Applications are invited for a fully funded PhD scholarship to explore feedback between humans, living things and artificial intelligence (AI).
Researchers are increasingly seeking to solve complex problems using computer generated predictions. These predictions are often applied to living things such as crops for productivity, or pests or diseases for their management. However, we rarely consider how living things will further adjust in response to the changes caused in their environment by the application of the outputs from the AI. This project will develop tools to explore how management methods based on artificial intelligence can lead to unexpected consequences when applied to living things.
You will tackle these problems by integrating diverse sources of information. This could include quantitative methods, mathematical modelling, fieldwork and interviews with end users.
This scholarship is currently only open to current residents of New Zealand and Australia. We are happy to consider students from a diverse range of fields. At a minimum, some university level of mathematics familiarity with scientific computation is expected. An interest in ecology 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 a relevant discipline.
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.
The best place for you to be based during your studies is Lincoln University (near Christchurch), New Zealand, although we can be flexible on this. You will be jointly supervised by William Godsoe (Lincoln, Bioprotection Research Centre), Claire Postlethwaite (Auckland, Mathematics) and Emma Sharp (Auckland, School of Environment).
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 thought, methods, and people to define and solve society’s thorny interconnected problems.
Informal enquiries are welcome by email:
- Full tuition fees
- Stipend of NZ$28,500 per year (tax free)
Start date is flexible but would preferably be between August 2021 and March 2022.
How to apply
Send an email expressing your interest, along with a CV, academic record, and list of three potential referees to William Godsoe at firstname.lastname@example.org.
Applications will be considered until the position is filled. Applications received by 30 July 2021 will receive full consideration.
Ecologist and science leader Dr Andrea Byrom has accepted a role as kairangi in Te Pūnaha Matatini.
Kairangi is a Māori word meaning ‘the finest pounamu’, which can be used to describe a person held in high esteem. This role acknowledges the important contributions of our senior colleagues.
Dr Andrea Byrom has been involved with Te Pūnaha Matatini as an associate investigator since the early days, and has contributed at many hui and supervised several early career researchers. She is currently co-supervising Te Pūnaha Matatini Whānau member Julie Mugford in the final stages of her thesis, alongside Associate Professor Alex James and Professor Michael Plank.
The project that Andrea is most proud of being involved with at Te Pūnaha Matatini was exploring the biodiversity benefits of large-scale pest control regimes with Dr Rachelle Binny. Their work quantified significant benefits for biodiversity from pest control over two decades. Andrea says that “I’m proud to have contributed to that research because it really demonstrated how important science is to the environment, and why we do large-scale conservation efforts like pest control or ecological restoration.”
She also particularly enjoyed collaborating with Professor Shaun Hendy and a group of summer interns on network analyses of the many types of people and organisations involved in environmental protection in Aotearoa. “That was a real introduction to network analyses and some of the things Te Pūnaha Matatini had to offer that I had not previously thought of applying to te taiao the environment.”
Andrea recently resigned from her role as director of Ngā Koiora Tuku Iho New Zealand’s Biological Heritage National Science Challenge. She has been working in the New Zealand science system since joining Manaaki Whenua Landcare Research as a postdoctoral researcher in 1997.
Over two decades working at Manaaki Whenua Andrea moved away from directly doing her own research and into leadership roles, after becoming interested in how science leadership could empower scientists to do their work, rather than add more bureaucracy to their lives.
She says that she “really loved that leadership style”.
“What I liked most about being a director of a National Science Challenge was having a view across all of the amazing talent that we have in the New Zealand science system.”
Her directorial responsibilities meant that Andrea did not have as much time as she would like to devote to Te Pūnaha Matatini in recent years. “I’ve been on a separate journey from Te Pūnaha Matatini for the last wee while, so to come back in as a kairangi now is quite an honour.”
“In the last few years, my interests have broadened to thinking about how we take our Te Tiriti o Waitangi partnership role seriously as scientists, and how we bring mātauranga Māori and kaupapa Māori research methods to the fore. I worked hard to facilitate a lot of that via the National Science Challenge and ended up in a co-director role in that area with Melanie Mark-Shadbolt.”
