The four-day annual event brings together the community of users, developers and collaborators of the ACCESS Earth system models. In 2024, the Program Committee seek to bring together people across all the ACCESS communities to present and discuss current and emerging topics, both technical and scientific around the following broad themes:
Model development and model evaluation for CMIP7
High resolution modelling
Paleoclimate and ice-sheet modelling
Build community around the ACCESS models
Showcase innovative science and modelling approaches
Learn about new components developed for ACCESS
Share knowledge between Community Working Groups and foster new and existing community- wide collaborations
TRAINING DAY PROGRAM
Monday 2nd September
Time Topic & Speaker
Training Day Coordinator: Paige Martin (ACCESS-NRI)
This is the second CMIP7 Evaluation Hackathon after one held in March of this year. The focus of this hackathon will be on the evaluation frameworks developed by ACCESS-NRI. We will give an overview of the available tools, including ILAMB and ESMValtool and the available data. There will be a series of hands-on exercises throughout the day, on topics including evaluation frameworks, running and writing recipes, and using Jupyter notebooks to interact with ESMValTool.
How to find and access datasets on NCI & handle large model output
Instructors: Anton Steketee, Paige Martin
Assistants: Jasmeen Kaur, Marc White
ACCESS-NRI
This session will show you some tools available for finding and accessing observational and model data on NCI. This will include an introduction to the ACCESS-NRI Intake data catalog where we will show you how to efficiently find, load and share ACCESS model output from your Python environment. In the second portion of this session, we will cover some best practices for processing and analyzing large output from ACCESS models. We will use software packages Xarray and Dask to optimize large dataset workflows
Learning Objectives
• Understand where and how climate data is stored on NCI
• Learn various methods of accessing climate data on NCI
• Understand how to use the ACCESS-NRI Intake data catalog and why it’s useful
• Learn basic best practices for handling large model output
Pre-Requisites
• Basic unix command-line knowledge
• Some exposure to Python programming
Advanced Git & GitHub best practices for software developers
Instructors: Micael Oliveira, Harshula Jayasuriya
ACCESS-NRI
Learn how to use Git and GitHub more effectively by becoming familiar with more advanced Git concepts and commands. If you have been using Git at a basic level, but would like to better understand what a commit is, what really happens during a merge, what the difference is between a merge and a rebase, or what the most commonly used Git workflows are, then this tutorial is for you.
Learning Objectives
• Deep understanding of how to best use Git and GitHub for scientific software development
• Exposure to common Git workflows in software development
Pre-Requisites
• Working knowledge of basic Git commands (add, commit, pull, push, etc.)
• Working knowledge of GitHub for software development and collaboration
• Experience with software development
Running a high-resolution regional model using a two-level nest
Instructors: Chermelle Engel, Paul Leopardi
Assistants: Heidi Nettelbeck, Martin Dix
ACCESS-NRI
Would you like to get started using the ACCESS high-resolution regional model to simulate an interesting weather event? This session will introduce you to a realistic experiment and help you prepare to run your own experiments using ERA5+ERA5- land initial conditions.
Learning Objectives
• Gain a basic understanding of what the ACCESS high resolution regional model is
• Understand how to run a double-level nest using either ERA5+ERA5-land or ERA5+BARRA-R2 initial conditions
Pre-Requisites
• Basic understanding of what limited area models are
• Basic familiarity with rose/cylc
TOPIC
Running model experiments with Payu and Git
Instructors: Aidan Heerdegen, Jo Basevi, Spencer Wong, Anton Steketee
ACCESS-NRI
Do you run ACCESS model experiments? Not sure how to best organize and share model runs? This session will cover how to use payu to run ACCESS-OM2 and ACCESS-ESM1.5 model experiments and organise experiment output. We will also cover how to use Git, and share experiment configurations on Github.
Learning Objectives
• Learn how to run ACCESS models
• Understand why it’s beneficial to organize and share model runs
• Learn how to organize and share model runs
Pre-Requisites
• Basic knowledge of Git
• Basic unix commandline knowledge
KEYNOTE SPEAKERS
KEY NOTE
The current status of global high-resolution modelling activities
Tomoki Miyakawa
Atmosphere and Ocean Research Institute (AORI), University of Tokyo
Many research groups around the world started to develop a new generation of global high-resolution models, which have the capability to resolve the vertical energy transport associated with deep convective clouds in the tropics and ocean eddies. These models matured to a degree, that they can be evaluated and tested in inter-comparison projects. In the Dynamics of the Atmospheric general circulation Modelled On Non-hydrostatic Domains (DYAMOND) project, first a winter period with only atmospheric models, than a summer period with atmospheric and coupled models was simulated with many of the new generation models.
NICAM is one of the earliest developed models of the kind. It shows signs of improvement in many aspects such as the diurnal cycle, tropical cyclone, MJO, and general circulation, when run at 3.5 km grid spacing.
Fugaku, a state-of-the-art supercomputer can run NICAM with 3.5 km configuration at a speed of 2 simulation year per day (2SYPD). NICAM can be run at a global 220m horizontal grid spacing at its finest, although only for a very short run. The ocean-coupled version NICOCO is used with its ocean component at 0.1 degree grid spacing, which is capable of resolving meso-scale eddies. Recent activities of NICAM/NICOCO will be introduced, along with results and on-going plans of the DYAMOND intercomparison project.
An introduction to ACCESS' ice sheet model: the Ice-sheet and Sea Level System Model (ISSM)
Felicity McCormack
Monash University and ARC Special Research Initiative
The ACCESS-NRI and Cryosphere Modelling Working Group recently completed an exercise to select an ice sheet model to include within the ACCESS model suite with the primary objective of facilitating decadal- to centennial-scale projections of sea level rise and climate change. Here, we outline the three phase, purpose-driven approach developed, which ultimately led to the selection of the Ice-sheet and Sea-level System Model (ISSM). ISSM is a finite-element thermomechanical ice sheet model that solves the full Stokes equations of momentum balance on an adaptive, anisotropic mesh, and has been widely applied across Antarctic, Greenland, and glacier domains. We outline some of the key capabilities of ISSM, highlighting some case studies from Australian-based users.
Title: K-Scale: A global to regional model hierarchy for assessing the costs and benefits of high-resolution global simulations
Richard
Jones Met Office, UK
The K-Scale model hierarchy includes global, regional and pan-tropical domains to investigate the relative importance of resolution, science configuration and domain size. It enables us to explore the question of whether, where, and how explicit convection global (and pan-tropical) simulations provide additional forecast information compared with our current approach to high-resolution forecasting – limited area regional models. We start to think about the costs and benefits of large-domain explicit convection forecasts focussing on two 40-day periods, one initialised in August 2016 and another in January 2020 – the DYAMOND summer and winter periods. We also look ahead to future plans for highresolution operational global forecasts at the UK Met Office.
KEY NOTE
Earth System Model Evaluation Tool (ESMValTool): Analyzing CMIP-style data made easy
Birgit Hassler
German Aerospace Centre DLR
The Earth System Model Evaluation Tool (ESMValTool) has been developed with the aim of taking model evaluation to the next level by facilitating analysis of many different ESM components, providing well-documented source code and scientific background of implemented diagnostics and metrics and allowing for traceability and reproducibility of results (provenance). This has been made possible by a lively and growing development community continuously improving the tool supported by multiple national and European projects. The latest major release (v2.0) of the ESMValTool has been officially introduced in August 2020 as a large community effort, and since then several additional smaller releases have followed.
