This page provides guidance on the use of climate model simulations from the 2014 NARCliM projections. This data is available from a suite of 12 different simulations. The projected changes from each of these simulations are considered equally plausible.
The 2014 NARCliM projections deliver two different types of data from the model simulations:
- Changes e.g. ‘temperature will rise by 3°C’
The AdaptNSW website has interactive maps that show climate changes (i.e. the difference between now and in the future). You can download the data that has generated those maps from NSW Climate Change Downloads. The changes shown via AdaptNSW are calculated from the multi-model average of 12 different climate model simulations (see NSW Climate projections map for 2060-2079). While useful for summarising, this multi-model average by definition does not show the values from individual simulations and therefore does not show the full range of possible futures from the 12 models.
- Actual model data
The 2014 NARCliM data available through the Climate Data Portal provides the actual values of climate variables from the 12 global climate model (GCM)-forced model simulations and the three reanalysis-forced simulations. Actual values can be converted into change values by subtracting present actual values from future actual values.
Table 1: Examples of model data in comma-separated values (CSV) and American Standard Code for Information Interchange (ASCII) format from the Climate Data Portal.
If you would like more information on how the modelling data has been generated please read the About NARCliM page. If you are unsure about what data you need to consider for your application please email us: email@example.com
Using NARCliM data
The model simulations used for the 2014 NARCliM projections were chosen to span the range of possible futures span the range of possible futures for temperature and rainfall under the A2 emissions scenario. Assessments of the impacts of climate change on your systems and services should therefore use climate data from the range of NARCliM simulations. We suggest the following five-step process to decide what data is best for your assessment.
Step 1: Decide which climate variables you need to examine for your system.
Step 2: Explore a range of future climate changes.
Step 3: Select NARCliM simulations for your analysis.
Step 4: Download the climate data for your impact assessment.
Step 5: Consider the limitations of modelling.
Step 1- Decide which climate variables you need to examine
How does the climate currently affect your system?
To assess the impacts of climate change that you are interested in, you first need to know which aspects of the climate your system is currently sensitive to. It is likely that you already know this but, if not, considering the following questions can help.
What climatic events have led to impacts in the past? For example, have severe impacts resulted from heatwaves in the past? Could it be that daily maximum temperatures are relevant?
Are you using a model in your work (statistical or numerical)? Does your model require inputs of multiple climate variables? For example, a hydrological model for estimating river flow might require precipitation and evaporation data as inputs.
Step 2 - Explore range of future climate changes
Now that you know what climate variables are going to impact your system, what direction of change will impact it the most or have the greatest risk to your system (e.g. increasing rainfall or decreasing rainfall – or both).
The 2014 NARCliM projections has generated 12 modelled futures. For a first-pass risk assessment, or as a prelude to more detailed impact analysis, we suggest that you explore the results below to understand that different models are projecting different changes in temperature and rainfall.
Figure 1 provides information about projected future changes in annual mean temperature and precipitation averaged over the whole of New South Wales. This information can help you decide on an ensemble that you want to test on your system (e.g. hotter ensemble or a wetter ensemble). This diagram shows:
- CSIRO-MK3.0 ensemble is projecting a decrease in average rainfall (drier) and an increase in average temperature (hotter).
- ECHAM5 ensemble is projecting a larger increase in temperature (hotter) with little change in precipitation.
- MIROC3.2 ensemble is projecting an increase in precipitation (wetter) and an increase in temperature (hotter).
- CCCMA3.1 ensemble is projecting a larger increase in precipitation (wetter) and a larger increase in temperature (hotter).
Figure 1: Changes in NSW-average annual mean temperature and annual mean precipitation between 1990-2009 and 2060-2079 for all 12 of the 2014 NARCliM projections and the multi-model mean.
Spatial and temporal variation
It is important to remember that future climate changes will be different for a different future time period, different seasons of the year and different locations within NSW.
The NARCliM Climatological Atlas (PDF 53MB) provides all the maps of the changes in annual and seasonal mean temperature and precipitation for the near and far future based on all 12 individual simulations. This information allows a more comprehensive comparison of the simulations at locations within NSW.
