Out and about

Modelling for transport policy interventions

Modelling for transport policy interventions

19 October 2021

Julie Mugford is applying the skills that she learned through her doctoral project with Te Pūnaha Matatini to a career across the public service.

As she entered the final year of her PhD in 2020, Julie Mugford spotted an advertisement for a role with the Ministry of Transport.

Julie had just completed an internship with the Ministry of Social Development, to see whether she enjoyed working in the public sector. She had found it a good fit, and when she saw a job available with ‘agent-based modelling’ in the description, she jumped at the chance.

From May 2020 to September 2021, Julie worked as a data analyst on the team that is creating an agent-based model of the Aotearoa New Zealand transport system.

“It has the whole New Zealand transport network,” explains Julie. “All the roads and public transport options, cycleways and footpaths. Then it has a population of typical New Zealanders with their activities that they want to do during the day, such as going to school or to work.”

“They decide what time of the day they’re going to go and what mode they’re going to use. The aim of the model is to be able to change policy settings – for example, road pricing – and see what affect it has on transport behaviour and assessing the social and environmental impacts of such changes.”

This is new territory for transport policy in Aotearoa New Zealand. Small detailed modelling tasks and larger scale aggregated modelling are already in use, but Julie is not aware of any agent-based modelling being used here. She notes that this approach has already been successfully applied in London, Melbourne, and some European countries.

Julie’s PhD project in applied mathematics looked at citizen science, where members of the public help scientists gather and analyse various forms of information. The central question of her project was whether scientists could get useful insight out of noisy, biased citizen science data.

She worked with her Te Pūnaha Matatini supervisors to develop methods to improve the reliability of data collected or analysed by members of the public through platforms like iNaturalist.

Being involved in Te Pūnaha Matatini and Te Pūnaha Matatini Whānau was the highlight of Julie’s PhD experience.

“Te Pūnaha Matatini is full of so many amazing people,”she says. “It was a welcoming environment, and the interdisciplinary focus of Te Pūnaha Matatini added more depth to my PhD than I could have ever imagined when I signed up to a PhD at the University of Canterbury School of Mathematics and Statistics.”

She says that there was always a lot of variation in the speakers and activities that the Whānau organised, and serving as the chair was a great opportunity to develop leadership experience while studying.

Julie is enjoying working in the public sector and plans to explore working at different ministries over time. She recently continued this journey by accepting a role as a Senior Analyst at the Ministry of Health.

“I like working in the public sector. It’s really great how there are these big problems that I can work on and they actually require all the experience that I’ve gathered from my education.”


Estimates of new onset of COVID-like illness, symptomatic population, and testing rates

18 October 2021

Using the FluTracking survey data for 2021 up until the week ending 29 August 2021, we can estimate the weekly incidence (new onset) of COVID-19-like symptoms within different age groups and different parts of the country. We use this to produce an estimate of the newly symptomatic population, and combine this with current testing data (up until 1 September 2021) to estimate how many people with new onset of symptoms (in the past seven days) would have been tested for COVID-19 through time.

  • Weekly incidence rates for any two or more COVID-19-like symptoms (CLI2+) in the week ending 29 August, 2021 was 6.7% for under 5s, 3.1% for 5-19 year olds, 3.3% for 20-64 year olds, 1.9% for 65+.
  • Weekly incidence rates for any one or more COVID-19-like symptoms (CLI1+) in the week ending 29 August, 2021 was 7.1% for under 5s, 4.8% for 5-19 year olds, 5.7% for 20-64 year olds, 3% for 65+.
  • New onset of illness, especially among pre-school and school age children has seen a large drop. This matches the declines seen in earlier Alert Level changes when schools were closed (Alert Level 3 and 4).
  • Despite the drop in testing seen in many regions, the lower incidence of illness still produces high symptomatic testing rate estimates in Auckland Region DHBs across all age bands except under 5s. A number of the approaches used to calculate this find testing rates up to 100%.

These estimates have a number of caveats. First, we are assuming that the weekly incidence (new onset of symptoms) estimated from the FluTracking survey is representative across a number of different communities. Secondly, we assume that all tests without an entry in the National Contact Tracing System (NCTS) database are seeking a test due to COVID-like symptoms. Finally, we assume that people seeking tests for symptoms would have had new onset of symptoms within the past seven days (although we test the sensitivity to this last assumption in section 3.2.2). In future it would be worth using more detailed information on the reason for seeking a test, including the symptoms people seek tests for, and the time from symptom onset to seeking a test. Unfortunately this data is not available yet.

