Home
Forgot your password?
Ambitious ILTER workshop on the potential for a Global Environmental Research Infrastructure
|
|
The International Long Term Ecological Research Network, ILTER, invited us to attend a ground-breaking workshop during the week of 16 April, hosted by the South China Botanical Gardens on behalf of the Chinese Ecosystems Research Network and funded by the Chinese Academy of Science.
Other attendees were from TERENO, ICOS, SMEAR, NEON, CERN, TERN and the journal Nature. Workshop participants were treated to a full-day field trip to the Dingushan Forest Research Station, a primary LTER site of CERN, followed by a traditional Cantonese meal.
Workshop rationale
The rationale was to develop conceptual inputs to next-generation global ecological/ecosystem/critical zone research infrastructure planning and implementation with reference to the Group of Senior Officers (GSO 2014) recommendations and progress report (GSO 2017). Importantly, South Africa’s Department of Science and Technology is a member and has hosted and chaired meetings of this organisation.
The GSO proactively works to identify opportunities for international collaboration among research infrastructures (RI) proposed by its members. This workshop resulted from a recent paper about the history and future of the ILTER co-authored by SAEON’s Wim Hugo and Johan Pauw (Mirtl et al., 2018), which mooted the future of ILTER as a global environmental research infrastructure. The focus of the workshop was on highly instrumented sites similar to the SMCRI and EFTEON SARIR projects.
Given that the GSO-GRI Framework (GSO 2014) includes only four environmental/ ecological global RIs, the invited workshop attendees were identified to lead the development of a Global Environmental RI and are from site-based networks and RIs dedicated to better understanding the functioning of indicator/benchmark ecosystems across global biomes.
The primary purpose was to generate conceptual inputs to next-generation global ecological/ecosystem/critical zone RIs planning and implementation (focus on highly instrumented sites) with reference to Group of Senior Officers recommendations. A secondary workshop purpose followed from the fact that many RIs and networks are facing similar challenges and fundamental issues where they could learn from each other, irrespective of the phase they are in, but for which time is never found to discuss.
Discussions and conclusions
We are reporting here on some key results from the workshop. With respect to the operational and management strategies of existing RIs, consensus was reached that:
- Universities in general, with the exception of a few, do not make good partners to run and manage long-term RIs.
- Science at the sites can be done by both RI staff and university scientists.
- Ownership of the sites is complex due to varied land ownership and funding entities contributing to the site (mixed ownership). Ownership and distribution of data become more complex in these cases.
- RIs need to facilitate science through the provision of platforms and high-quality data and it is important to provide model/analyses-ready data.
- RIs are at a critical point and a key challenge will be to find a middle ground between service and science. Several RIs are not mandated to provide a service.
A discussion on the tasks and job profiles of a global RI was introduced by SAEON Managing Director Johan Pauw, who stressed that globally recognised scientists cannot be expected to run RIs but should be involved in all aspects of a global RI. Science advisory councils can provide strategic advice, but in-house expertise will still be required.
It is a different category of scientist that works at RIs. They have an interest in and understanding of the science, but also have managerial skills. RI scientists can moreover provide advice as to the direction that the science should take because of the interest and understanding of the RI and the ecosystem in which it is based. There must be a certain level of flexibility in answering scientific research questions, but consistent monitoring must be predominant.
In the structure of a global RI the scientist in charge acts as a PO (Principal Operator) rather than a PI (Principal Investigator) and must have an even-handed approach. After losing their good scientists, NEON implemented a matrix system whereby scientists are provided an opportunity to conduct good science.
It is therefore important to obtain staff that are highly skilled but not ambitious to obtain excellent h-indexes. RIs should not focus on the end of the science pipeline, but on the start of the pipeline. Young scientists would find work as big data analysts profitable as it results in highly rated papers and they can easily find jobs elsewhere should they want to move on.
Technicians, instrument operators and data operators are essential for the successful functioning of an RI. The dilemma is that the RI feeds data to the big data scientists and external research scientists, who then receive all the kudos. One way forward is to assign Digital Object Identifiers (DOIs) to data sets so that RI scientists and technicians may also be recognised for their contribution to the science. There should be opportunity and an expectation that RI scientists publish the small- or field-scale research results.
RIs must have or at least interact with a modelling interface centre. The scientists and data managers at such a centre must know both the data production side and the requirements of the models. Model feedback must be used to improve the RI’s monitoring programmes.
|
Way forward
Environmental RIs concentrate their science at single sites; often called supersites or master sites. The advantages of this approach are in the reduced cost for basic operations, opportunities for sharing and training personnel, common metadata, smooth interoperability of measurements and data operations, increased number of users and citations, and the additional value (e.g. understanding of feedback and feedforward ecosystem drivers and responses) derived from integrated measurements.
The workshop made the case for some 10 ILTER sites to form a global environmental research infrastructure. It was concluded that an LTER perspective would improve systems ecology by identifying the relevant processes and determining thresholds and extremes, will reduce source of model uncertainty, reduce uncertainties regarding the boundaries of the observed system, reduce uncertainty regarding interactions between observed variables and errors of abstraction and reduce uncertainties regarding calibration.
Subsequently the workshop participants have started working to formalise the proposal for an LTER-based global research infrastructure by way of a published statement in a highly-rated scientific journal.