Figures 2 and 3 are examples from the NARCliM Climatological Atlas showing changes in annual mean temperature and annual mean precipitation between 1990-2009 and 2060-2079. In these figures you can see that the projections vary spatially. It is important to consider this in your selection of models for your region.
Figure 2: Maps of changes in annual mean temperature between 1990-2009 and 2060-2079 for the 12 simulations.
Figure 3: Maps of changes in annual mean precipitation between 1990-2009 and 2060-2079 for the 12 simulations.
Step 3 – Select the simulations for impact analysis
Now that you understand which models are projecting changes that show the greatest risk to your system, you need to select models for your analysis.
After looking at the range of possible futures in Step 2 you will be better able to select what simulations will suite your impact assessment. To capture the range of risk your system may face under climate change, you should aim to select a set of simulations that are likely to capture the largest changes that could arise from the full set of 12 simulations. For example, the wettest and hottest scenario for your region or driest scenario for your region.
You might perform an analysis of all 12 simulations. However, this is often unnecessary. For example, communication of your results may be more effective if you communicate impacts for a smaller number of future climate scenarios. It may also be impractical. For example, if you are using a complex numerical impact model to translate future climate data into impact data (e.g. a hydrological or crop simulation model), then you may be constrained by the amount of time this model takes to run on the computers available to you. In these cases, you can use a diagram like Figure 1 to select a smaller number of NARCliM simulations to analyse.
Step 4 – Access the climate data for analysis
Now that you know the models you want, you need to download the data, in the right format! What data format does your model use?
The Climate Data Portal provides data in a range of formats to accommodate different data users. This includes comma-separated values (CSV), network Common Data Form (NetCDF), and GeoTif formats.
Step 5 – Consider the limitations of the projections your analysis
Now you have the data, you need to know models are limited in what they can do. How will this affect your results? You also probably know your model/analysis has limitations too. What do those limitations mean when analysing climate change impacts?
In drawing conclusions about the impacts of climate change from the 2014 NARCliM data, it is important to consider the limitations of the NARCliM simulations and the methodology that you have applied to it. Specifically, it is important to understand whether any of these limitations will affect your results and, if so, how. Some of the questions below may be relevant here, but others that are not listed may also be relevant.
Is your study looking at global scale systems e.g. El Ninos events?
If so, are you aware of what global climate models (GCM) have been used in the modelling and how they were able to simulate the larger system processes?
Have you been able to account for all relevant biases in the NARCliM simulations?
If you are a specialist user, you may be concerned about bias. It is important to understand how any biases in variables from the NARCliM simulations may affect your analysis. You can test whether there are biases in NARCliM data by comparing them with observations of your variable. If you do not have observations then this is a limitation you need to be aware of when analysing the impacts.
Temperature and precipitation have been bias corrected and are available from the Climate Data Portal. More information on the method applied in the bias correction of temperature and precipitation is available in the NARCliM Technical Note 3. For a more in-depth analysis of each model, please see: Performance of models.
If you discover a bias in another variable, we suggest two techniques for generating data suitable as input to impact models:
- Bias correction – actual values of climate variables from the climate model simulations for a historical period and corresponding climate observations are compared. Statistical relationships are derived that transform the climate model data into data that more closely resembles the observations. These relationships are then applied to the climate model output for future time periods.
- Scaling observations – actual values of climate variables from the climate model simulations are not used directly. Instead, they are used to calculate future climate changes between a recent time period and a future period. The changes are then used to modify climate observations of the recent period.
Both of these techniques have advantages and disadvantages and rely on accurate observations of the climate variables of interest. Neither technique can be used if observations are not available.
Have you had to make any assumptions about future climate conditions in your analysis that may prove to be untrue?
For example, have you assumed that future changes in daily maximum temperatures will be the same as future changes in daily minimum temperatures?
Have you used an impact model that assumes that relationships between different climate variables will be the same in the future as they have been in the past?
For example, have you used a hydrological model that derives evaporation from temperature variables using a relationship developed from observations of the past climate?
More guidance on using climate model data
CSIRO has prepared a range of guidance material as part of their Climate Change in Australia website. Although some of this material is specific to the data presented in that website, much of it is intended for the general user of climate model data. We recommend users take the time to read this as it covers topics such as how to conduct an impact assessment and common mistakes when using climate change data.