Estimates of effects on changing Alert Levels for the August 2021 outbreak

18 October 2021

Report delivered 9 September 2021, this version compiled 15 October 2021.

We use an individual-based, Aotearoa-specific Contagion Network Model (CNM) to simulate the spread of COVID-19 in the community for an outbreak comparable to that detected on 17 August 2021. The CNM explicitly simulates spreading processes and the effect of any interventions, in order to predict the spread of COVID-19. It is therefore able to avoid the need for input assumptions or external estimates of the effective reproduction number (Reff) during the outbreak.

We use this model to consider two scenarios:

  1. An intervention with parameters comparable to Alert Level 4 (AL4) is applied from 18 August 2021 onwards and remains in place for the duration of the simulations.
  2. An intervention with parameters comparable to AL4 is applied from 18 August to 15 September 2021, after which the intervention parameters are changed to match those expected for Alert Level 3 (AL3).

We complement the CNM simulations with results from a Branching Process Model (BPM). We use this to simulate scenarios where an Alert Level change on 16 September 2021 results in a multiplicative increase of Reff of factors of 1, 1.5, 2, or 3 times that estimated during AL4.

After 60 days at AL4, it is predicted that almost all trajectories will be either eliminated or contained (i.e. all cases are either in quarantine or isolation). In contrast, only a few trajectories are eliminated or contained for the optimistic and pessimistic AL3 scenarios. In the former, many trajectories exhibit suppression like behaviour, remaining at low numbers (under 20 new cases per day) or growing only slowly.

The pessimistic AL3 scenario shows a growing number of daily new cases for the majority of trajectories. The pessimistic AL3 scenario also suggests that it would initially be difficult to distinguish between uncontrolled growth and suppression or elimination like behaviour – it takes approximately another 10 days post-deescalation before the growing trajectories start to become distinguishable.


Making a global impact in predicting and preventing pandemics

Making a global impact in predicting and preventing pandemics

11 October 2021

Professor David Hayman made a global impact in 2020 with his contributions to the report on biodiversity and pandemics by the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES).

David Hayman is an epidemiologist and Principal Investigator at Te Pūnaha Matatini who uses multidisciplinary approaches to address how infectious diseases are maintained within their hosts and how the process of emergence occurs.

Dave has spent a long time working on emerging infectious diseases and bats, making him a natural candidate for Aotearoa New Zealand to put forward when the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) put out a call for nominations for an expert panel to produce a report on the interactions between biodiversity and human drivers of disease emergence.

The report on escaping the ‘era of pandemics’ was produced at pace during a week-long virtual workshop to review the scientific evidence on the origin, emergence and impact of COVID-19 and other pandemics, as well as on options for controlling and preventing pandemics.

Dave now sits on the One Health High Level Expert Panel (OHHLEP), a high-level expert panel that gives advice across four international agencies: the World Health Organization, the World Organization for Animal Health, the Food and Agriculture Organisation and the United Nations Environmental Programme.

He says that the IPBES report has been influential across these agencies, and is often referred to. “We are a high level expert panel that provides expertise and advice to these major global organisations about how they can work better together,” he says. “And I think the IPBES report has actually influenced that.”

The IPBES report has been an important step in these four agencies coming to terms with the complexity and interrelatedness of disease and the environment, and they are recognising the need to address these issues in a transdisciplinary way.

“There’s a lot of things from Te Pūnaha Matatini and working in Aotearoa New Zealand that influenced my contributions to the report. There’s lots I’ve Iearned from Te Pūnaha Matatini about style of working and things like respect for Māori and Indigenous knowledge.”

Dave describes himself as both a pessimist and an optimist as we face a future of increasing pandemics and the effects of climate change.

“It can seem all bad,” he says. “But on the plus side a lot of the drivers for climate change, biodiversity crises and extinction crises are the same as the things that are driving disease emergence. So we can potentially have win, win solutions.”

“We can look at things like reducing industrial-scale trafficking of wildlife or agricultural encroachment into rainforest, both of which are bad for the environment and may also be bad for human health.”

“You can potentially reduce one risk and improve things in another way.”

