The R-based climate4R open framework for reproducible climate data access and post-processing

Abstract

Climate-driven sectoral applications commonly require different types of climate data (e.g. observations, re-analysis, climate change projections) from different providers. Data access, harmonization and post-processing(e.g. bias correction) are time-consuming error-prone tasks requiring different specialized software tools at eachstage of the data workflow, thus hindering reproducibility. Here we introduce climate4R, an R-based climateservices oriented framework tailored to the needs of the vulnerability and impact assessment community thatintegrates in the same computing environment harmonized data access, post-processing, visualization and aprovenance metadata model for traceability and reproducibility of results. climate4R allows accessing localand remote (OPeNDAP) data sources, such as the Santander User Data Gateway (UDG), a THREDDS-basedservice including a wide catalogue of popular datasets (e.g. ERA-Interim, CORDEX, etc.). This provides a uniquecomprehensive open framework for end-to-end sectoral reproducible applications. All the packages, data anddocumentation for reproducing the experiments in this paper are available from http://www.meteo.unican.es/climate4R.