Since it is designed to assess output from ESMs participating in the Coupled Model Intercomparison Project (CMIP), ESMValTool expects input data to be formatted according to the CMOR (Climate Model Output Rewriter) standard. While this CMORization of model output is a quasi-standard for large model intercomparison projects like CMIP, this complicates the application of ESMValTool to non-CMOR-compliant data like native climate model output (i.e., the raw output directly provided by a climate model). In one of the recent versions of ESMValTool an extension was added that allows reading and processing native climate model output. This is implemented via a CMOR-like reformatting of the input data during runtime. This extension opens up the large collection of diagnostics provided by ESMValTool for these models. The latest improvements to ESMValTool focus on optimizing the memory efficiency when reading and processing input data, and the addition of new metrics to benchmark new model simulations against observations and other models (e.g. the suite of CMIP6 simulations).
“climatearchive.org”: 500 million years of climate data at your fingertips
Sebastian Steinig University of Bristol, UK
Earth’s climate shows a remarkable variability on geological timescales, ranging from widespread glaciation to ice -free greenhouse conditions over the course of the Phanerozoic, i.e. the last 540 million years. Earth system modelling allows us to better understand and constrain the drivers of these changes and provides valuable reference data for other paleoclimate disciplines (e.g., chemistry, geology, hydrology). However, the sheer volume and complexity of these datasets often prevents direct access and use by non-modellers, limiting their benefits for large parts of our community.
I present the online platform “climatearchive.org” to break down these barriers and provide intuitive access to (paleo) climate data for everyone. More than 100 coupled climate model simulations with the Unified Model (UM) build the backbone of the web application. Key climate variables (e.g. temperature, precipitation, vegetation and circulation) are displayed on a virtual globe in an intuitive three-dimensional environment and on a continuous time axis throughout the Phanerozoic. The software runs in any web browser — including smartphones — and promotes visual data exploration, streamlines model-data comparisons, and supports public outreach efforts. I will discuss our novel modelling approach, the current web application and outline the future integration of new sources of model and geochemical proxy data to streamline and advance interdisciplinary paleoclimate research.
ABSTRACT
ORAL PRESENTATIONS
The role of sea-surface temperatures in a June 2016 east coast heavy rain event: insights from AUS2200 simulations.
Chris Chambers
The University of Melbourne
In early June 2016 an east coast low complex brought extensive heavy rainfall that produced flooding in areas of Queensland, New South Wales, Victoria, and Tasmania. In the lead-up to this event, sea-surface temperatures (SSTs) in the Coral and Tasman Seas were the warmest on record. Our aim is to determine how the high SSTs, and their distribution, influenced the rainfall and the storm-system’s evolution. To do this, three AUS2200 sensitivity to SST simulations, for the period 0000 UTC 3 June to 0000 UTC, 8 June 2016, have been run using ERA5 skin-temperature data over water. The three simulations are run with 1)1980-2019, 3 June average ERA5 skin temperature (Climatology), 2) constant 3 June 2016 ERA5 skin temperature (Fixed), and 3) daily evolving ERA5 skin temperature (Evolving). The Fixed and Evolving simulations, that include the observed warmer SSTs, produce greater rainfall than the Climatology simulation over much of the ocean area and most of the east-coastal mountains. The Fixed and Evolving simulations also produce a deeper east coast cyclone that intensifies over a prominent warm eddy. These cases also keep the low pressure further to the north, off the New South Wales coast, for longer than the Climatology run. Towards the latter stage of the simulation period, a complicated multicentred low-pressure system around Tasmania, seen in observations, correlates better with the Fixed and Evolving simulations than the Climatology simulation. The potential mechanisms responsible for these differences will be discussed. Additional simulation results will be presented to investigate the relative roles of the warm eddy and the broad-scale warmer SST.
ABSTRACT
ACCESS-NRI support of high-resolution modelling using ACCESS-RM with ERA5 driving condition
Chermelle Engel
ACCESS-NRI
Researchers working outside of operational centres, such as the Australian Bureau of Meteorology, that wish to work on high-resolution modelling have different barriers to entry and ability to access data.
High-resolution ACCESS model runs require initial- and lateral boundary conditions. Instead of using initial- and lateral boundary conditions from the parent (ACCESS) model, researchers outside of operational centres tend to use initial- and lateral boundary conditions from reanalysis datasets such as ERA5 or ERA5-land.
There were known issues when using ERA5 soil moistures as an initial condition.
ACCESS-NRI has worked to:
• Circumvent the known issues when using ERA5 soil moistures as an initial condition.
• Add the ability to instead use ERA5-land land/surface (including soil moisture) information at 10km rather than 31km, or BARRA-R2 land/surface (also including soil moisture) at 11km NOT 31km.
• demonstrate the impact of using a double- rather than single-level nesting strategy. The results of the ACCESS-NRI work will be shown in this talk along with the strategy to release suites that support similar work for interested researchers.
ABSTRACT
Evaluating ACCESS-AM2 against aerosol observations
Sonya Fiddes
Australian Antarctic Program Partnership, IMAS, UTAS
ACCESS-CM2 and AM2 include a double moment aerosol scheme that can prognostically simulate a range of aerosols. Natural marine aerosol and their interaction with clouds and radiation is one of the most significant sources of climate model uncertainty. The Southern Ocean provides a relatively unpolluted region to study natural processes and has some of the most significant cloud-radiation biases globally.
Here we present an evaluation of ACCESS-AM2 aerosol over the Southern Ocean. We evaluate against seven ship campaigns, two station campaigns, and the permanent instrumentation at Kennaook-Cape Grim (KCG). The observations span the years 2015-2019 and while biased towards summer, cover a range of seasons. We use observations of condensation nuclei (N10) and cloud condensation nuclei (CCN), measures of aerosol number concentrations at different sizes.
We find that N10 is heavily underestimated across all regions and seasons, with the bias worse at higher latitudes. Outside of northern latitudes, CCN is also significantly underestimated, while at KGC, the model performs within 1% of observed. We then perform eight experiments to explore different aerosol configurations. Increasing the sea spray flux improves CCN for marine regions but has detrimental impacts on the radiation budget. This presents a concern whereby improving one aspect of the system makes other parts significantly worse. Updating to the most recent dimethyl sulfide climatology and turning on the primary marine organic aerosol flux marginally improves CCN and the Southern Ocean radiative bias with limited compensating errors. We recommend this configuration is considered for future releases of ACCESS.
ABSTRACT
BARPA-C - Convective-Scale Regional Climate Modelling in Australia
Emma Howard Bureau of Meteorology
Natural hazards are changing across Australia and globally in response to anthropogenic climate change. Accurate projections of Australian climate hazards, such as intense rainfall, extreme winds, bushfires, and cyclones, depend to varying degrees on the representation of atmospheric convection in climate models and its upscale effects on synoptic weather systems. Furthermore, the locations of exposed populations such as urban centres and coastlines are often poorly resolved in low resolution climate models. The Bureau of Meteorology is developing a suite of modelling systems, BARPA, based on the Met Office Unified Model (MetUM) coupled to the JULES land surface model, for generating an ensemble of projections for Australian Climate Service (ACS).