Dave concludes that tackling these issues will require quite big societal changes, but “what COVID-19 did show is that you can do large-scale stuff. You can shut down whole countries. I’m not saying that’s a good thing, but it showed us the scale and pace at which societies can change and adapt.”


A biosecurity risk framework for forestry in Aotearoa New Zealand

A biosecurity risk framework for forestry in Aotearoa New Zealand

8 October 2021

In her role with Scion, Dr Rebecca Turner is working with stakeholders in Aotearoa New Zealand and internationally using data to predict biosecurity risk.

Dr Rebecca Turner joined Scion as a postgraduate fellow in 2018, co-funded by the Biological Heritage National Science Challenge and Te Pūnaha Matatini. In 2020 she was promoted to a full-time biosecurity scientist.

Scion is a Crown research institute that specialises in research, science and technology development for the forestry, wood product, wood-derived materials, and other biomaterial sectors.

“My postdoc led directly into this role,” says Rebecca. “It created the opportunity for me to get to know Scion systems, plus New Zealand researchers and international collaborators through Te Pūnaha Matatini.

“Crown research institutes focus on applied science for the sectors they serve, so publishing reports for industry use is important. Having the postdoc and the TPM funding helped me build academic credibility by publishing papers.”

Rebecca’s background is in ecology, molecular biology and mathematical modelling. She is interested in research using mathematical techniques to understand biology and other applications.

Rebecca was involved with Te Pūnaha Matatini Whānau during her postdoctoral fellowship, and appreciated building her network at our Annual Hui each year. She also contributed to our Mycoplasma bovis response, working with data from the National Animal Identification and Tracing (NAIT) system that tracks cattle movement around New Zealand.

At Scion, Rebecca’s initial project explored the potential of using border interception data to predict arrivals and establishments of invasive pests in New Zealand. She says that the team hopes to be able to use border interception data to warn people what invasive species to look out for in orchards and forests.

The project quickly became complicated and grew to include international interception data, and Rebecca is now working with the United States Forest Service on an extension of this project looking at data about beetles, which include large groups of potential forestry pests.

“Although we’re really good at biosecurity in New Zealand and we’ve got a really good rate of interceptions and getting them down to species level relative to our population size, we’re still a small country, and we can only collect a certain amount of data. So we then started working with international stakeholders to get interception data from other countries as well.”

Three years in, they are now in a place where they have all the data, have started analysing it, and are starting to see where some of that data is useful for predicting establishments.

“In New Zealand forestry the major plant species that we have is Pinus radiata, so we’re looking for insects that are associated with Pinus radiata, and trying to predict which are going to establish in New Zealand, using things like interception data and climate matching.”

“We’re creating a biosecurity risk framework specifically for the forestry industry. I’m also collaborating with AgResearch and Plant & Food Research through Better Border Biosecurity (B3) to create frameworks for the agricultural industry and the pasture industry.”

Find out more about Rebecca

PhD scholarship on past human-environment interactions on Ahuahu Great Mercury Island

Applications are invited for a fully funded PhD scholarship to examine past human-environment interactions on Ahuahu Great Mercury Island, Coromandel, Aotearoa.

Indigenous societies have often developed unique ways of sustaining ecological complexity. In Aotearoa, Polynesians arrived in the 13th century with a set of commensal plants and animals and a view of the natural world adapted to small, tropical and sub-tropical islands. Guided by foundational Polynesian principles, Māori learned to live in Aotearoa’s fundamentally different environment, developing unique kaitiakitanga (values, principles, and practices of guardianship) and tikanga (customary practices).

Our transdisciplinary project is investigating how Māori drew on the knowledge of the founding Polynesian ancestors and developed unique perspectives and practices in response to intimate experiences in and interactions with the Ahuahu Great Mercury Island landscape. Horticultural development and processes of niche construction are foci of the project. Palaeo-environmental and archaeological data will be integrated with modelling of Māori tikanga and kaitiakitanga and social-ecological systems.

As part of our research team, you will contribute to understanding complex human-environment interactions and the emergence of organizational strategies associated with sustainability and resiliency.


This scholarship is open to anyone who can be in New Zealand and meets the requirements to enrol in a PhD at the relevant institution.

Nau mai, haere mai! Applicants from a diverse range of fields are welcome, with experience in any one or more of the following advantageous: Mātauranga Māori; archaeological ecodynamics; geoarchaeology; paleoenvironmental reconstruction; and analytical and statistical approaches used to identify the nature and causes of environmental change.

