater levels and can be routinely extracted from time-series data. The CDF project was designed to identify which parameters and plots would be most useful to distinguish effects of abstraction or other anthropogenic influences from natural processes to enable improved groundwater understanding and management in New Zealand. The project involved four distinct segments of work: (i) a review of council groundwater-level-monitoring practise and data delivery; (ii) a review of methods used for processing and testing groundwatertime-series data; (iii) development of open-source and in-house code (hosted in a Jupiter Notebook) to process data and test viability of product development; (iv) an online consultation and surveys of council and GNS Science staff to understand which methods might be most useful and applicable in day-to day-work and resource management, as well as to improve understanding of groundwater dynamics. The majority of New Zealand councils have online data portals for presenting results of monitoring; these portals display some (but not all) of the parameters that the councils measure, such as groundwater levels, which is telemetered via radio or cell-phone networks and most commonly displayed as simple hydrographs. Internationally, a wide variety of processing methods are applied to groundwater time-series, with numerous examples in scientific literature, reports and authors’ experiences. Ten examples of processing and plots were selected for evaluating applicability and usefulness: (1) temporal filtering and trend analysis, (2) Julian Day plots, (3) temporal statistics, (4) probability distributions, (5) extreme-value analysis, (6) classification and cluster analysis, (7) frequency-domain analysis, (8) tidal analysis, (9) barometric efficiency and (10) rainfall correlation.Stakeholder consultation involved an online workshop with 46 participants from 14 regional and unitary authorities and three external research providers (Aqualinc, Lincoln Agritech and MHV Water). Participants responded to questions as to whether they had seen their groundwater data presentedor analysed in a particular fashion and whether they thought it was novel, useful and/or relevant. Of the 10 methods presented and discussed, four were deemed to be ‘relatively routine’. All were seen to be useful and would be used if processing was relatively simple and straightforward. Participants were overwhelmingly in favour of an open-source tool(s) that could be easily (non-technically) applied for data exploration and provide some standards in groundwater-level processing, particularly for informing groundwater-flow models (auths)
