Chapter three

Climate change projections

Using models to project changes in rainfall, temperature, water availability and extremes in climatic variables enables us to identify the threats or risks imposed on agriculture. Our understanding of climate change, its impacts and the actions we can take to deal with it are changing and evolving rapidly. Global Climate Models (GCMs), the models from which projections are made, are improving and growing steadily more reliable. Projections of impacts and consequences are becoming more specific and options for action are becoming clearer.

Climate change projections

CSIRO Marine & Atmospheric Research released an updated report in 2006 on climate change (projections) under enhanced greenhouse conditions in South Australia (6), commissioned by the South Australian Government. The researchers selected 13 GCMs to produce projections of rainfall and temperature to 2030 and 2070. These 13 GCMs were selected from 23 as being those that best simulate observed average patterns of mean sea level pressure, temperature and rainfall (1961–1990) in the SA region.

Increased minimum, maximum and average temperatures are projected, along with increased frequency of extreme maximum temperatures, a decrease in the frequency of extreme minimum temperatures, increased evaporation rates, variations in rainfall patterns and increased occurrences of extreme events, such as floods, heat waves and high fire danger conditions.

Temperature change projections

Figure 9 shows the temperatures projected for South Australia by 11 of the global climate models. You can compare simulations for past years with observed temperatures as indicated by the dark black line. The graph shows that the temperature trends simulated by the eleven models predict the observed data reasonably well, therefore giving us some degree of confidence in their ability to predict temperature change into the future.

The models predicting the lower range of temperatures show temperatures may increase by 2oC by 2100, while those models predicting the higher range of temperatures show temperatures may rise by 4oC.

Figure 9: Observed and simulated temperature anomalies using 11 GCMs for South Australia, from 1860 to 2100. The annual anomalies are from a 30-year period (1975 to 2004) and smoothed by an 11-year running mean (6).

Figure 9: Observed and simulated temperature anomalies using 11 GCMs for South Australia, from 1860 to 2100. The annual anomalies are from a 30-year period (1975 to 2004) and smoothed by an 11-year running mean (6)

* Note an anomaly is defined as a difference from the average over the observed period. In the above case the observed period was 1975 to 2004. So the simulated anomalies reflect how much our average temperature may change after this period.

It is increasingly suggested that temperatures in the future are looking to be more likely to trend towards the upper end of the range of the TAR temperature projections that the CSIRO figures are based on. This is due to the impacts of:

  • A decrease in the reflectivity of the Earth’s surface due to the melting of snow and ice.
  • The release of extra carbon dioxide and methane from the terrestrial biosphere.
  • A projected decrease in the concentration of aerosols, which have a cooling effect in the atmosphere (6).
The following maps (Figure 10) illustrate the projected ranges of changes in temperature across South Australia for 2030 and 2070, for the annual averages, as well as each season.

The charts show the projected temperature change ranges (for 2030 and 2070) corresponding with the colours used in the maps. The ‘SRES’ charts (from the IPCC Special Report on Emission Scenarios - SRES) are based on a range of assumptions about population changes, energy sources, levels of regional or global co-operation and socio-economic arrangements, but do not include any specific greenhouse gas mitigation activities. The SRES scenarios were used as a basis for the climate projections in the Third and Fourth Assessment Report written by the IPCC (5).

The 550 ppm and 450 ppm charts, on the other hand, show the projections for 2030 and 2070 for scenarios where atmospheric CO2 concentrations are stabilised at 550 parts per million or 450 parts per million respectively. This shows the impacts on projections from reducing greenhouse emissions.
Using the wider SRES range of carbon dioxide concentrations, the findings from the CSIRO report summarised in the box below:
 
Shows Seasons temperatures by 2030 and 2070
 
Figure 10: Average seasonal and annual warming ranges (oC) for 2030 and 2070 relative to 1990 for SRES scenarios, and CO2 concentrations stabilised at 450 ppm by 2100 and 550 ppm by 2150. The coloured bars show ranges of change for areas with corresponding colours in the maps (6).
Figure 10: Average seasonal and annual warming ranges (oC) for 2030 and 2070 relative to 1990 for SRES scenarios, and CO2 concentrations stabilised at 450 ppm by 2100 and 550 ppm by 2150. The coloured bars show ranges of change for areas with corresponding colours in the maps (6).
 