The successful candidate will hold, or expect to complete soon, an honours or masters degree, or similar, in a relevant discipline, including anthropology, archaeology, environmental science, geography, Māori studies, or mathematical modelling.

We encourage applicants from all backgrounds to apply and we especially encourage applications from Māori and Pasifika students. Our research group aims to achieve work-life balance within a productive scientific environment.


Ideally you will be based in Auckland at the University of Auckland, although the University of Canterbury in Christchurch or the University of Otago in Dunedin may also be options. You will be jointly supervised by Professor Thegn Ladefoged (University of Auckland), and depending on your skills and focus, Dr Matiu Prebble (University of Canterbury), Dr Rebecca Phillipps (University of Auckland), Professor Melinda Allen (University of Auckland), Associate Professor Priscilla Wehi (Otago University), Associate Professor Paul Augustinus (University of Auckland), and Dr Dion O’Neale (University of Auckland).

You will be part of Te Pūnaha Matatini, the Aotearoa New Zealand Centre of Research Excellence for Complex Systems. Te Pūnaha Matatini brings together different disciplines, ways of thinking, methods, and people to define and solve society’s thorny interconnected problems.

Te Pūnaha Matatini has an active whānau group which supports early career researchers, committed to the Te Pūnaha Matatini values of manaakitanga and whakawhanaungatanga, offering supportive tuakana / teina learning environments.


Please get in touch if you have pātai about this position.

Financial details

  • Full tuition fees
  • Stipend of NZ$28,500 per year (tax free)

Start date

Start date is flexible but would preferably be between January and June 2022.

How to apply

Interested candidates should send an email expressing their interest, along with a CV, academic record, and list of three potential referees to Thegn Ladefoged at t.ladefoged@auckland.ac.nz.

Due date

Applications will be considered until the position is filled. Applications received by Wednesday 1 December will receive full consideration.

PhD scholarship on modelling spreading processes on real-world networks

Applications are invited for a fully-funded PhD studentship to work on a project modelling and understanding spreading processes on multilayer and multiplex networks.

There is growing understanding that spreading processes on real-world networks are typically moderated or influenced by additional factors, which themselves occur on networks and with feedback loops between the processes on the two networks. These dual spreading processes can occur either across a single node set with multiple edge types (multiplex networks), or multiple distinct node sets, with different edge types within and between the node sets (multilayer networks).

This project aims to understand how the outcomes of spreading processes are affected by the multilayer and multiplex network structures and by different network topologies arising in different applications, for example, by considering how behavioural dynamics can affect contagion in networks of epidemic spread.

You will combine mathematical modelling and dynamical systems methods with practical applications using concrete data to understand the role of social factors in epidemic spread and to investigate the dynamics of these processes.


This scholarship is open to anyone who can be in New Zealand and meets the requirements to enrol in a PhD at the relevant institution.

We are happy to consider students from a diverse range of fields including mathematics, engineering, physics, computer science and statistics. At a minimum, some university-level mathematics is expected, as is some level of familiarity with scientific computation. An interest in epidemiology will be an advantage for this project, but no prior knowledge is required. The successful candidate will hold, or expect to complete soon, a masters degree, or similar, in their disciplinary area.

Applicants from all backgrounds are actively encouraged to apply. Members of underrepresented groups are very welcome, as are students with families. Our research group aims to achieve work-life balance within a productive scientific environment.


You will be based at the University of Auckland and will be jointly supervised by Dr Dion O’Neale (Department of Physics) and Associate Professor Claire Postlethwaite (Department of Mathematics).

The PhD position will be embedded within Te Pūnaha Matatini, the Aotearoa New Zealand Centre of Research Excellence for Complex Systems. Te Pūnaha Matatini brings together ‘many faces’ – different disciplines, ways of thought, methods, and crucially, people – to define, and then solve, society’s thorny interconnected problems.

The expertise of our researchers spans the breadth of human knowledge, from computational sciences to environmental economics, and from linguistics to Indigenous philosophy to mathematical biology. This deeply transdisciplinary approach characterises Te Pūnaha Matatini and is unique within the New Zealand research system; it carries methods, approaches, and tools over from one discipline to another, and in doing so, develops integrated and transformative insights.


Informal enquiries are welcome by email:

Financial details

  • Full tuition fees
  • Stipend of NZ$28,500 per year (tax free)

Start date

Start date is flexible but would preferably be between March and June 2022.

