Plotting your data

Once you have run your workflow, you will likely want to visualise the output data. There are several ways you can do this. Here we cover two methods: using terminal GUI tools and using Python with the Iris library.

Using GUI tools

For quick visualisation and inspection of NetCDF data files, you can use graphical tools such as ncview and xconv. These tools provide a fast way to explore datasets without writing any code, making them useful for checking file contents, verifying dimensions, and performing an initial quality check before more detailed analysis.

Running graphical applications over SSH

If you are connected to a remote system over SSH, graphical applications need X11 forwarding to display on your computer. Start an X server on your computer first (for example, XQuartz on macOS), then connect from a terminal using generic host details supplied by your service:

$ ssh -AY <remote-host>

Repeat the options for every additional SSH hop. The -Y option enables trusted X11 forwarding. The -A option forwards your SSH authentication agent and should only be used with systems you trust. A terminal integrated into an editor such as VS Code works in the same way as any other terminal.

On the system where you will start the graphical application, run echo "$DISPLAY". It should return a value. If it is blank, disconnect and reconnect with X11 forwarding enabled at every hop before continuing.

ncview

ncview is a lightweight graphical browser for NetCDF files. It can display 2D fields, animate data through time, inspect values, and switch between variables and dimensions. It is particularly useful for obtaining a quick overview of model output or observational datasets.

Example:

Met Office

ncview temperature.nc

Monsoon

See also

For help choosing the correct Monsoon3 service and finding the official setup guidance, see Using Monsoon3.

ncview temperature.nc

Useful tasks include:

• Browsing available variables in a file

• Stepping through time slices

• Viewing simple animations

• Examining data ranges and missing values

Documentation and setup:

• ncview documentation: https://cirrus.ucsd.edu/ncview/

xconv

xconv is a graphical data exploration and conversion tool commonly used within the atmospheric and climate modelling community. It supports NetCDF files as well as several meteorological data formats and provides facilities for plotting, inspecting dimensions, and converting data between formats.

Example:

Met Office

xconv

Monsoon

xconv

After launching the application:

1. Open a NetCDF file.

2. Select a variable from the variable list.

3. Use Plot Data to display the selected field.

4. Navigate dimensions such as time, level, latitude, and longitude using the GUI controls.

Documentation and setup:

• xconv documentation: https://ncas-cms.github.io/xconv-doc/html/index.html

• xconv setup guide: https://ncas-cms.github.io/xconv-doc/html/setup.html

When should I use GUI tools?

GUI tools are ideal for:

• Quickly checking that a file contains the expected variables

• Exploring dimensions and metadata

• Verifying model output before analysis

• Performing a rapid visual quality-control check

For reproducible analysis, processing large datasets, or creating publication-quality figures, command-line tools and scripting languages such as Python are generally more appropriate.

If these tools are installed via environment modules on your system, you may first need to load them before use:

Met Office

module load ncview module load xconv

Monsoon

module load ncview module load xconv

Refer to your local HPC documentation for the exact module names available on your platform.

Using Python

The output files are in NetCDF format, which can be read using a variety of software packages. Python provides more flexibility and allows for more complex visualisations. Below is an example of how to read and plot NetCDF data using the Iris and Matplotlib libraries.

import os from pathlib import Path

import iris import iris.quickplot as qplt import matplotlib.pyplot as plt

# Set the path to the data. path_to_data = Path("path/to/your/data")

# List what that path contains (up to 10 files).
print(os.listdir(path_to_data)[:10])

# Load one output file to inspect available cubes.
print(iris.load(path_to_data / "data_file.pp"))

# Load a cube from a NetCDF file. surface_temperature_cube = iris.load_cube(
    path_to_data / "output_file.nc", "surface_temperature"
)

# Change the units from Kelvin to Celsius.
surface_temperature_cube.convert_units("celsius")

# Choose the time slice to plot (e.g. the first time slice). time_slice = 0

# Plot the data. qplt.contourf(surface_temperature_cube[time_slice, :, :])
plt.gca().coastlines() plt.show()

The resulting plot can be seen below.

Fig. 27 Example output from iris.quickplot.