University of Auckland researcher Tra Dinh is studying how changes in atmospheric processes contribute to future climate change.

Investigating climate sensitivity

“The numerical simulations and predictions of the Earth’s climate, which are an essential component of my research, are not possible without NeSI’s high computing facility.”

The more we understand about the relationships between atmospheric CO2, water vapour, clouds, and global circulation, the better we will be able to predict climate change. 

That’s what University of Auckland researcher Tra Dinh is exploring. Recently, she authored an article for the university’s “The Big Q”, to answer some of the most frequently asked questions about climate change the IPCC special report.

“I use theory and numerical tools, in combination with observations, to study how the multi-scale interactions of atmospheric processes underline the basic structure of the atmosphere and how these interactions will contribute to future climate change,” she says.

Using a circulation model and NeSI’s computing facilities, she is investigating how changes in CO2 level is linked to changes in atmospheric humidity and clouds, which feedback on the radiative energy budget of the Earth’s climate system, and in turn affecting climate sensitivity.

“The numerical simulations and predictions of the Earth’s climate, which are an essential component of my research, are not possible without NeSI’s high computing facility,” she says.

Using NeSI facilities, Tra is able to run and analyse her simulations in a much shorter time, making it easier to identify the results and answers to her research questions. 

“Tra is using the High Resolution Atmospheric Model (HiRAM) and so access to NeSI is critical for her to run her numerical simulations. The model runs most efficiently on parallel computing platforms,” says NeSI’s Chris Scott, who worked with Tra on this project.

With multiple interacting variables (e.g. solar insolation, greenhouse gases, humidity, temperature, circulations, clouds) our climate system is incredibly complex and dynamic. For Tra, the multi-variable interactions and their roles in determining Earth’s climate are particularly interesting. 

“Some of these interactions are nonlinear and they may not be intuitive at first,” she says. “For example, recently I found that the seasonal variations of solar insolation have a large influence on the annual-mean spatial distribution of water vapour in the atmosphere. This is work in progress, and this result was found in numerical simulations of Earth’s climate using the NeSI computing facility.”

For more information on Tra’s research, visit her University of Auckland webpage


Need help with your research project? NeSI’s computational science team can assist with code optimisation, parallelisation, porting to GPUs, custom code development, and many other tasks. Email if you are interested to find out more or if you have a task you’d like NeSI to help you tackle.


The Technical Details:

Below you’ll find a more technical description of the support NeSI offered Tra Dinh. If you have any questions about these details, contact

Some of the ways our team supported Tra Dinh, include:

  • Compiled the HiRAM code on NeSI platforms (Pan, Kupe and Māui)
  • Ported HiRAM scripts to run on NeSI clusters under Slurm, including checkpointing capability and making sure output is stored in the correct format and locations (e.g. making use of the fast “nobackup” file system for performance)
  • Ported and installed the IVE visualisation tool on NeSI platforms to improve Tra’s productivity. This was a quite a tricky code to get installed as it looked like it hadn’t been updated for some time and had dependencies on some obscure/old packages.
  • On Kupe and Māui, HiRAM takes around 1 hour of wall time per simulated year - this is about 1.7 times faster than on the previous machine Tra was using.


Next Case Study

NeSI Case Study Yiming Peng

Taking machine learning further

“I couldn't imagine how long my PhD journey would go if I hadn't used the high performance computing resources of NeSI."