A simplified palaeoceanography archiving system (PARIS) and GUI for storage and visualisation of marine sediment core proxy data vs age and depth.

Scientific discovery can be aided when data is shared following the principles of findability, accessibility, interoperability, reusability (FAIR) data (Wilkinson et al., 2016). Recent discussions in the palaeoclimate literature have focussed on defining the ideal database format for storing data and associated metadata. Here, we highlight an often overlooked primary process in widespread adoption of FAIR data, namely the systematic creation of machine readable data at source (i.e. at the field and laboratory level). We detail a file naming and structuring method that was used at LSCE to store data in text file format in a way that is machine-readable, and also human-friendly to persons of all levels of computer proficiency, thus encouraging the adoption of a machine-readable ethos at the very start of a project. Thanks to the relative simplicity of downcore palaeoclimate data, we demonstrate the power of this simple but powerful file format to function as a basic database in itself: we provide a Matlab-based GUI tool that allows users to search and visualise data by sediment core location, proxy type and species type. The adoption of similarily accessible, machine-readable file formats at other laboratories will promote data sharing within projects, while also allowing for the automation of submission of data to online database repositories with particular formatting and/or metadata requirements, thus reducing post-hoc workload.

遵循可发现性、可访问性、互操作性、可复用性（FAIR，即Findability、Accessibility、Interoperability、Reusability的首字母缩写）数据原则共享数据，可助力科学发现（Wilkinson等，2016）。古气候学领域近期的学术研讨多聚焦于界定存储数据及关联元数据的理想数据库格式。本文着重指出FAIR数据广泛推广中常被忽视的核心环节：即在数据源端（即野外与实验室层面）系统性创建机器可读数据。本文详述了法国气候与环境科学实验室（LSCE）所采用的文件命名与结构化方法：该方法以文本文件格式存储数据，既具备机器可读性，又对不同计算机操作水平的人员友好易用，从而可在项目伊始就推广机器可读的工作理念。鉴于沉积物岩芯古气候数据相对简洁，本文验证了这一简洁却高效的文件格式本身即可作为基础数据库使用：我们开发了一款基于Matlab的图形用户界面（Graphical User Interface，GUI）工具，支持用户按沉积物岩芯点位、代用指标类型及物种类型对数据进行检索与可视化。若其他实验室也采用此类易用且机器可读的文件格式，将有助于推动项目内部的数据共享，同时可实现向有特定格式或元数据要求的在线数据库存储库自动提交数据，从而减少事后的工作量。