From the outset, cartopy’s purpose has been to simplify and improve the quality of mapping visualisations available for scientific data. Thanks to the simplicity of the cartopy interface, in many cases the hardest part of producing such visualisations is getting hold of the data in the first place. To address this, a Python package, Iris, has been created to make loading and saving data from a variety of gridded datasets easier. Some of the following examples make use of the Iris loading capabilities, while others use the netCDF4 Python package so as to show a range of different approaches to data loading. Cartopy comes with powerful vector field plotting functionality. There are 3 distinct options for visualising vector fields: quivers (example), barbs (example) and streamplots (example) each with their own benefits for displaying certain vector field forms. Since both quiver() and barbs() are visualisations which draw every vector supplied, there is an additional option to “regrid” the vector field into a regular grid on the target projection (done via cartopy.vector_transform.vector_scalar_to_grid()). This is enabled with the regrid_shape keyword and can have a massive impact on the effectiveness of the visualisation: Source.