As a convection-permitting regional climate model, BARPA-C uses the latest regional atmosphere and land- configuration of the models RAL3.2. BARPA-C builds on BARPA-R, the Bureau's core set of downscaled 17-km projections, which relies on parametrisation to provide atmospheric convection. Following regional climate modelling protocols, BARPA-C has been evaluated by downscaling reanalysis data and comparing modelled to observed climatologies. 10 years of ERA5 have been downscaled to 4-km grid-spacing using BARPA-C, from 2013 to 2022, via the regional-scale BARPA-R nest. The application of spectral nudging techniques to improve the representation of large-scale climate features has been investigated. This presentation will evaluate BARPA-C against high-resolution observational datasets, including radar data, Himawari8 and BARRA-C2. This assessment will also evaluate the ability of BARPA-C to improve the representation of climate hazards, such as intense rainfall and extreme winds, compared to BARPA-R and global driving models.
ABSTRACT
Estimating the climate in a global model with small samples: how many do we need?
Sramana Neogi Monash University
The need for high-resolution global climate model simulations has been widely recognised by the atmospheric science community for decades. One major roadblock for developing this capability is the large amount of computational resources required to carry out the model simulations. Here, we present an approach that could reduce this cost by using relatively short simulations in studying model errors and responses.
As a first step, we try to estimate how much we can learn about the climatology in a global climate model (GCM) simulation run in the AMIP (Atmospheric Model Intercomparison Project) configuration for 35-years from relatively short samples. Analysis of the 35-year AMIP simulation from a GCM shows that creating an ensemble of 4-5 years (randomly selected) from these 35-years is sufficient to replicate climatological model errors. We then run the Australian Community Climate and Earth System Simulator Coupled Model (ACCESS-CM2) in the AMIP configuration for 3-4 months to see if we can replicate the results. We also explore how well the response to El-Nino and La-Nina can be captured by a few simulated years. In this talk, we will discuss the feasibility of running short simulations and its possible applications in high-resolution atmospheric modelling.
ABSTRACT
A machine learning approach to rapidly project climate responses using CMIP data
Vassili Kitsios CSIRO
The world economy is currently grappling with how to transition to net-zero emissions later this century. There is hence a need to assess the physical climate risks for a broad range of future economic, and hence emissions, scenarios. A small selection of the possible scenarios is simulated by numerous climate models as part of the international Coupled Model Inter-comparison Projects (CMIP). These simulations provide estimates of the future temperature, rainfall and other properties required to assess physical climate risk. Here, we employ model reduction and machine learning techniques to generate climate simulation output at a fraction of the computational time and cost. The presentation will outline the methodology and validate the method via its ability to reconstruct CMIP data excluded from the learning phase. One can now rapidly estimate the climate response for many more emissions pathways that would ordinarily be possible using CMIP models alone. We refer to this approach as QuickClim. We use QuickClim to reconstruct plausible climates for a multitude of concentration pathways. As expected, higher mean temperatures coincide with higher end-of-century carbon concentrations. However, the uncertainty in the climate variability saturates earlier, in the mid-century, during the transition between current and future climates. For pathways converging to the same end-of-century concentration, the climate is found to be sensitive to the choice of trajectory. In net-zero emission type pathways, this sensitivity is of comparable magnitude to the projected changes over the century.
ABSTRACT
ACCESS-OM3: Australia’s next-generation global ocean and sea ice model for CMIP7 and beyond
Andrew Kiss ANU
ACCESS-NRI and COSIMA are developing the ACCESS-OM3 multi-resolution global ocean and sea ice model suite, which couples the state-of-the art MOM6 and CICE6 ocean and sea ice models, and (optionally) the WaveWatch III surface wave model and upgraded WOMBAT ocean biogeochemistry. ACCESS-OM3 will be driven by prescribed atmospheric data and available at resolutions of 1°, 0.25°, 0.1°, and beyond. The MOM6-CICE6 configurations at 1° and 0.25° will also be incorporated into the ACCESS-CM3 and ACCESSESM3 coupled physical and Earth system models intended for CMIP7. This presentation will provide an overview of the new research capabilities provided by the model suite, an update on progress and early results, and plans for the future, including integration into CMIP7 development efforts.
ABSTRACT
The Response of the Southern Ocean to Climatological Iceberg Freshwater Forcing
Jingwei Zhang UTAS
Antarctic icebergs, as a moving source of freshwater while drifting in the Southern Ocean, have been suggested to play a significant role in altering water masses and sea ice dynamics, especially during recent increases in Antarctic ice discharge. Most current climate models, however, misrepresent Antarctic iceberg melting by assuming it releases uniformly along the coast. In this study, we investigate the impacts of the climatological spatial distribution of iceberg melt on the Southern Ocean by implementing a recent simulation of iceberg melt pattern in a global ocean-sea ice model. Our results demonstrate that widespread iceberg melting leads to decreased surface salinity and enhanced stratification, which in turn suppresses vertical convection and mixing, ultimately leading to a reduction in upward heat transport. Such reduced heat transport results in an accumulation of heat in the subsurface and a cooling effect at the surface, further intensifying the formation of sea ice. Additionally, we emphasize the significance of incorporating a more realistic distribution of iceberg melt into ocean models, which can reduce model bias of climatological temperature and salinity by up to 30% compared with observations. Our results underscore the necessity of considering iceberg drifting patterns in climate/ocean models. Current models that neglect drifting icebergs must be interpreted based on the responses presented in this study.
First Impressions of ACCESS-S2 Antarctic Sea Ice Forecast
Elio Campitelli Monash University
ACCESS-S2, the Australian Bureau of Meteorology's current operational seasonal forecast system, generates initial conditions by assimilating oceanic and atmospheric data. Unlike its predecessor ACCESS-S1, it does not incorporate sea ice concentration observations. Hence, the sea ice initial state is formed by the ocean initial conditions.
Our study evaluates ACCESS-S2's performance in forecasting sea ice concentration in polar regions, with a focus on Antarctica. Using hindcast and reanalysis data from 1981 to 2018, we assess the system's skill and investigate sources of error. Results indicate that ACCESS-S2 generally exhibits low skill in forecasting sea ice extent and concentration anomalies when compared to a simple persistence forecast. The system also displays high biases, particularly during summer months.
Analysis reveals that a significant portion of forecast errors stems from inaccuracies in initial conditions. This suggests that incorporating sea ice concentration data into the forecasting system could potentially reduce these errors, thereby improving overall forecast accuracy. Our findings highlight the importance of sea ice data assimilation in enhancing polar region predictions and provide valuable insights for future improvements to the ACCESS-S2 system.
ABSTRACT
Modelling the Weather of the 21st Century
Julie Arblaster Monash University
Improving the models we work with is fundamental to improving our ability to predict and understand future weather change. In the Modelling Science project of the Centre of Excellence for 21st Century weather, we are building km-scale versions of the ACCESS model, both regional and global, coupled and uncoupled. This talk will provide an update of the science and improved modelling capability we are undertaking to e.g. understand the role of small-scale SST gradients on atmospheric phenomena, including tropical cyclones and coastal processes; quantify the role of surface heterogeneity on weather systems; and identify the impact of small-small coupling on the large scale weather and climate.