The images show that temperatures inland are likely to increase more than in coastal regions. The further inland you go, the warmer it may become. If greenhouse emissions can be restricted, and increases in atmospheric concentrations mitigated, the temperature changes projected at the other end of the scale are reduced. However the lower end will not change.

Rainfall change projections

While a warmer world is overall expected to be a wetter world, in latitudes such as southern Australia projections point towards drier conditions. For farmers, this projection can be one of the most worrying aspects, but it is also the most uncertain aspect of the projections (7,8). There was a relatively high level of consistency between global circulation models used by the IPCC in 2001 with respect to the drying trend in southern Australia. This trend is similar across all Mediterranean climates, as shown in Figure 11. This consistency is repeated in the latest round of models used for the IPCC report due during 2007. Previous assessments of rainfall change over Australia (e.g. CSIRO, 2001) have all indicated that the potential for rainfall decreases, particularly in winter, but with increases in rainfall possible in summer. Figure 12 shows the actual and simulated rainfall using 11 of the global climate models. For past years, you can compare the simulated rainfall with observed rainfall shown by the dark black line. Unlike the temperature projection figure, the observed data does not follow the rainfall trends simulated by the eleven models at all well. This emphasises the fact that the rainfall projections are most uncertain for South Australia.
 
Figure 11: Inter-model consistency in direction of simulated annual rainfall change in ten global climate models. Large changes are occurring where the average change across the models is greater in magnitude than 5% per OC global warming (5, 9

Figure 11: Inter-model consistency in direction of simulated annual rainfall change in ten global climate models. Large changes are occurring where the average change across the models is greater in magnitude than 5% per OC global warming (5, 9)

Evaporation levels, combined with rainfall figures, are very important to agriculture in determining the net water balance and the amount of water available to plants. In South Australia a small increasing trend in annual pan evaporation has been recorded since 1970.

Figure 12 shows that rainfall projections do not seem to be strongly increasing or decreasing. Rainfall trends differ from location to location and large decadal fluctuations in rainfall means it will be difficult to detect an enhanced greenhouse signal in rainfall from the natural variability until after 2050. It is important to look in more detail at the trends for individual locations in your region, available from the CSIRO report (6).
 
Figure 12: Observed and simulated rainfall anomalies using 11 global climate models for South Australia, from 1860 to 2100. The annual anomalies are variations from the average over the 30‑year period from 1975 to 2004 and are smoothed by an 11-year running mean (6). 
 
Figure 12: Observed and simulated rainfall anomalies using 11 global climate models for South Australia, from 1860 to 2100. The annual anomalies are variations from the average over the 30‑year period from 1975 to 2004 and are smoothed by an 11-year running mean (6).

The following maps (Figure 13) show the projected changes in rainfall across South Australia to 2030 and 2070.
 
Figure 13: Average seasonal and annual rainfall change (%) for 2030 and 2070 relative to 1990 for SRES scenarios, and CO2 concentrations stabilised at 450 ppm by 2100 and 550 ppm by 2150. The coloured bars show ranges of change for areas with corresponding colours in the maps (6)
 
Figure 13: Average seasonal and annual rainfall change (%) for 2030 and 2070 relative to 1990 for SRES scenarios, and CO2 concentrations stabilised at 450 ppm by 2100 and 550 ppm by 2150. The coloured bars show ranges of change for areas with corresponding colours in the maps (6)
 
It is likely that rainfall will decrease in areas shaded in red or dark orange and increase in the areas shaded in green and grey (although it is much less certain in the grey areas). According to these projections, decreases in rainfall are most probable along the coast and in spring. These charts show the level of uncertainty still associated with rainfall projections at this stage.

