Manifesto (work in progress)

(This page is being discussed on this issue)

Industrial ecology (IE) research relies upon data describing interactions among natural and anthropogenic systems. Such data may be physical (e.g., mass, energy, monetary) or non-physical (e.g., education quality, happiness, monetary), may be complex (e.g., the global supply chain of a product) or simple (e.g., the sale of a car), and may be qualitative or quantitative. This diversity in what these data are and how these data are, makes transparency and reproducibility in IE research challenging. Significant data proprietary/confidentiality issues, and the increasing burden being placed on the field as more data and research gets published and accumulated, has added to this challenge.

This IE Open Science Manifesto has been written in an attempt to address these issues, attempting to represent the whole IE community. It represents ground-roots interest in advancing Open Science within the IE community, and is also somewhat of a response to the 'call to action' Forum piece published in the Journal of Industrial Ecology (JIE) in early 2018 [1]. Therefore, this Manifesto lays out a preliminary vision for advancing Open Science within IE research and an agenda to realise more reproducible, transparent, and impactful research.

This is a community effort. If you're interested in Open Science within IE, join us!

Agenda

The IE Open Science community identifies the following issues in publishing IE research. We propose open working groups on each of these topics

IE data inventory/catalogue

Often times IE researchers require similar data. A central IE data inventory would facilitate and also harmonize the use of data, such as scenario assumptions, etc. Furthermore, directly feeding research results into a community database would facilitate reproducibility, data reuse. As an added benefit the work of researchers is promoted by making reuse more likely.

The IE data inventory is a curated data catalogue, more than a mere collection of links to datasources, but far away from being a proper database. Both machine-readable and non-machine readable data can be inventorized, a common data structure and classification is not used.

This topic is closely related to the ontology group, as a harmonized database requires harmonized data structures.

Group members: Stefan, Rupert, Niko, Rick

Challenges/Working items

What format? Google spreadsheet? Should be searchable, open/easily accessible to entire community, descriptive. We need a better version of [2].

Ontology

In order to harmonize IE research data, it is necessary to harmonize IE data structures. The main challenges come from the fact that IE uses different methods (MFA, LCA, etc.) and different kinds of data (monetary, material, impacts, etc.). The data model (e.g. ontology) could also facilitate model / method integration, like coupling LCA and IO.

More details on the IE Ontology page

Group members: Chris, Brandon, Rick

Challenges/Working items

Ontology vs relational DB vs a combination of both - which one is best/feasible for IE data? To clarify differences, see [3] or [4]. Elaborate on reasoning for MySQL or triple store in relation to needs with respect to storing/exchanging IE data.
Look at major existing IE databases (such as ecoinvent, EXIOBASE, and others inventoried) and their schemas as a starting point for the data model development
Ontology development for IE. Start with system description --> end with something like HS codes [5]? How light/heavy do we want to develop the ontology? Glossary or...? What is appropriate?
Attempts at structuring data and concepts... (add yours) [6] Unified Material Information System (UMIS) - material stocks and flows data; .

IE Community database / IE data commons

Should there be a central/super 'IE community database'? Should there be separate databases that are interoperable, supported by an IE ontology or relational database structure? What type of database should it be? Who should host it? How will it be funded and maintained? Should there be an information system instead or as a predecessor that, thanks to the data harmonization, allows to query and extract data from the distributed, autonomous databases inventoried)?

Vision: Based on the IE ontology, the data from the IE data inventory can be parsed and formatted into a common database available for a wide array of search and download options. The database can have the format of a multi-layer supply and use table, depicting the socioeconomic network with balances for the different chemical elements, energy, and money. The SUT can be hierarchical, meaning that it allows for simultaneous accounting of processes and sectors at different aggregation levels. The SUT contains appendices such as: environmental and social extensions, capacity data, lifetime data, material composition data, etc. From the general SUT the different method-specific databases can be derived. For example, non-allocated and allocated unit process dataset can be determined and linked via a system model/construct to form a process life cycle database. MFA process parameters can be determined, for example, by collecting all plastic-containing flows across processes and calculating yield coefficients from them.

Group members: Tbd.

Challenges/Working items

Allow for recording incomplete data
Distinguish between measured data, reconciled data, and proxy or representative data
Find ways to anonymize confidential data or data that would allow to infer the production recipe of individual companies
Link database with other field
Link database with ontology and write reusable and replicable parsing routines.

Badge candidate review

Check Vienna CE paper as paper badge candidate. Other suggestions welcome and can always keep an eye out if reviewing any.

IE Open Science Wiki

This wiki should become the go-to place for IE researchers that are interested in Open Science.

Group members: (Niko)

Challenges/Working items

Open data for sustainability Wiki on GitHub? Should we make it somewhere more 'friendly'?
Collect links and provide guidelines on open science publishing
Collect data templates, etc.