ABSTRACT
CMIP6 evaluation on the dynamical evolution of the SH polar vortex anomalies
Eun-Pa Lim Bureau of Meteorology
The springtime stratospheric polar vortex variability over Antarctica is an important source of predictability of the southern annular mode (SAM) and associated southern hemisphere (SH) regional climate in austral spring-summer seasons. The Antarctic spring vortex anomalies often result from the poleward and downward progression of the meridional dipole anomalies of the SH stratospheric winter jet with vigorous wave-mean flow feedback. This dynamical process is conveniently captured by the anti-correlation of the zonal mean zonal wind anomalies at 60°S in the upper stratosphere between winter and spring. In this study, we evaluate the ability of CMIP6 models to simulate the dynamical connections from the stratospheric winter jet to the spring polar vortex to its downward coupling to the tropospheric SAM. We examine zonal winds from 41 models at all available pressure levels with the historical forcing’s and the future forcing’s following the Shared Socioeconomic Pathway 5-8.5.
Our results show that models, including ACCESS ESM1.5 and CM2, tend to simulate the observed anticorrelation between the winter and spring westerlies at much higher latitudes of 10-20°, which is significantly linked to the models’ biases in the mean position of the winter stratospheric jet. Furthermore, the relationship between the stratospheric polar vortex and the surface SAM tends to be overestimated compared to the observed in winter, whereas it is skillfully simulated in spring. The identified winter stratospheric jet-spring vortex biases appear to affect the models’ future projections of the Antarctic stratospheric vortex with increasing greenhouse gases by the end of the 21st century.
ACCESS-AM3 – Progress towards the 3rd generation of Australia’s Earth system model
Ian Harman
CSIRO Environment
The next generation of ACCESS Earth system model, ACCESS-ESM3, comprising advances to all components of ACCESS is being developed. This configuration is targeted at a broader and longer-term CMIP7 submission, post FastTrack. A key step in the development of ACCESS-ESM3 is the creation of a landatmosphere physics-only configuration, ACCESS-AM3, involving the technical coupling between the land and atmospheric components, CABLE and the Unified Model.
We will describe the configuration of ACCESS-AM3, its current status (alpha version), and present some first results from test historical runs. We will also outline how the broader community can engage with the development of the model.
ABSTRACT
RRR: Reliability, Replicability, Reproducibility for Climate Models
Aidan Heerdegen ACCESS-NRI
It is difficult to reliably build climate models, reproduce results and so replicate scientific findings. Modern software engineering coupled with the right tools can make this easier.
Some sources of complexity that make this a difficult problem:
Climate models are an imperfect translation of extremely complex scientific understanding into computer code. Imperfect because many assumptions are made to make the problems tractable.
Climate models are typically a number of separate models of different realms of the earth system, which run independently while exchanging information at their boundaries.
Building multiple completely separate models and their many dependencies, all with varying standards of software engineering and architecture.
Computational complexity requires high performance computing (HPC) centres, which contain exotic hardware utilising specially tuned software.
ACCESS-NRI uses spack, a build-from-source package manager that targets HPC, and which gives full build provenance and guaranteed build reproducibility. This makes building climate models easier and reliable. Continuous integration testing of build correctness and reproducibility, model replicability, and scientific reproducibility eliminates a source of complexity and uncertainty. The model is guaranteed to produce the same results from the same code, or modified code, when those changes should not alter answers. Scientists can be confident that any variation in their climate model experiments is due to factors under their control, rather than changes in software dependencies, or the tools used to build the model.
ABSTRACT
ENSO and tropical basin interactions in idealized worlds
Dietmar Dommenget Monash University
In this study we discuss a set of fully coupled general circulation model simulations with idealised geometries of the tropical ocean basins and land with a focus on important characteristics of El Niño Southern Oscillation (ENSO) type of variability and tropical basin interaction. In a series of 15 simulations we first vary the zonal width of a single tropical ocean basin from 50^o to 360^o, while the rest of the tropical zone is set as land. Further we discuss different simplified configurations of two or three tropical ocean basins. We will discuss the difficulties on how to setup such runs with the ACCESS-ESM and GFDL CM2.1 models and present results about the ENSO dynamics. The results show remarkable changes in ENSO characteristics as function of basin width and due to the interaction with other basins that challenge our current understanding of ENSO dynamics. A single basin ENSO has an optimal basin width of about 150^o at which ENSO preferred period is the longest, the wind stress feedback is the strongest and variability is stronger than in all other basin widths, expect for the 350o basin. Tropical basin interactions substantially affect ENSO strength, periodicity, feedbacks, non-linearity, spatial scale and pattern. In experiments with two or three identical ocean basins we find highly synchronised ENSO modes that are identical between basins and far more energetic and oscillatory then the single basin modes. The results suggest that tropical basin interaction is an essential part of ENSO.
ABSTRACT
Role of distinct glacial boundary conditions in simulating a glacial climate state.
Himadri Saini UNSW
Earth’s climate history has oscillated through several glacial and interglacial cycles. Among these, the Marine Isotope Stage 3 (MIS 3), spanning from 65,000 to 27,000 years ago, exhibited distinct millennialscale variability. This study presents the simulation of the glacial climate at around 49,000 years ago (49ka), within MIS 3, using the Australian Earth System Model (ACCESS-ESM1.5). In a first step, only orbital parameters and greenhouse gases consistent with 49ka conditions were modified. In a second step, the continental ice-sheet topography and extent as well as associated changes in vegetation and land-sea mask were included. Here we will present the impact of the combined 49ka boundary conditions on climate as well as the separate impacts of orbital parameters plus greenhouse gases, and continental ice-sheets.
ABSTRACT
ACCESS-ESM1.6 – Australia’s contribution to the CMIP7 Fast Track
Tilo Ziehn CSIRO
Australia has participated in all international Coupled Model Intercomparison Project (CMIP) activities so far and submitted ACCESS model versions to the last two rounds of CMIP – CMIP5 and CMIP6. Preparations for CMIP7, deploying new ACCESS model versions, are well underway. Two configurations are planned. ACCESSESM1.6, based on our successful CMIP6 submission, is being prepared to meet the tight CMIP7 Fast Track timeline which requires publication of the results by the end of 2026. The next generation ACCESS-ESM3 configuration is targeted at a broader and longer-term CMIP7 submission. This presentation describes the configuration design of ACCESS-ESM1.6 with a focus on significant upgrades to the land and ocean biogeochemistry and other improvements and changes. We will present some first results from test runs including preliminary evaluation of the performance.
ABSTRACT
Benchcab: a scientific evaluation framework for the CABLE land surface model
Sean Bryan ACCESS-NRI
Model developers often test the scientific validity of their modifications using ad-hoc scripts/tooling which target the scientific expertise of the model developer. A lot of the evaluation is also done via published results which tend to overrepresent positive effects on simulations. For CABLE, we decided to develop benchcab: a scientific evaluation framework to provide a standardised and systematic evaluation of the model. With benchcab, a developer can run several versions of CABLE across a range of model configurations and then get access to a statistical analysis of the performed simulations via the modelevaluation.org platform. The results show a range of analyses from overall performance across several statistical metrics all the way to detailed timeseries. The analyses look at all aspects of a land surface model and are consistent to allow comparison between developments.
ABSTRACT
Developing the next standard configuration for standalone JULES using a benchmarking system based on ModelEvaluation.org
Heather Rumbold UK Met Office
Benchmarking of land surface models (LSMs) involves adopting widely agreed standards for judging performance. Unlike evaluation or validation, benchmarking requires comparison of outputs with predefined targets or thresholds, allowing meaningful inter-comparisons of independent models or different configurations of a single model.
