by Claire Carouge


The Australian Community Climate and Earth System Simulator (ACCESS) has been actively updated over the last few years in view to contribute to the latest Coupled Model Intercomparison Project Phase 6 (CMIP6). In particular, the CSIRO has developed ACCESS Earth System Model version 1.5 (ACCESS-ESM1.5). This model is based on ACCESS 1.3 which was used for submissions to CMIP5. The atmosphere model, the Unified Model (UM), is the same version in both ACCESS models but the other components have been updated:

  • The ocean model MOM to version 5.0
  • The sea-ice model CICE to version 4.1
  • Recent developments have been added to the land model CABLE (although not as extensively as in ACCESS-CM2).

Additionally, the carbon cycle is now modelled in the ocean with the model WOMBAT. And the carbon, nitrates and phosphate cycles are modelled over the land using CASA-CNP within CABLE.

The version of MOM5 used in ACCESS-CM2 was harmonised with the latest version of the code on the official GitHub repository. This was a collaborative effort between CSIRO (Russ Fiedler, Peter Dobrohotoff) and the CMS team (Aidan Heerdegen). A similar effort is underway for the MOM version used by ACCESS-ESM-1.5, but was not delivered in time for the CMIP6 model runs, it will be available at a later date for CLEx users.

Since ACCESS-ESM1.5 is based on ACCESS1.3, it originally used the same suite of shell scripts to configure and run the model. These are not user friendly and are difficult to maintain even by experts. Jhan Srbinovsky at CSIRO has worked on simplifying these. At CMS, we decided to put some effort into upgrading the workflow management software payu to support ACCESS-ESM1.5. Holger Wolff, Aidan Heerdegen, Scott Wales and Marshall Ward (at NCI) have been involved in this work.

Payu was developed at ARCCSS by Marshall Ward and is used extensively by the Centre’s researchers to configure and run the ocean models. Using it for ACCESS-ESM1.5 gives users a simpler way to manage their simulations. Additionally, it allows us at CMS to better support the model as we can build on more expertise rather than dividing our expertise on supporting several workflow systems.

If you want to use ACCESS ESM1.5, please refer to this page of the CMS wiki. Please note, we currently only have 2 configurations available: piControl and last interglacial. CSIRO is currently finishing the control runs for CMIP6 configurations.

NU-WRF: CABLE coupled to WRF

We have updated the WRF code coupled to LIS. This code is developed by NASA and is called NASA Unified WRF (NU-WRF). It contains WRF, LIS and LIS pre-processor, LDT, as well as a lot of other WRF utilities. Note, we are using this code to couple CABLE to WRF so no effort has been made to port any of these other utilities.

NU-WRF has now been updated to its latest version, v9 patch 2. It contains WRF v3.9.1 (modified) and LIS/LDT v7.2.

Claire Carouge has updated NU-WRF v9p2 to include CABLE based on v2.3.4 with modifications from Mark Decker to include a ground water scheme. This code is being evaluated by Annette Hirsch and Jatin Kala. The first evaluations show a warm bias in the tropics over Australia. Following these results, it was decided to update the CABLE code base to be identical to ACCESS-CM2 CABLE code. This work is in progress now. More evaluations of CABLE-WRF are also underway in particular an ensemble of simulations to compare different WRF physics options with CABLE.

NU-WRF is still under license. If you are interested in using this model, please contact the CMS team at


CMIP6 data is now beginning to be available at NCI. NCI has worked with CLEx, the CSIRO and the Bureau of Meteorology to improve the management and discoverability of the Earth System Grid Federation (ESGF) datasets at NCI. All ESGF datasets will be managed by NCI directly from now on. In particular, NCI will send all communication, handle requests for new data and downloads, host the official wiki space for CMIP (this requires a login to NCI). Note that currently, there have been very little data published, hence all available data has already been downloaded by NCI.

Paola Petrelli and Scott Wales have worked with NCI to update the ARCCSSive tool. The new tool is now called clef. This tool allows you to search for CMIP5 and CMIP6 data available at any ESGF node. It also gives you the possibility to put in a request for download if there is data available on the ESGF but not locally at NCI. Additionally to search for data, you can also use clef directly in your Python code to define the files you need for your analysis but also access other metadata (e.g. errata, version).

If you do not have an NCI login, you can find general information about CMIP6 on the WCRP website. Information on the data hosted at NCI, can be found on the NCI ESGF node. Finally, there is work under way to expand clef capabilities to other datasets hosted at NCI: CORDEX, reanalysis datasets etc. This would give users a flexible way to discover datasets that might be of interest to them. For example, one could search for which dataset currently hosted at NCI have sea surface temperatures, or cover a given period in time etc.


We have started organising the community around ERA5. Paola Petrelli has started downloading a subset of the ERA5 dataset. Following very slow download speeds, she worked with Matt Nethery at NCI to optimise the workflow. As such, the downloads are now performed by NCI using their data mover nodes with fast connections. The workflow includes an automated system for Paola (and potentially others) to put in a request for download.

The Bureau of Meteorology through its BARRA team is providing some quality control on the data and the community is discussing how best to expand this quality control.

Currently, we can not fulfill all requests as we are limited by the amount of storage available to us. We are hoping to obtain funds for storage via a LIEF grant or other sources of more permanent funding. Despite these limitations, we encourage you to let the CMS team know of your needs for ERA5. We might be able to download more data now. If not, knowing what the needs are will help us to design proper priorities when the storage becomes available.