Climate Data

Analysing model output

Overview

Teaching: 30 min
Exercises: 10 min
Questions

  • How can I analyse model output?

  • How can I create plots of my data?

Objectives

  • Use Iris to load a Unified Model file

  • Explore auxillary dimensions

  • Create a Cartopy plot

As a final topic, let’s look at another library for working with climate data, Iris.

Iris represents variables as ‘data cubes’, which like Xarray combine data, co-ordinates and metadata

Iris will also create ‘auxillary co-ordinates’, that are alternate views of the other co-ordinates e.g. model and pressure levels

Iris will also load GRIB2 (used in weather forecasting) and Unified Model files, the atmospheric model used in ACCESS

Example: Inspecting a model output file with Iris

Let’s look at what we can see in an output file from the ACCESS model

Notebook

What are the differences between Iris and Xarray’s datasets?

Try loading the same NetCDF file in each - do they show the same information?

Like Xarray iris has tools for indexing and reducing data. It can also convert GRIB2 or UM files to CF-NetCDF

Example: Iris cube operations

What can we do with an Iris cube?

Plotting your data

Just like netCDF libraries there are a number of ways you can go about plotting data, inside and outside of Python.

http://matplotlib.org/ http://scitools.org.uk/cartopy/ http://geo.holoviews.org/

Example: Plotting with Matplotlib and Cartopy

Let’s explore options for creating plots

Create an interactive plot in Ipython

Check out the GeoViews documentation and create a plot in Ipython showing how surface temperature in Australia changes over a year

Start by finding a relevant dataset in THREDDS and loading it into Xarray

http://scitools.org.uk/iris/ http://scitools.org.uk/cartopy/

Key Points