UK /ɪkˈsper.ɪ.mənt/ US /ɪkˈsper.ə.mənt/
a test done in order to learn something or to discover if something works or is true
– From the Cambridge Dictionary of English
When most people think of experimental science, the image that comes to mind is usually someone in a white coat standing over a microscope and a petri dish. But when it comes to climate science, how do we conduct experiments when the entire planet is our laboratory? How do we answer fundamental questions about how the oceans and atmosphere interact, and temperature changes in one region might impact temperatures on the other side of the world, several years later?
The answer is through climate models.
Sophisticated computer models solve the fundamental physical equations that govern how oceans, sea ice, and the atmosphere interact. These models are developed by research groups around the world. They contain code which simulate (represent) a range of climate processes – from the formation of storm tracks that circumvent the globe to deep ocean circulation. The resulting simulations from these independently developed models are compared to one another and verified through the international Climate Model Intercomparison Project (CMIP).
Researchers from the Earth Systems and Climate Change Hub under Project 2.2 Enhancing Australia’s capacity to manage climate variability and climate extremes in a changing climate used Australian-developed climate models, such as the Australian Community Climate and Earth-System Simulator (ACCESS-CM2j, developed by Hub Project 2.1), CSIRO-Mk3L, and the Predictive Ocean Atmosphere Model for Australia (POAMA) to investigate the role of the Pacific Ocean in global warming, and how changes in the South Pacific Ocean can impact the Tropical Pacific.
Different climate models are used by researchers for different experiments depending on their strengths and weaknesses. For example, ACCESS-CM2j features a sophisticated atmospheric physics model coupled to ocean and sea-ice models and allows researchers to simulate processes occurring at a reasonably high spatial resolution. But this can make running the code computationally expensive, meaning that it takes a long time to simulate the Earth system on the National Computational Infrastructure’s supercomputer. In comparison, the CSIRO-Mk3L model runs at a lower resolution, which means that simulations can be run much quicker. This allows researchers to run, for example, millennial-long simulations, enabling them to study large-scale changes in the climate system that occur over multiple decades.
Using climate models in climate change research
Key findings from research using climate models to investigate climate variability and change undertaken in Project 2.2 is outlined below.
The role of the Pacific in global warming
The Pacific Walker Circulation is the term that describes the trade winds which blow from east to west across the tropical Pacific Ocean. In the past, climate models have projected that these winds will get weaker as the Earth continues to warm due to increasing greenhouse gas emission. However, between 1992-2011 the strength of the trade winds increased. This increase was accompanied by a 10-year plateau in the rise of global surface temperatures, although temperatures in the deep oceans continued to rise. This raised questions about the relationship between the Walker Circulation and global surface temperature.
To get to the bottom of this, Hub researchers joined forces with researchers from the ARC Centre for Climate Extremes to conduct experiments using three different climate models, including CSIRO-Mk3L. The experiments explored how decadal variability in the Pacific Walker Circulation influences the rate of global surface warming through a low-frequency climate phenomenon called the Interdecadal Pacific Oscillation (IPO). The IPO swings between two phases (a positive and negative phase) over decadal timescales. When the IPO is in its negative phase, the trade winds strengthen and global surface warming slows down. When the IPO is positive, the opposite happens, and trade winds weaken, and global warming accelerates.
Our team of researchers also found that they could partially predict when a positive or negative IPO would occur. By association, this means we can estimate if global warming trends in the next few decades will be faster or slower than average. This new knowledge can be used to help our industries, communities and business better plan and prepare for climate risks.
We are currently in a negative IPO phase (and are therefore experiencing a slower rate of warming). However recent observations suggest that the IPO is starting to shift towards a positive phase. It is therefore likely that we may be entering a phase of accelerated surface temperature warming.
The role of the South Pacific in modulating Tropical Pacific variability
The Tropical Pacific Ocean is particularly important to Australia’s climate because it is the ‘engine room’ of El Niño-Southern Oscillation (ENSO). ENSO has large impacts on Australia’s climate, including on rainfall, drought and extreme events, making it one of the most important climate drivers for our region. ENSO describes the fluctuation between the El Niño phase and the La Niña phase, with a neutral phase in between. El Niño conditions generally result in below-average rainfall over much of eastern Australia, while the La Niña phase generally results in above-average rainfall over much of Australia. Therefore, the Tropical Pacific Ocean is an important region to study in order to understand what causes and drives ENSO.