“I feel like the tide’s turning and that people are starting to listen. But it’s really important to put different perspectives and stories out there.”
After a demanding period as a director, Andrea is focusing on spending more time with her partner, as well as doing environment consultancy work and board roles. “I’m particularly interested in how important governance is to science and the environment. That’s my new passion, and as a kairangi I would like to contribute where I can – particularly around complex environmental research.”
“I love being a sounding board for students and I love coming to hui where there are great minds contributing things that I hadn’t thought of and ideas that I’m interested in.”
Since stepping back as a director, Andrea and her partner have been making the most of their time together by killing of a large amount of lawn on their half-hectare property in mid-Canterbury and replanting it with over 5,000 native plants.
How to kill your lawn with Andrea Byrom
- Acquire large quantities of cardboard boxes and flatten them
- Lay cardboard over lawn on non-windy day
- Cover cardboard with a whole lot of mulch
- Water it all down
- Leave for two months
- Replant with native plants
Aotearoa New Zealand government communication during the 2020 lockdown
25 May 2021
This paper is under review and is currently available through First Look on SSRN to provide early access prior to publication.
Aotearoa New Zealand’s response to the COVID-19 pandemic is considered one of the best in the world. A major component of the response was the communication of public health measures. A reflexive thematic analysis of Prime Minister Jacinda Ardern’s addresses to the nation and her daily press briefings with the Director-General of Health, Ashley Bloomfield, led us to identify three key themes:
- Open, honest and straightforward communication
- Distinctive and motivational language
- Expressions of care
We argue that the messages presented in the daily briefings supported the New Zealand Government’s COVID-19 elimination strategy through building trust with the audience and framing the ‘lockdown’ as an urgent, collective and meaningful cause, mobilising New Zealanders to support public health measures.
The 2020 Prime Minister’s Science Prize has been awarded to Te Pūnaha Matatini for our contribution to Aotearoa New Zealand’s COVID-19 response.
The Prime Minister’s Science Prize is awarded for transformative science which has had a significant economic, health, social or environmental impact.
Te Pūnaha Matatini are being recognised for our work that developed a series of mathematical models, analysed data and communicated the results to inform the New Zealand Government’s world-leading response to the global pandemic.
Te Pūnaha Matatini is a Centre of Research Excellence funded by the Tertiary Education Commission and hosted by the University of Auckland. Over the past six years, Te Pūnaha Matatini has grown from the kernel of an idea into a diverse national network of over a hundred investigators and students who are tackling the interconnected and deeply interdisciplinary challenges of our time. Our values, expertise and focus on communication made us uniquely positioned to grapple with the COVID-19 pandemic in Aotearoa New Zealand.
Te Pūnaha Matatini’s modelling was key in helping the government make good decisions about lockdowns, particularly in April and May when the need to relax Alert Levels arrived, and in August, when a tailored lockdown was used in Auckland to eliminate a large outbreak. These public health interventions have had an immense impact on New Zealanders’ lives, not the least of which was preventing a considerable number of deaths due to COVID-19 if the virus had been allowed to spread unimpeded.
“Even I underestimated the centrality of [science] advice for me, in this time in office, and just how important it would become to us as a government.” – Jacinda Ardern, Prime Minister of New Zealand
The team made sure their models served the health system by working with Orion Health data scientists to ensure information got to where it was needed. Orion Health works with healthcare sector clients to deploy and manage machine learning models, which meant they were able to offer their technology and processes to support the Te Pūnaha Matatini team.
Te Pūnaha Matatini’s work and related research from around the globe was actively communicated to the public throughout 2020, and several of Te Pūnaha Matatini’s researchers were the most prominent science communicators during the crisis.
“I want to thank the many, many, many people in this room who were a part in your own ways in either helping us generate the information we needed to make those decisions, or who helped us communicate those decisions when it mattered most.” – Jacinda Ardern, Prime Minister of New Zealand
The transdisciplinary team working on COVID-19 that received this award brought together researchers from the University of Auckland, University of Canterbury, Victoria University of Wellington, Manaaki Whenua Landcare Research, Market Economics, and Orion Health.