The benchmarking approach can be used to determine the suitability of new model configurations by comparison with the performance of previous configurations of the same model. The work developed here assesses new components of JULES for use in future model configurations, using a benchmarking system based on the ModelEvaluation.org web application. The configurations generated have been run through a newly developed benchmarking suite that uses predefined metrics and previous standard configurations as benchmarks. A workflow has been developed that runs JULES for all the available FLUXNET sites and utilizes existing meteorological driving data from PLUMBER2 (PALS Land Surface Model Benchmarking Evaluation Project 2). The suite then calculates statistical metrics for every site, variable, model configuration and benchmark. Each configuration is ranked relative to the benchmarks and these rankings are averaged over all statistics and sites to give an average ranking for the variables separately. A final averaging is performed over all variables to give an overall ranking. This method allows a clear comparison to be made on the performance of the new configuration relative to the benchmarks.
This talk will outline the development cycle used to generate the new standard standalone JULES configurations and demonstrate how the benchmarks have been used to assess the suitability of new science code.
ORAL PRESENTATIONS
ACCESS-AE: a national, km-scale numerical weather prediction ensemble system
Shaun Cooper Bureau of Meteorology
The Bureau of Meteorology is developing a km-scale numerical weather prediction system for a domain that encompasses the entire nation. The system consists of a deterministic model ACCESS-A, and an ensemble system ACCESS-AE.
ACCESS-A is a 90 vertical level, 1.5 km horizontally grid spaced model that will provide rapid update, short-range forecasts. ACCESS-AE is a 12-member ensemble system with 90 vertical levels and a horizontal grid spacing of 2.2 km. The ACCESS-AE convective-scale ensemble system quantifies uncertainties in the forecast and enables the probabilistic prediction of high impact weather events across the entire country.
ACCESS-A and ACCESS-AE use the Regional Atmosphere and Land configuration of the UK Met Office Unified Model (UM) called RAL3.2. This regional model configuration unifies the previously separate mid-latitude and tropical RAL configurations. RAL3.2 has been shown to have an improved representation of convection, more realistic precipitation distributions compared to previous configurations, and a reduction in 10m windspeeds that results in improved nocturnal winds. The surface characteristics of these models have higher resolution datasets than the Bureau’s previous km-scale models to define properties such as the land-sea mask and urban fractions. As part of the ACCESS-AE development several high impact weather case studies have been run and assessed, along with some weeklong trials. Initial evaluation and verification of these model forecasts will be presented.
Process-focused evaluations of GCM performance using ACCESS
and other CMIP6 simulations: An example with ENSO
Harun Rashid
CSIRO
The evaluation of climate and earth system model performance is a key part of model development. Usually, a selection of metrics relevant to the climate feature of interest are calculated from single-model simulations and compared with similar metrics calculated from observations. Using this approach, much can be learned about the climate feature, and improved scientific understanding can also benefit model developers by improving the relevant aspects of model performance. However, improving the assessed model biases in many emergent properties, such as El Niño–Southern Oscillation (ENSO), is more difficult. It would help if more model evaluation studies provided deeper process-focused information about the investigated climate feature, potentially benefiting the model development process.
Here, I illustrate this approach using the evaluation of ENSO performance in ACCESS-CM2 and other CMIP6 simulations as an example. This approach is based on using a multi- or single-model large ensemble, in which the basic-state and anomalous ENSO properties show large variations across the ensemble members. For a single-model ensemble, this could be achieved by including simulations from multiple experiments with different external forcings or perturbed parameter sets. The key ENSO metrics and the relevant basic -state properties are then calculated for each ensemble member and correlated across all ensemble members. This reveals the basic-state biases, which are arguably easier to fix, that are most correlated with the biased ENSO properties and processes. I will present and discuss results obtained from the ACCESS-CM2 single-model and CMIP6 multi-model ensembles.
ATMOSPHERE
ACCESS-CM3 Update
Kieran Ricardo ACCESS-NRI
This presentation introduces the latest developments in the ACCESS coupled model, ACCESS-CM3. The new version, CM3, integrates updated versions of all model components and adopts the NUOPC coupling framework, leveraging the established NUOPC support in MOM6 and CICE6. This talk provides an overview of the NUOPC framework and outlines the coupling approach implemented in CM3. Preliminary findings from the model's initial runs will be shared, alongside ongoing work and future developments.
Joint modulation of coastal rainfall in Northeast Australia by local and large-scale forcings: Observations versus AUS2200 simulations
Thi Lan Dao
The University of Melbourne
This study investigates the interaction between large-scale and local-scale forcings in regulating precipitation and its diurnal variation over coastal areas in Northeast (NE) Australia using the convection-permitting UK Met-Office Unified Model simulations with a horizontal grid-spacing of 2.2 km (AUS2200). The AUS2200 simulates well the spatial distribution of rainfall and its variation with large-scale variabilities such as El Niño-Southern Oscillation (ENSO) and the Madden Julian Oscillation (MJO) over Northeast Australia during a total 180 simulations days. Over the coastal areas, inhomogeneous rainfall patterns are evident in both radar observations and model simulations. By classifying the characteristics of offshore and onshore rainfall propagation, we found that the rainfall propagation modulates the average rainfall patterns. Modelling results suggest that the large-scale background wind and local-scale land-sea breeze circulations are two important factors driving rainfall propagation. For offshore rainfall propagation, a dominant propagation type over NE Australia, rainfall is triggered during the afternoon by strong sea breezes, and then propagates offshore during the nighttime with the strong upper-level westerly wind. In contrast, onshore rainfall propagation occurs during days with strong background easterlies from the surface to the upper levels. Rainfall tends to occur and stay over the land during days with strong sea breezes and weak background upper-level westerlies. On days with strong low-level easterly winds and strong upper-level westerly winds, rainfall is mainly concentrated over the ocean. Background wind regimes associated with different phases of the MJO modulate the direction and strength of rainfall propagation, leading to different coastal rainfall patterns.
Towards a convection-permitting reanalysis for Australia
Chun-Hsu Su Bureau of Meteorology
The new version of Bureau of Atmosphere regional atmospheric reanalysis (BARRA2) comprises of a 12 km full-input reanalysis (BARRA-R2) over Australasia. BARRA-R2 employs a 4D-Var assimilation technique with the ACCESS model (UK Met Office Unified Model and JULES land model), assimilating both conventional and satellite observations. The analyses from BARRA -R2 drive a convection-permitting regional ACCESS model (BARRA -C2), which provides information at a 4 km scale. While this approach shows promise in delivering new insights into high precipitation extremes, organised storm cells, and convective wind gusts, it also inherits biases from BARRA-R2 and exhibits other model biases including an overestimation of high rain rates. Looking ahead, the focus will be on developing convective-scale data assimilation reanalysis, overcoming the limitations of Australia's relatively sparse observation coverage, to generate more climatologically consistent data. With the release of BARRA2 data, here I will present the use of ACCESS for delivering higher resolution reanalysis for the Australian community, and highlight the areas for community contributions for future advancement to address.
Introducing mCDrive: the marine carbon dioxide removal integrated validation experiment.
