Climate Models are our primary tools for understanding climate change and its likely impacts. Such models use observed data and mathematical equations to estimate the future climate either in the form of short-term forecast or long-term projections.
Unlike weather forecasts, which describe a detailed picture of the expected daily conditions starting from the present, climate models are based on probability, indicating areas with higher chances to be warmer or cooler and wetter or drier than usual. Climate models are based on global patterns in the ocean and atmosphere, and records of the types of weather that occurred under similar patterns in the past.
Climate models are based on well documented physical processes governing global circulation in the atmosphere and ocean. To project climate into the future, the initial conditions are set to reflect a finite set of possible future conditions called scenarios. Scenarios are possible stories about how quickly human population will grow, how land will be used, how economies will evolve, and the atmospheric emissions of greenhouse gases that would result for each storyline. In 2010, climate scientists agreed upon a new set of scenarios for the future concentration of atmospheric greenhouse gases. Collectively, these scenarios are known as Representative Concentration Pathways or RCPs. Each RCP indicates the amount of climate forcing, expressed in Watts per square metre, which would result from greenhouse gases in the atmosphere in 2100. Learn more about RCPs and climate implications of different emission pathways.
Global climate models, such as those used in the various assessment activities coordinated by the Intergovernmental Panel on Climate Change (IPCC), attempt to predict the climate systems response to human-induced increases in greenhouse gases among other factors. These models are known as fully coupled ocean-atmosphere models which means they simulate both oceanic and atmospheric processes and important connections between the two.
Over 40 global climate models were used to support development of the Fifth Assessment Report of Intergovernmental Panel on Climate Change.
The most recent projections for Queensland are based on the modelling completed by the Coupled Model Intercomparison Project Phase 5 (CMIP5), organized under the auspices of the World Climate Research Programme.
A common factor affecting the usefulness of a global climate model for predicting local climate conditions is their resolution, that is the ‘grid’ on which the calculations are based.
Global climate models typically use a grid cell size of 100-200km so they will project the same climate for any place within a given grid cell up to 200 km across, without reference to local conditions.
In order to better represent regional climate, a smaller grid cell size is used through a process of downscaling. With recent advancements in computation we can now simulate future climate with higher spatial resolution state-wide.
Let’s check the downscaling benefits locally by comparing global and regional climate simulations at a single global model grid cell.
Global climate models attribute the same climate to a large extent ignoring local topography and sea-land contrasts due to coarse spatial resolution limitations.
By improving spatial resolution, regional climate models better simulate local climate. Sea-land distinction and topography-driven processes like orography are clear advantages.
The high resolution climate change projections for Queensland were produced using a dynamical downscaling approach. Dynamical downscaling means running a dynamical model also known as a Regional Climate Model using output from the Global Climate Model as input. This is in contrast to other methods, such as statistical downscaling, which is based on statistical relationships between large-scale and small-scale variables.
In this approach a global variable resolution climate model CCAM (conformal‐cubic atmospheric model) developed by CSIRO was used (See Katzfey et al., 2016). Downscaling process consisted of two steps. In first step a global 50 km uniform resolution simulations with CCAM were completed using bias- and variance-corrected (see Hoffman et al., 2016) sea surface temperature as well as sea ice concentrations from eleven Global Climate Models for the period 1950 to 2099.
|CMIP5 model name||Model name||Institution Name(s)||Country of origin|
|ACCESS1-0||Australian Community Climate and Earth-System Simulator, version 1.0||CSIRO & BoM||Australia|
|ACCESS1-3||Australian Community Climate and Earth-System Simulator, version 1.3||CSIRO & BoM||Australia|
|CCSM4||Community Climate System Model, version 4||NCAR||USA|
|CNRM-CM5||Centre National de Recherches Météorologiques Coupled Global Climate Model, version 5||CNRM-CERFACS||France|
|CSIRO-Mk3.6||Commonwealth Scientific and Industrial Research Organisation Mark 3.6.0||CSIRO & Qld Govt||Australia|
|GFDL-CM3||Geophysical Fluid Dynamics Laboratory Climate Model, version 3||GFDL NOAA||USA|
|GFDL-ESM2M||Geophysical Fluid Dynamics Laboratory Earth System Model with Modular Ocean Model, version 4 component||GFDL NOAA||USA|
|HadGEM2||Hadley Centre Global Environment Model, version 2||Met Office Hadley Centre||UK|
|MIROC5||Model for Interdisciplinary Research on Climate, version 5||AORI Japan||Japan|
|MPI-ESM-LR||Max Planck Institute Earth System Model, low resolution||Max Planck Institute||Germany|
|NorESM1-M||Norwegian Earth System Model, version 1 (intermediate resolution)||Norwegian Climate Centre||Norway|
The global 50 km CCAM simulations were further downscaled in step two using global stretched version of CCAM with spatial resolution of ~10 km over Queensland region. These high resolution simulations were completed for the period 1980 to 2099.