The COVID-19 programme at Te Pūnaha Matatini continues into 2021 with projects focusing on branching process models, complex network models, phylodynamics, and the spread of disinformation and misinformation.
A modelling study
22 March 2021
- Vaccination of New Zealand’s frontline border workforce is a priority in order to protect this high-exposure group from the health impacts of COVID-19.
- Although vaccines are highly effective in preventing disease, their effectiveness in preventing transmission of COVID-19 is less certain.
- There is a danger that vaccination could prevent or reduce symptoms of COVID-19 but not prevent transmission. Counterintuitively, this means that vaccinating frontline border workers could increase the risk of a community outbreak.
- In a scenario where the vaccine reduces transmission by 50%, vaccinating border workers could increase the risk of a significant community outbreak from around 7% per seed case to around 9% per seed case.
- Until more is known about the effect of the vaccine on transmission, we recommend increasing the routine testing of vaccinated border workers to mitigate this risk. Regular saliva testing may be a good way to achieve this.
- Careful attention should be paid to any groups, such as frontline workers’ family members, who may be vaccinated but who are not undergoing routine testing to ensure they do not become asymptomatic spreaders.
Australia and New Zealand have a strategy to eliminate community transmission of COVID-19 and require overseas arrivals to quarantine in government-managed facilities at the border. In both countries, community outbreaks of COVID-19 have been sparked following infection of a border worker. This workforce is rightly being prioritised for vaccination. However, although vaccines are highly effective in preventing disease, their effectiveness in preventing transmission of COVID-19 is less certain. There is a danger that vaccination could prevent symptoms of COVID-19 but not prevent transmission. Here, we use a stochastic model of COVID-19 transmission and testing to investigate the effect that vaccination of border workers has on the risk of an outbreak in an unvaccinated community. We simulate the model starting with a single infected border worker and measure the number of people who are infected before the first case is detected by testing. We show that if a vaccine reduces transmission by 50%, vaccination of border workers increases the risk of a major outbreak from around 7% per seed case to around 9% per seed case. The lower the vaccine effectiveness against transmission, the higher the risk. The increase in risk as a result of vaccination can be mitigated by increasing the frequency of routine testing for high-exposure vaccinated groups.
4 September 2020
There is limited evidence as to how clinical outcomes of COVID-19 including fatality rates may vary by ethnicity. We aim to estimate inequities in infection fatality rates (IFR) in New Zealand by ethnicity. We combine existing demographic and health data for ethnic groups in New Zealand with international data on COVID-19 IFR for different age groups. We adjust age-specific IFRs for differences in unmet healthcare need, and comorbidities by ethnicity. We also adjust for life expectancy reflecting evidence that COVID-19 amplifies the existing mortality risk of different groups.
The IFR for Māori is estimated to be 50% higher than that of non-Māori, and could be even higher depending on the relative contributions of age and underlying health conditions to mortality risk. There are likely to be significant inequities in the health burden from COVID-19 in New Zealand by ethnicity. These will be exacerbated by racism within the healthcare system and other inequities not reflected in official data. Highest risk communities include those with elderly populations, and Māori and Pacific communities. These factors should be included in future disease incidence and impact modelling.
NEW RESEARCH — LINK TO FULL PDF
22 February 2021
Mathematical modelling to inform New Zealand’s COVID-19 response
Between February and May 2020, New Zealand recorded 1504 cases of Covid-19 before eliminating community transmission of the virus in June 2020. During this period, a series of control measures were used including population-wide interventions implemented via a four-level alert system, border restrictions, and a test, trace, and isolate system.
Mathematical modelling played a key role in informing the government response and guiding policy development. In this paper, we describe the development of a stochastic mathematical model for the transmission and control of Covid-19 in New Zealand. This includes features such as superspreading, case under-ascertainment, testing and reporting delays, and population-wide and case-targeted control measures.
