A Model interface for ERA5

The ERA5 dataset provides a comprehensive view on recent climate data by assimilating vast amounts of historical observations into the ECMWF integrated forecast system, and as such establishing a reference point in the field of weather and climate modelling. The successor of ERA-interim is ubiquitous in the earth sciences, with applications such as boundary conditions for regional simulations, atmospheric forcings to ocean or land surface models, initial conditions to climate prediction experiments, etc.. The conventional workflow for such applications is to download the data, extract the necessary variables, optionally regrid or resample and save it in a model specific format. This procedure is time consuming, difficult to document properly and generates a lot of intermediate data of low reuse value. Here, we provide an alternative to this by wrapping access to the ERA5 dataset in a standardized OMUSE model interface. OMUSE is a Python framework for Earth System modelling, developed to simplify the use of simulation codes and enable new model couplings. Within OMUSE the ERA5 dataset is transparently accessed using the CDSAPI and the resulting interface is very much like an OMUSE interface for a simulation code. Data is pulled from the online climate data store only when needed and cached for later reuse. This approach simplifies the access and coupling of the ERA5 dataset with OMUSE model components and makes it trivially easy to repeat a model run with a different dataset or even replace it with a life model.