Development of interoperable web applications for paleoclimate research

Studying how the Earth’s climate changed in the past requires a joint interdisciplinary effort of scientists from different scientific domains. Paleoclimatic records are increasingly obtained on multiple archives (e.g. marine and terrestrial sediments, ice cores, speleothems, corals) and they document past changes in various climatic variables of the different components of the climatic system (e.g. ocean, atmosphere, vegetation, ice). Most paleoclimatic records still rely on independent observations with no standard format describing their data or metadata, resulting in a progressive increase of variables and taxonomies. Therefore, despite the achievements of the last decades (e.g. NOAA, NEOTOMA and PANGAEA databases), the lack of a common language strongly limits the systematic reusability of paleoclimate data, for example for the construction of paleoclimatic data syntheses or the evaluation of climate model simulations.The international project “Abrupt Change in Climate and Ecosystems: Data and e-infrastructure” (ACCEDE, funded by the Belmont Forum) aims at creating an ecosystem for paleoclimatic data in order to investigate the tipping points of past climatic changes. In this context, the recently formalized Linked PaleoData (LiPD) format is the core for the standardization of paleoclimate data and metadata, and it is acting as communication protocol between the different databases that compose the e-infrastructure.Here we show two web-based solutions that are part of this effort and that take advantage of the LiPD ecosystem. The African Pollen Database, and the IPSL Paleoclimate Database, both hosted and developed by Institut Pierre-Simon Laplace, France, have the objectives (1) to give open access, while respecting the FAIR principles, to a variety of paleoclimate datasets - from pollen fossils to various tracers measured on marine sediments, ice cores or tree rings -, and (2) to combine and compare, using visualization tools, carefully selected and well dated paleoclimatic records from different disciplines to address specific research questions. The two databases are the result of data recovery from pre-existing and obsolete archives that followed a process of data (and metadata) consolidation, enrichment and formatting, in order to respect the LiPD specification and ensure the interoperability between them and the already existing databases. We designed harmonised web interfaces and REST APIs to explore and export existing datasets with the help of filtering tools. Datasets are published with DOI under an open license, allowing free access to the completeness of information. A LiPD upload form is embedded to the websites, in order to encourage both users and data stewards to propose, edit, add new records, and to bring the community into the use of LiPD format. We are currently working on finalizing visualization tools to evaluate aggregate data for research and education purposes.With this effort we are developing a framework in which heterogeneous paleoclimatic records are fully interoperable, allowing scientists from the whole community to take advantage of the completeness of the available data, and to reuse them for very different research applications.