How to apply

Interested candidates should send an email expressing their interest, along with a CV, academic record, and list of three potential referees to Dion O’Neale at d.oneale@auckland.ac.nz.

Due date

Applications will be considered until the position is filled. Applications received by 15 December 2021 will receive full consideration.

Modelling to support a future COVID-19 strategy for Aotearoa New Zealand

23 September 2021

Updated modelling exploring how high rates of vaccine coverage might reduce the health burden from COVID-19 if combined with moderate public health measures to reduce transmission of the virus.

Aotearoa New Zealand is on track to vaccinate upwards of 80% of those aged over 12 against COVID-19 with the Pfizer/BioNTech mRNA vaccine. Recent announcements by Pfizer and BioNTech suggest that the vaccine may soon be approved for use in children aged 5-11 years. This means it is possible that Aotearoa could achieve vaccine coverage across the total population of more than 90%.

We consider how these high rates of coverage might reduce the health burden from COVID-19 if combined with moderate public health measures to reduce transmission of the virus. Scenarios are evaluated using the Te Pūnaha Matatini vaccine model, using current data about vaccine effectiveness with respect to the Delta variant. As the effects of the vaccine on transmission remain uncertain, we consider three levels of vaccine effectiveness: high, central, and low to illustrate a range of possibilities.

The results suggest that a combination of high levels of vaccination within the community, a strong test-trace-isolate-quarantine system (assuming case numbers are kept sufficiently low) and moderate public health measures may be enough to attain population immunity, greatly reducing the need for strong public health measures, such as stay-at-home orders and workplace closures.


FluTracking incidence calculation methods

20 September 2021

This report details the methods used for calculating the estimated weekly incidence of COVID-19-like and influenza-like illness in Aotearoa New Zealand, using data from the FluTracking weekly survey.

FluTracking is a participatory health surveillance system for Australia and New Zealand that uses an online, voluntary survey to detect the potential spread of influenza, and more recently COVID-19. Volunteer participants from the general public receive a weekly email prompting them to fill out a survey which asks whether they have experienced any cold, flu, or COVID-19-like symptoms in the previous week.

FluTracking data is an incredibly useful resource for identifying patterns in symptom onset across the New Zealand population. It allows us to report on how symptom onset differs over time and between seasons, as well as across different regions in Aotearoa. However, there are limitations in this data that can reduce the utility of some results.

This report details a methodology that can be used to mitigate some of the limitations in the FluTracking data, and provides case definitions that can be used to inform responses to outbreaks of infectious disease, such as COVID-19. This report also includes a brief summary of FluTracking estimates for May 2020 to April 2021 as an example of applying this new methodology.

Contagion network modelling in the first weeks of the August 2021 outbreak

In the first weeks of the 2021 Auckland August COVID-19 outbreak, the contagion network team provided a number of reports to officials with estimates of the likely size of the outbreak and updates on the estimated effect of Alert Level interventions on curtailing spread.

The report ‘Modelling Estimates of Expected Size: the August 2021 COVID-19 Outbreak in Aotearoa’ describes the modelling approaches used in these estimates, including both a Contagion Network Model and a Branching Process Model. Output from both of these models was conditioned to real-world case data as it became available, using a Gaussian Process Model.

The reports ‘Preliminary modelling of a new community case of COVID-19 as of 17 August 2021’ and ‘Impact of wastewater testing results on preliminary estimates of COVID-19 community transmission as of 18 August 2021’ were compiled in the first two days of the outbreak when very little was known about the details of the outbreak. The first of these was produced when all that was known was that a single case of COVID-19 had been discovered in the community, linked to a man living in Devonport. The second included consideration of the information of a negative wastewater test from the weeks before the detection of the first case. Both of these reports significantly underestimate the size of the outbreak at detection, though both did predict that the outbreak was likely to be significantly larger than was known at the time of writing.

As the outbreak progressed, daily updates were produced to incorporate the effect of new cases discovered over the previous 24 hour period. These semi-automated daily updates used the same methods as described in the report of 26th August, with both a Contagion Network Model and a Branching Process Model treated with Gaussian Process conditioning. As the outbreak progressed, and new daily cases decreased, detailed information from contact tracing became relatively more important for estimating near-term risk and outbreak progression. Hence, the example shown here from 7th September is the last daily update provided for the initial phase of the outbreak.

Read the reports