ArcGIS and QGIS support them, but it can be difficult to work them into simulations without converting to a more benign data type like an ASCII file. They’re great for managing lots of spatial data, but if you’ve never dealt with them before, they can be a bit of a bear to work with. It’s also a convenient way to keep track of data obtained from satellite measurements over time. Folks who work with climate and earth systems tend to store their data this way. These chunks of data can then be accessed as combined spatial coverages over time.

Three variables are represented this way, and these are stored together in a single time step. The actual structure of the file might be simpler (that is, it might consist of a single variable and/or single time step) or more complex (with many more variables or where each variable is actually a set of coverages representing a range of values for that variable imagine water temperature readings taken at a series of depths). In this diagram, each table represents a gridded spatial coverage for a single variable. To make this a little more clear, here’s a diagram: The basic structure of a NetCDF file A NetCDF file (which stands for Network Common Data Form) is a multidimensional array, where each layer represents the spatial gridded distribution of a different variable or set of variables, and sets of grids can be stacked into time slices. NetCDF is one of the more notorious of these.

Some government agency or environmental science department put together some brilliant time series GIS package and you want to find a way to import it into your model. But oftentimes the data may be in a format which isn’t readable by your modeling software, or takes some finagling to get the data in there. You found it: the holy grail of palaeoenvironmental datasets.

UPDATE 22 SEPT 2016: Please see the updated version of this tutorial here.

Working with net cdf files in r