As with temperature projections, the ‘SRES’ charts (from the IPCC Special Report on Emission Scenarios - SRES) are based on a range of assumptions about population changes, energy sources, levels of regional or global co-operation and socio-economic arrangements, but do not include any specific greenhouse gas mitigation activities.

The 550 ppm and 450 ppm charts, on the other hand, show the projections for 2030and 2070 for scenarios where atmospheric  CO2  concentrations are stabilised at 550 parts per million or 450 parts per million respectively. This shows the impacts on projections from reducing greenhouse emissions.

Water balance and evaporation projections

One of the most important interactions between rainfall and temperature is the moisture balance. Generally as temperature rises evaporation increases. The small rise in average temperatures in Australia has generally not been accompanied by very slight rise in evaporation. There are a number of possible reasons for this with changes in wind and changes in instruments for recording evaporation being considered the most likely (20, 23).

It is expected that further rises in temperature from global warming will be associated with increased evaporation and decreased soil moisture. This would exacerbate the consequences of a drying trend.

Extreme events projections

  • A range of extreme events is expected to occur under conditions of climate change and may already be evident, including unusually violent storms, high winds, extreme storm surges, more intense heatwaves, bushfires, drought and flooding.
  • The frequency of extreme maximum temperatures will increase while the frequency of extreme minimum temperatures will decrease.
  • The frequency of hot spells above 35°C and 40°C are projected to increase across most of South Australia with the largest increases in the north.
  • Despite decreases of up to 30% in average rainfall over parts of South Australia in some seasons, the incidence of heavy rainfall is projected to increase by 0 to 10%. The specific weather patterns associated with heavy summer rainfall in the north of the state are projected to increase both in terms of frequency of events and magnitude of rainfall, with a projected 20% increase in flood frequency in northern South Australia.
  • All climate models show an increase in the frequency of droughts in Australia towards the end of this century (9).

Extreme events at the coast

  • Storm surges of at least half a metre in height occur year round along the South Australian coast with the greatest frequency of events occurring during the winter and spring months. They are caused by the westerlies or south-westerlies following the passage of cold fronts and their associated mid-latitude low pressure systems further to the south.
  • The frequency of winter time low pressure systems (lows) is projected to decrease by about 20% in the vicinity of South Australia under enhanced greenhouse conditions. Central pressures of the most extreme lows in model projections were lower by about 2 hPa on average, indicating slightly more intense lows under enhanced greenhouse conditions. Accumulated rainfall accompanying the lows in the South Australian region decreased by between 10 and 20% under enhanced greenhouse conditions owing to fewer low systems occurring. The amount of rainfall per low, however, tended to increase by up to 10% over the Bight and coastal regions in the western half of the state. The frequency of mid-latitude lows in spring increased by 2% while the most extreme lows deepened by about 1hPa (6).
  • Extreme wind speeds are projected to decrease across much of South Australia and the Bight in winter and summer. Increases in extreme wind speed occurred in model projections over the north of the state in autumn and while spring decreases occurred in the south of the State and over the Bight.
  • Examination of wind direction changes, particularly in westerlies, south-westerlies and southerlies, that can be responsible for storm surge occurrence in South Australia, revealed only minor changes in winter in South Australia with south-westerly coastal regions tending towards decreases in frequency. However, there were relatively larger increases in frequency of westerlies, south-westerlies and southerlies in eastern coastal regions in spring. Patterns of change were qualitatively similar in autumn but weaker than in spring while in summer, increases occurred only in the frequency of southerlies (9).
  • In addition, the sea level has been projected to rise by 9 to 88 centimetres by 2100 (9). However, recent observations have shown that Arctic and Antarctic ice is being lost at a rate significantly greater than that which had been projected, which means there is a risk of considerable increase in sea level rise projections.