We show how the model was calibrated to New Zealand and international data. We describe how the model was used to compare the effects of various interventions in reducing spread of the virus and to estimate the probability of elimination. We conclude with a discussion of the policy-modelling interface and preparedness for future epidemic outbreaks.
NEW RESEARCH — LINK TO FULL PDF
8 December 2020
Modelling support for the continued elimination strategy
- We model the effects on the risk of COVID-19 border reincursions of a wide variety of different border policies, including changes in managed isolation requirements for travellers as well as different testing regimes for frontline border workers.
- A more detailed modelling study and risk analysis of a specific policy change would be recommended before any implementation.
- One potential change in policy that could be considered is to replace the current requirement for 14 days in MIQ with 7 days in MIQ followed by 7 days in home isolation (including a second PCR test) for arrivals from countries with low prevalence of COVID-19 such as Australia.
- However, any increase in the number of arrivals from high-prevalence countries, for example due to an increase in MIQ capacity or repurposing of existing MIQ capacity, will lead to an increase in the risk of border reincursions.
- Weekly PCR testing of frontline border workers helps to ensure most border reincursions are detected before they grow too large. Supplementing this with an additional weekly rapid test would be an extra safeguard that decreases the risk of a large outbreak.
NEW RESEARCH — LINK TO FULL PDF
9 November 2020
Early intervention is the key to success in COVID-19 control
- Evaluating the effectiveness of New Zealand’s COVID-19 response, relative to counterfactual (alternative ‘what-if’) scenarios, is important for guiding future response strategies. We assess the importance of early implementation of interventions for controlling COVID-19.
- We model counterfactual scenarios in which the timings of three policy interventions are varied: border restrictions requiring 14-day quarantine of all international arrivals, border closure except to returning residents and citizens, and Alert Level 4 restrictions. We compare these to a modelled factual scenario in which intervention timings are the same as occurred in reality.
- Key measures describing the dynamics of a COVID-19 outbreak (notably peak load on the contact tracing system, the total number of reported COVID-19 cases and deaths, and the probability of elimination within a specified time frame), are used to compare outcomes between scenarios.
- Key measures were more sensitive to the timing of Alert Level 4, than to timing of border restrictions and border closure. Of the counterfactual scenarios, an earlier start to Alert Level 4 would have resulted in the greatest reduction in numbers of cases and deaths.
- Delaying the start of Alert Level 4 by 20 days could have led to over 11,500 cases and 200 deaths, and would have substantially reduced the probability of eliminating community transmission of COVID-19, requring a longer period at Alert Level 4 to achieve control.
NEW RESEARCH — LINK TO FULL PDF
21 October 2020
Economic comparison of the use of Alert Levels 3 and 4 in eliminating the Auckland August outbreak: a cost-effectiveness analysis
- We compare the economic costs of containing the Auckland August outbreak of COVID-19 using Alert Level 3 to those that might have been incurred from the use of Alert Level 4.
- We estimate the effectiveness of Alert Level 3 using data from the actual August outbreak. The effectiveness of a putative regional Alert Level 4 is less certain, but we consider an optimistic estimate based on what was achieved in the March-April outbreak, as well as a more pessimistic estimate, which reflects the higher transmission rates observed in August.
- We use a decision-making model for de-escalation of alert levels based on observations of weekly case numbers, which is a simpler decision-making criterion to that used in New Zealand and likely underestimates the duration of Alert Level 4 periods that would be used in practise.
- To achieve the same likelihood of elimination, we find that both the optimistic and pessimistic Alert Level 4 period has a shorter duration than the period needed at Level 3.
- To achieve the same likelihood of elimination, the optimistic Alert Level 4 controls have a lower economic cost than the Alert Level 3 controls.
- To achieve the same likelihood of elimination, the pessimistic Alert Level 4 controls come at a comparable economic cost to the Alert Level 3 controls.
- This analysis does not take into account the longer term economic costs of these measures, nor does it consider social, or health impacts that might differ between strategies.