Manipulation of netCDF data with R for climate change research: Multi-model analysis for CMIP5 models

Geoscientists now live in a world with an exponential growth in digital data and methods. Climate change studies usually describe computational methods informally. Climate scientists seek to share their information, the justification of reproducible research has received increasing attention in geosciences. To have it in an open-source format makes it easier to interchange not only with fellow scientists but also a variety of sources including funders, publishers, and journalists. R is a open-source computer language powerful and highly extensible that can promotes reproductive science techniques in a easier way. R is highly accessible for non-computational scientists when coupled with packages like ‘raster'', ‘netcdf'', ´rgdal`and ‘rasterVis'', R enables scientists to make sense of their data and to carry out complex data analysis. In this paper we have assessed the power of R language for manipulating climate data from a huge dataset: the Coupled Model Intercomparison Project Phase 5 (CMIP5). Moreover we have proposed an example of best practices to handle model ensembles. This is the first study to our knowledge to promote best practices for CMIP5 ensemble. The NetCDF data accessible to R via raster package capabilities provides efficient access to the multi-model, with crucial applications in climate change research. In recent years more than 100 peer-reviewed scientific publications have used the CMIP5 data sets. We envision that in the near future (5-10 years), scientists will use radically new tools to author papers and disseminate information about the process and products of their research.