Tyler Rohr IMAS/UTAS
To keep global warming below 1.5°C by 2100 the 6th IPCC Assessment Report considered 230 pathways and all of them required carbon dioxide removal (CDR) technologies. Accordingly, CDR has emerged as a possibly trillion-dollar industry that is both morally-imperative but highly invasive. Many of these technologies intend to engineer the oceans by biologically or chemically manipulating the marine carbon cycle (mCDR). Before we can responsibly implement mCDR for climate intervention we must understand its impact on marine ecosystems and its efficiency for carbon sequestration. Adhering to the precautionary principle, this will require a comprehensive modelling effort to evaluate mCDR technologies in coupled ESMs before they can be tested at scale in the open ocean. As part of a recent DECRA and philthropic grant, I am leading a suite of coordinated modelling experiments in ACCESS-ESM to assess the impact and efficiency of mCDR deployed on different timelines, with different technologies, and alongside different emissions pathways. In this talk I will review 1) the primary strategies available for mCDR, 2) the key processes, feedbacks, and uncertainties that will underwrite their viability, 3) the ongoing developments being made to WOMBAT to enable ACCESS-ESM to resolve all requisite processes, and 4) the suite of CDR deployment scenarios that I plan to simulate. My hope is that others will be able to leverage/engage with these experiments and I am eager to collaborate with the community to further develop scenarios that are of interest to others on the ACCESS community.
ESM A Slowdown
Jingwei Zhang UTAS
of Subsurface Freshening in the Southwest Pacific Since 1990
The Southwest Pacific, as the western boundary of the South Pacific Subtropical Gyre, has exhibited notable subsurface freshening trends over the past decades. However, using an observation-based reanalysis dataset, we demonstrate that the subsurface freshening in the Southwest Pacific Ocean (160oE-160oW, 20oS-40oS and 300-700m depth) has slowed down in the most recent three decades (1990-2020), occurring at a rate less than one-fourth of that observed during the prior three decades (1960-1990). To understand the drivers of this subsurface salinity variability, we conducted flux-form model experiments using a global ocean-sea ice model, specifically targeting the distinct impacts of various surface fluxes. Our results indicate that this multidecadal variability of subsurface salinity is predominantly induced by surface wind stress (~113%), through deepening isopycnal and anomalous Ekman transport. The contribution from the surface heat flux ( ∼32%) has comparable magnitude to that of surface freshwater flux but with opposite sign ( ∼-45%), leading to a relatively muted combined contribution from combined buoyancy forcings. Further empirical orthogonal function (EOF) analysis suggests that the Interdecadal Pacific Oscillation likely contributes to this multidecadal variability through modifying climate variables such as SST, net precipitation, and large-scale atmospheric circulation. The distinct contribution of individual surface forcings and perturbation experiments conducted here provide a reference for identifying the impact of climate variability on ocean-interior hydrography.
Pacific and Atlantic modes of deep-water formation in the Miocene climate optimum
David Hutchinson Climate Change Research Centre, University of New South Wales, Sydney
We are investigating modes of deep-water formation in the Miocene climate optimum (approx. 15 million years ago); a warm interval which was up to 7 °C warmer than present day and elevated CO2 concentrations. We are currently running simulations of this period with the coupled climate models GFDL CM2.1 and ACCESS-ESM1.5. These simulations require extensive changes to the boundary conditions, including the land-sea mask, topographybathymetry, vegetation, river runoff and greenhouse gases. We have found that the Miocene can exhibit either North Pacific or North Atlantic deep-water formation, depending on the details of the paleogeography. The Arctic -Atlantic gateways are critical to controlling the North Atlantic freshwater balance, which in turn determines the preferred basin of sinking. Furthermore, we show that hotter conditions (3x pre-industrial CO2) can trigger North Pacific sinking, while colder conditions (1x pre-industrial CO2) can trigger North Atlantic sinking. The results may help to explain the longterm evolution of the meridional overturning circulation, which has existed in both Pacific - and Atlantic- sinking modes according to geological proxies of deep ocean circulation.
LIGHTNING TALKS
Improving testing in CABLE: Range checks for variables
Abhaas Goyal
ACCESS-NRI
A new feature in CABLE has been implemented, which adds streamlined functionality for checking input and output variables at certain timesteps against valid ranges. This sanity check helps with improved testing, providing the enduser with logs and faster feedback loop to develop new features more accurately. The user can set a flag for doing range checks, which include:
- No checks for bounds.
- Checks at every timestep.
- Checks only when writing to the output file.
In case the range check fails, the user can also set a switch to (a) provide warning with logs or (b) exit the program. The first step of the work was identifying and grouping different types of variables across the codebase. The next step was defining stages, which now includes:
1. Constant parameters before running the spin-loop.
2. Constant parameters when creating the NetCDF output file.
3. Variables in creating the restart file (including land-only grid cell).
4. Variables at every timestep or only when writing to the output file (set with user flag). We also did additional refactoring work related to writing output to improve readability/maintainability. This includes:
- Making output inclusion type dependencies clear between group and individual values.
- Decoupling range checks and writing variables to the output file.
- Applying DRY and functional programming principles by introducing functions for (a) aggregating temporary output variables and (b) conditional checks before writing.
Finally, regression testing was done using benchcab by doing bitwise-comparison of the feature branch with the main branch using default CABLE configurations.
LIGHTNING
SVN to Git: lessons learnt
Claire Carouge
ACCESS-NRI
Authors: Claire Carouge, Ian Harman, Juergen Knauer, Jhan Srbinovsky, Paul Leopardi
The UK Met Office (UKMO) is actively preparing to transition from SVN-based version control to Git-based. This change will increase the visibility and access to the UKMO models, tools and documentation. However, this is a major change that will impact all developers of the UM, JULES and associated tools. Recently, we have gone through a similar transition by moving CABLE to git and GitHub.
In this poster, we propose to highlight the benefits and possible pitfalls of moving from SVN to git such as we saw when moving the CABLE’s source code. Although we saw some clear benefits on collaboration very early on, it also became clear that detailed contribution guidelines and training are required as git and GitHub can be used following subtly different workflows.
AMAMI, a Multitool Auxiliary Interface for ACCESS Models
Davide Marchegiani
ACCESS-NRI
The existing tools for interacting with ACCESS Models are complex, fragmented, outdated, poorly documented, and often exist in multiple simultaneous versions. Additionally, many standard operations, particularly those involving UM fieldsfiles, are not fully supported.
AMAMI (ACCESS Models Auxiliary Multitool Interface) is a Python interface designed to simplify interactions with ACCESS Models. It provides streamlined access to common workflows and operations through a series of commands, all within a single, well-documented, and versioned package.
LIGHTNING
Ocean Warming and Antarctic Sea Ice Loss
Edward Doddridge IMAS
Can we use a coupled climate model to assess the sensitivity of Antarctic sea ice to subsurface ocean warming? Following a dramatic decline in 2015 and 2016, Antarctic sea ice has consistently been below the long-term climatology. Recent work has suggested that subsurface warming in the Southern Ocean helped push Antarctic sea ice into this new low coverage state. We use the Australian Community Climate Earth System Simulator – Earth System Model v1.5 (ACCESS-ESM1.5) to investigate the response of Southern Ocean sea ice to subsurface ocean warming. Through ensembles of initial condition perturbations, we identify the magnitude and timescale of the resulting sea ice reduction.
The ACCESS-Hive: what is it and how do I use it?
