A Back to Basics Approach to Enable Earth and Environmental Science Data and Information Services to be FAIR and Key Players in Modern Transdisciplinary Research and Applications

Abstract:

Earth and environmental science data have so much to contribute to help unravel global crises such as COVID-19, unprecedented wildfire episodes and increased frequency of extreme weather events. Effectively predicting and managing these emergency situations requires faster than real time integration of FAIR data at their highest resolution and at scale. Live data sources need to ‘born FAIR’ and machine actionable so that computational systems can find, access, interoperate and reuse data with no or minimal human intervention.

Published in 2016 the FAIR principles do not enunciate the exact specifications required to make data machine actionable. FAIR Implementation Profiles (FIPs) are now emerging to help create more specific ‘community’ profiles to improve interoperability and reusability. But who is the ‘community’? How many ‘community’ profiles do we need?

To make data FAIR, Sustkova et al (2020) noted there are both generic aspects that can be shared across multiple communities (eg., licensing, identifiers) and domain-specific aspects (eg., metadata profiles; vocabs). Within the domain specific aspects there are common protocols that apply to multiple scientific domains (eg., units of measure; how location is specified) and there are attributes unique to the fundamental branches of the physical sciences (eg., the periodic table used in chemistry, biochemistry, geochemistry, hydrochemistry data; common descriptions of physical specimens). We need the various expert domains to specify FIPs for these basic scientific building blocks, possibly in collaboration with the International Science Unions and other recognised authoritative groups. Vocabularies also need to be better harmonised, and where possible be multilingual.

But the multiple crises of 2020 show there is also an urgent requirement for integration of any data across the physical sciences, social sciences, health, infrastructure, etc, regardless of whether it is derived from the research, industry, government or private sectors. We need international collaboration, both within and across all groups that make data FAIR and machine actionable in any domain, sector or country. But where will this collaboration occur? Will the 17 UN Sustainable development goals provide the impetus for harmonised FIPs across multiple data types and sectors?