The Xarray libray is an open source project and Python package that introduces labels in the form of dimensions, coordinates, and attributes on top of raw NumPy-like arrays,which allows for more intuitive, more concise, and less error-prone user experience. I personally consider it like a Pandas designed for geosciences, with very powerful tools to easily manipulate 3D large datasets. In this code, I make some computation in a huge dataset (daily global maps from 1993 to 2019 of geostrophic velocity obtained by altimeters) and plot the results.

The computations includes slicing data for a specific region (the South Brazil), compute a 60-day running mean and then obtain spatial derivates, to compute the horizontal shear production and the gradient of the total vorticity, after that, I group these variables by months and seasons. These calculations are done with few lines of commands and would require much more effort (and time) if I decide to work with 3D arrays only instead of xarray built-in functions.