Jasmeen Kaur
ACCESS-NRI
The ACCESS-Hive is an online, publicly accessible portal designed for the ACCESS community to learn, convene, and engage in meaningful discussions regarding the utilization of ACCESS models and datasets. There are two components: (1) Hive Docs - a central repository of model-related documentation, (2) Hive Forum - a discussion platform aimed at promoting collaboration among researchers. The Hive Docs was established to consolidate existing ACCESS Community documentation into a unified location, and hosts information about tools released and curated by the ACCESS-NRI team, in addition to resources contributed by members of the ACCESS community. On the Hive Forum, researchers can ask questions, initiate discussions with peers, and organize productive knowledge-sharing endeavors. In this interactive poster, we will walk visitors through the online portal, and seek feedback on how to improve this community resource.
Model Analogues: a statistical forecasting technique to explore the climate system
Jemma Jeffree ANU
Models are valuable, but computationally expensive, tools to explore the earth system. The computational cost means that it is preferable to use model simulations for multiple purposes and that the scope of some research projects is constrained by what is feasible to run. This presentation will introduce a technique model analogue forecasting to address both these points, with two example use cases.
Model analogue forecasting is related to initialised model forecasting. We select months from a pre-existing long model run, such as a control run or large ensemble, which approximate some target initial condition. The analogue forecasts are made by tracking the way these months evolve in the existing model run. Previous work has shown that they exhibit similar skill to an initialised forecast.
These forecasts are much faster and cheaper than traditional initialised forecasts, and can be used to broaden pacemaker-type studies to explore multiple models, lead times and initialisation months. In our first example, we explore whether analogue forecasts of the El Niño Southern Oscillation are improved by considering information from regions outside the tropical Pacific, and then infer how these regions contribute to ENSO evolution.
Furthermore, model analogue forecasts offer some advantages over a traditional forecasting system, such as being able to look backwards in time to see what the consistent precursors of a particular event might be. In our second example, we show preliminary results identifying high-rainfall months with similar spatial patterns in ACCESS-ESM1.5, and investigate the oceanic precursors of these high-rainfall events.
LIGHTNING
Implementing Sea Ice Thermodynamics from Laboratory Experiments into CICE
Kial Stewart
ANU
Sea ice exists in some of the most inaccessible, inhospitable, and extreme environments on Earth, meaning direct observations of sea ice are scarce, biased, and uncertain. This dearth of in situ sea ice observations inhibits our knowledge of the complex, finescale thermodynamic properties of sea ice and their accurate representation in numerical sea ice models. These properties – which include the temperature, salinity, brine fraction, porosity, and thermal conductivity – control how rapidly sea ice grows, to what thickness, and how rapidly it melts in the spring. Here we employ a novel laboratory apparatus to grow sea ice in controllable conditions and directly measure its thermodynamics under a wide range of atmospheric forcings. The sea ice thermodynamics model obtained from the experiments exhibit substantial differences to existing models. We implement these new sea ice thermodynamics into the CICE model in ACCESS-OM2, and investigate their impacts on the sea ice and ocean fields.
LIGHTNING
Running CABLE configurations with Payu
Lachlan Whyborn ACCESS-NRI
CABLE simulations often entail complex spin-up phases to bring the model to a quasi-equilibrium state, to act as a starting point for further simulations. These spin-up stages are currently handled in an ad-hoc manner, which is opaque to new users. A unified and adaptable interface for running complex multi-phase spin-ups has been built both for stand-alone execution,or through the payu tool for improved provenance.
LIGHTNING
High resolution urban and regional modelling with JULES+UM
Mathew Lipson
Weather of the 21st Century
The Bureau of Meteorology has recently developed new configurations of JULES+UM which better represent Australian and global cities at high resolution. These configurations are now openly available for use by researchers in the ACCESS community.
The configurations include the urban-canyon model "MORUSUS", urban landcover from the 10 m resolution "WorldCover" product, and the most recent science configuration for UM regional modelling (RAL3.2). Configurations at 300 m and 1.5 km resolutions over Sydney can be run “out of the box”, while other domains and resolutions can easily be established for any city or region globally.
A thorough analysis of urban-related impacts in Australia is yet to be undertaken. Therefore, much work remains to fully understand the interaction of regional weather and climate with Australian cities. Any interested researchers are encouraged to contact the author for assistance developing a configuration suitable for their needs.
LIGHTNING
Software Transformation at ACCESS-NRI: Empowering Climate Science with Advanced Computing
Micael J. T. Oliveira
ACCESS-NRI
At ACCESS-NRI, the newly established Software Transformation Team is dedicated to enhancing the computational capabilities of our climate scientists. Our mission is to optimize climate models for cutting-edge computer architectures and to facilitate the integration of emerging technologies, such as Graphics Processing Units (GPUs) and Machine Learning/Artificial Intelligence (ML/AI).
This poster presents an overview of the Software Transformation Team, its initiatives and plans. By collaborating closely with the ACCESS community, we hope to identify computational bottlenecks and redesign software to leverage parallel computing, GPU acceleration, and innovative ML/AI solutions that enhance model accuracy and predictive capabilities.
LIGHTNING
Using large ensembles for model evaluation
Nicola Maher ANU
Observations can be considered as one realisation of the climate system that we live in. To provide a fair comparison of climate models with observations one must use multiple realisations or "ensemble members" from a single model and assess where the observations sit within the ensemble spread. Single model initial-condition large ensembles (LEs) are valuable tools to complete such an evaluation. Here, we present the new multi-model large ensemble archive (MMLEAv2) which has been extended to include 16 models and 11 two dimensional variables. Data in this archive has been remapped to a common 2.5 x 2.5 degree grid for ease of inter-model comparison. We additionally introduce the newly updated climate model and diagnostics package for large ensembles (updated CVDP), which is the only such package designed specifically for use with LEs. By applying this package to the MMLEAv2 we highlight it's use for model evaluation against observations and model-to-model inter-comparisons. We demonstrate that for some metrics depending on which ensemble member is chosen, a model could evaluate poorly or favourably due to a large ensemble spread, and emphasise that for highly variable metrics, LEs present a much fairer model evaluation than either a single ensemble member, ensemble mean, or multi-model mean.
The ACCESS-NRI Training Program: Does it meet your needs?
Paige Martin ACCESS-NRI
Are our training programs hitting the mark? What would be your ideal vision for ACCESS-related training activities? Or maybe you just want to learn about the available training resources for the ACCESS community! This poster will detail our current training plan, outlining all of the training-related resources that ACCESS-NRI has to offer, including training events, online documentation, and other online learning resources. But we also want your input! Stop by our interactive poster to leave a note with your ideas and feedback about ACCESS-related training – the good, the bad, and your suggestions for improvement. Your honest feedback is crucial to help us refine and enhance our training initiatives, and to make sure that we are tailoring our events to YOU – the ACCESS community!
Opportunities to contribute to the ACCESS-ESM1.6 CMIP7 Fast Track submission
Rachel Law CSIRO
The World Climate Research Program’s Coupled Model Intercomparison Project phase 7 (CMIP7) has defined a set of ‘Fast Track’ experiments. These sit alongside the core ‘DECK’ experiments and have been chosen to feed into the next Intergovernmental Panel on Climate Change (IPCC) Assessment Report. As such, model datasets for these experiments are anticipated to be required by the end of 2026. In order to meet this timeline, experiments will initially be run with ACCESS-ESM1.6, an updated version of ACCESS-ESM1.5 as used for CMIP6. In parallel, the next generation ACCESS-ESM3, will continue to be developed, aiming for a slightly later CMIP7 submission. This poster will introduce the DECK and Fast Track experiments and planned timeline for completing selected Fast Track experiments. We are seeking community involvement to enable a wider range of Fast Track experiments to be completed, as well as in evaluation and analysis of any of the experiments.