Watch the video below of model simulations to see the advantages of the higher resolution.
Climate modelling has been steadily improving over the past decades. For the CMIP5 report, 40 models were used and climate projections are typically based on the output from an ensemble of models. Combining several model simulations into a model ensemble allows for an assessment of uncertainty.
Eleven high-resolution (~10 km grid cell size) simulations were completed for period 1980 to 2100 over the Queensland region using RCP8.5 emission scenario. The choice of CMIP5 models was dictated by the availability of global sea surface temperatures and sea ice at monthly intervals from the global models.
Climate Change in Australia Technical Report describes model evaluation and scoring based on performance of various CMIP5 models in Australian region. The models used in dynamical downscaling for Queensland region perform well and represent a good spread of models with different patterns of warming (Mizuta et al., 2014) and complexity of physical processes. These were selected as they are likely to best represent the range of uncertainty in climate change projections for Queensland.
The main advantage of illustrating results from all eleven models is in the transparency. This is vital when using the data to better estimate climate risk. For example, showing the output from all models allows users to not just look at model performance, but the full extent of variability across projections. This is more informative than just focusing on the median (50th percentile) and allows output for the upper (90th percentile) and lower (10th percentile) bounds to be considered as well. This allows for more realistic scenario- based testing of worst and best case examples.
To illustrate, we have used the full range of model estimates on mean precipitation to illustrate the effects of upper and lower bounds in dam water storage in the following interactive diagram.
Click on the different model buttons to see how it affects dam levels.
In addition to improved spatial resolution, our projections are continuous over the period 1980-2099 unlike other downscaled projections in Australia, which adopted time-slices modelling approach. Time-slices approach focuses on simulations covering selected discrete periods, such as 1991-2010, 2041-2060 and 2081-2100.
The video below compares the continuous simulations and time-slices approach.
The Queensland Future Climate Dashboard summarises information of 11 state-of-the-art climate models with regional scale simulations until the end of the current century. The dashboard is a visualisation platform composed of drop-down menus, maps, plots and tables whereby users can customise, visualise and export summarised future climate information according to their interest.
Use the map to find out more about a region.
Get fine resolution detail on each region and model output data.
Change the dropdowns to view data on the map and graphs for different regions, variables, seasons and years.
Click on the graphs to find out more about the model data.
Each bar shows the range and mean for all models, and the outputs for each individual model.
The bottom graph shows predicted indicator changes into the future.
The bottom graph shows predicted indicator changes into the future.
Data are available for everyone to use. It's best to first determine the level of data required for your needs. Summary reports and existing tables can be easily accessed through the dashboard.
The major purpose behind constructing these data sets is to aid in decision making in an environment of uncertainty. This is often referred to as climate change risk assessment.
Climate risk is the product of the consequences of climate change and the likelihood of those consequences. Queensland Fire and Emergency Services (QFES) is responsible for supporting the risk based planning across Queensland’s Disaster Management Arrangements (QDMA) through the Queensland Emergency Risk Management Framework (QERMF). The risk assessment methodology put forward by the QERMF illustrates how climate change projections can be used in risk assessment framework. The risk analysis considers likelihood of occurrence (L), vulnerability (V) and consequence or likely impact (C). QFES risk assessment methodology takes into account the climate risk obtained from both historical records and future simulations.