Recent research has shown that multi-year and decadal scale variability in the Tropical Pacific Ocean is influenced by variability in other regions, such as the South Pacific and Atlantic Oceans. But how is this occurring, and why? To address these questions, Hub researchers used the new ACCESS-CM2j coupled climate model to investigate how the different regions of the Pacific interact.
Our researchers simulated two climate systems: one where all the oceans were allowed to interact normally, and a second one in which variability in the South Pacific was ‘switched off’. In this climate system, South Pacific temperatures were held fixed and not allowed to vary from year-to-year. This allowed us to measure the impact of South Pacific variability on ENSO and the IPO. We found that with the South Pacific ‘switched off’, the IPO was disrupted and decadal temperature variability in the Tropical Pacific was reduced. We also found that switching off the South Pacific reduced the extent of sea-surface temperature variations during extreme El Niños and La Niñas.
What does this mean? These result shows that the climate processes originating in the South Pacific can have an impact on the magnitude of extreme ENSO events. These processes also affect the IPO, which can influence the rate of surface global warming. Understanding the causes and drivers of extreme ENSO events, as well as the drivers of decadal variability in the Tropical Pacific, is important in ultimately helping communities be better prepared for climate variability and change.
Possible changes to El Niño in the future, and how this may change the predictability of the Southern Annular Mode
Another important driver of Australian climate is the Southern Annular Mode (SAM), which is a wind belt over the Southern Ocean that encircles the globe. SAM heavily influences temperature and rainfall variabilities over different regions of Australia in different seasons. Previous research has shown that SAM and ENSO are linked in spring and summer. During El Niño events, warmer waters over the equatorial Pacific trigger a chain reaction in the atmosphere that leads to the wind belt expanding towards the equator (which is defined as the negative phase of SAM) for several months, increasing the likelihood of high temperatures and dry conditions over subtropical Australia. Because ENSO can be predictable 2-3 seasons in advance, this SAM and ENSO relationship allows researchers to use ENSO to predict what SAM may do in the future.
Over the last 60 years, observations show that sea surface and subsurface ocean temperatures have warmed faster in the western Pacific than in the east Pacific.
Sea surface temperatures have warmed faster in the west Pacific in the last 60 years. Observed September to December mean sea surface temperature trends from 1960-2014. Stippling indicates where the trend is statistically significant. Red colours indicate warmer sea surface temperatures, while blue colours indicate cooler temperatures
Hub researchers looked at how the observed extreme El Niño events experienced in 1982, 1997 and 2015 would behave under future climate conditions, assuming this current ocean warming trend continues. Using the POAMA model, they simulated these three extremes El Niño events alongside the observed warming trend. Their results suggest that under future conditions, the intensity of extreme El Niño events would weaken. These results imply that if the observed increases in ocean temperature continues, SAM and associated surface climate in the Southern Hemisphere, including Australia, would be less predictable during El Niño events.
Interestingly, this observed warmer-in-the-west-Pacific pattern differs from what most climate models project to occur in the future. Most models project that the equatorial Pacific will warm more in the east. If this were to occur, El Niño events would instead become more extreme. Much work has also been done by Hub researchers to investigate how ENSO and other modes of climate variability may change under these conditions.
Research undertaken in Hub Project 2.2 over the last three years has made a valuable contribution to the highly relevant and rapidly expanding knowledge base on Pacific variability. This research will continue under the Hub’s current Project 5.2. A deeper understanding of the drivers of Pacific variability, and in particular how this variability impacts projections of Australian temperatures and rainfall, is crucial in supporting the long-term planning and management of resources and risks associated with climate change and extreme events.
This research was conducted by Hub researchers from Project 2.2 Enhancing Australia’s capacity to manage climate variability and climate extremes in a changing climate. Project 2.2 ran from July 2016 until June 2019.
- Bordbar, MH, England, M, Sen Gupta A, Santoso, A, Taschetto, AS, Martin, T, Park, W, Latif, M. 2019. Uncertainty in near-term global surface warming linked to tropical Pacific climate variability. Nature Communications. 10, 1990, doi:10.1038/s41467-019-09761-2 | Full paper
- Chung C, Power SB, Sullivan A, Delage F. The role of the South Pacific in modulating Tropical Pacific variability. Scientific Reports (submitted)
- Lim EP, Hendon HH, Hope P, Chung C, McPhaden M. Continuation of tropical Pacific Ocean temperature trend will weaken linkage of Southern Annular Mode and extreme El Niño, Scientific Reports, 9, doi:10.1038/s441598-019-53371-3 | Full paper