LIGHTNING
The carbon and climate impacts of forestation in Australia
Tammas Loughran
CSIRO
Forestation is considered to be a feasible and cost effective strategy to remove CO2 from the atmosphere and store it in natural reservoirs. However, it is highly uncertain how much carbon can be removed by forests or how much they will affect the climate at the local to global scales. While Australia has some forestation potential to support meeting its Paris Agreement pledges, it is unknown how changes in the climate might affect this potential and how the associated decreased albedo of forested areas can offset the climate cooling effect achieved through increased uptake of CO2. We use the ACCESS-ESM1.5 Earth system model to perform idealized experiments of forestation to investigate the effects of forest cover has on the Australian climate at a range of global warming levels. Experiments include sensitivity tests by replacing all croplands or a specific fraction of croplands with up to 0.58 M km2 of forests. We find that forestation on these lands can remove up to 80 M tonnes of carbon per year and cool Australia's mean climate by ~0.4 °C, however, ACCESS-ESM1.5 projects some localized warming where forestation occurs. With careful forest management to account for localized warming, forestation has considerable potential in Australia to remove CO2 and contribute to meeting net-zero targets.
LIGHTNING
Model Configuration Management and Automated Testing at ACCESS-NRI
Tommy Gatti ACCESS-NRI
Model Configurations, released and supported by ACCESS-NRI, are stored within git repositories which allow changes to be automatically tracked.
Configurations are also versioned. Users can know that, if the major version number is the same, the configurations all produce the same answers- they are bitwise compatible. This is important when comparing experiments using these configurations, and when upgrading a configuration to a newer version. Generic automated testing using GitHub Actions is performed on these configurations to ensure metadata correctness, bit-reproducibility (if required) and efficiency (in future work). These checks are run when changes to the configuration are proposed (via pull requests), ensuring configurations perform as expected before they are merged.
Furthermore, scheduled self-tests are run on select configurations to verify that they continue to produce exactly the same answers over time. This verifies that nothing has changed in the HPC system that will affect answers.
This configuration management workflow produces high-quality model configurations and allows developers to skip menial tasks. It also provides certainty to users that any changes in the output of an experiment are a result of their choices, not due to an unexpected system or configuration change.
POSTERS
POSTER
POSTER
POSTE
POSTER
POSTER
POSTER
POSTER
meorg_client - connecting evaluation tools for the LSM community
Benjamin Schroeter
ACCESS-NRI
Migrating ACCESS projects from Trac and Subversion to GitHub
Paul Leopardi ACCESS-NRI
Tracking the provenance of ACCESS model data
Paul Leopardi ACCESS-NRI
Making ACCESS ESM1.5 configurations more robust, tested and flexible
Spencer Wong ACCESS-NRI
ACCESS-EMS-GBR: A regional coupled atmosphere-ocean model of the Great Barrier Reef
Rebecca Jackson, Matthew Woodhouse, Clothilde Langlais, Mathieu Mongin, Mark Baird CSIRO Environment
CICE6 adds Wave-Ice Interaction, Fast-Ice and Horizontally Resolved BGC
Anton Steketee ACCESS-NRI
The representation of urban land cover in the CABLE land surface modelling system
Jiachen Lu UNSW
POSTER
POSTER
POSTER
Exploring geothermal heat flow, subglacial hydrology, and ice flow in East Antarctica's Aurora Subglacial Basin
Felicity McCormack
Monash University
Enhancing ACCESS-OM3 with Integrated Wave-Ice Interactions
Ezhilsabareesh Kannadasan
ACCESS-NRI
Southern Ocean Sea Ice Changes and the Carbon Cycle: How worried should we be?
Ariaan Purich
Monash University
POSTER
POSTER
Joint modulation of coastal rainfall in Northeast Australia by local and large-scale forcings: Observations versus AUS2200 simulations
Thi Lan Dao
The University of Melbourne
Estimating the climate in a global model with small samples: how many do we need?
Sramana Neogi Monash University
POSTER
POSTER
POSTER
BARRA2: A new version of regional atmospheric reanalysis for Australia
Chun-Hsu Su Bureau of Meteorology
RRR: Reliability, Replicability, Reproducibility for Climate Models
Aidan Heerdegen
ACCESS-NRI
Model Configuration Management and Automated Testing at ACCESS-NRI
Tommy Gatti
ACCESS-NRI
POSTER
POSTE
POSTER
POSTER
POSTER
POSTER
POSTER
Understanding the role of extratropical eddies on the Southern Hemisphere Hadley cell edge
Julie Arblaster
Monash University
Role of distinct glacial boundary conditions in simulating a glacial climate state.
Himadri Saini University of Melbourne
The carbon and climate impacts of forestation in Australia
Tammas Loughran CSIRO
Running CABLE Configurations with Payu
Lachlan Whyborn ACCESS-NRI
SVN to Git: lessons learnt
Claire Carouge ACCESS-NRI
Using large ensembles for model evaluation
Nicola Maher ANU
The ACCESS-NRI Training Program: Does it meet your needs?
Paige Martin ACCESS-NRI
POSTER
POSTER
POSTER
POSTER
POSTER
POSTER
POSTER
Improving testing in CABLE: Range checks for variables
Abhaas Goyal
ACCESS-NRI
AMAMI, a Multitool Auxiliary Interface for ACCESS Models
Davide Marchegiani ACCESS-NRI
The ACCESS-Hive: what is it and how do I use it?
Jasmeen Kaur
ACCESS-NRI
Implementing Sea Ice Thermodynamics from Laboratory Experiments into CICE
Kial Stewart ANU
Opportunities to contribute to the ACCESS-ESM1.6 CMIP7 Fast Track submission
Rachel Law CSIRO
Ocean Warming and Antarctic Sea Ice Loss
Edward Doddridge IMAS
Software Transformation at ACCESS-NRI: Empowering Climate Science with Advanced Computing
Micael J. T. Oliveira
ACCESS-NRI
POSTER
If you had 3 data wishes, what would they be?
Clare Richards and Kelsey Druken
ACCESS-NRI
POSTER
Using pacemaker experiments to understand tropical basin interactions and associated impacts on Australian climate
Andrea Taschetto, ZoeGillett and Sebastian McKenna
UNSW
POSTER
POSTER
POSTER
POSTER
POSTER
POSTER
CMIP7 Data Request – contribute now
Sonya Fiddes (UTas) & Vanessa Hernaman (CSIRO)
UTas, CSIRO
COMPAS: A coastal unstructured ocean model using Voronoi meshes
Mike Herzfeld, Darren Engwirda and Farhan Rizwi
CSIRO
POSTE
Ice Sheet Modelling and the ISSM at ACCESS-NRI
Mike Tetley & Felicity McCormack
ACCESS-NRI
Developing a consistent view of future climate projections in the global coastal ocean
Clothilde Langlais
CSIRO
An Early Warning System for Australian Coastal Sea Level Extremes
Ryan Holmes, Oceane Richet, Ben Hague, Grant Smith and Claire Spillman
Australian Bureau of Meteorology, Research
High resolution urban and regional modelling with JULES+UM
Mathew Lipson
Weather of the 21st Century
Model Analogues: a statistical forecasting technique to explore the climate system
