Previous PageTable Of ContentsNext Page

Risk assessment of climate change impacts on Australia’s wheat industry

Mark Howden1 and Roger N. Jones2

1 CSIRO Sustainable Ecosystems, GPO Box 284, ACT, 2601.
CSIRO Atmospheric Research, PMB 1 Aspendale Vic. 3195.


We describe a systematic risk assessment of the combined effects of possible increases in atmospheric carbon dioxide concentrations and associated temperature increases and rainfall changes on the Australian wheat industry for the year 2070. The likely impacts of these three factors varied markedly among regions, with those in Western Australia having a high likelihood of significant yield reductions, but those in north-eastern Australia having a high likelihood of moderate increases in yield but also a small probability of substantial yield reductions. Nationally, whilst median yields are little changed (without adaptation) there is a significant risk to the industry as increases in crop value are limited (to about 10% or $0.4B p.a.) but potential losses are large (about 50% or $2B p.a.). Adaptation strategies of changing planting dates and varieties could be highly effective by offsetting the negative impacts of global change and enhancing positive aspects. The median benefit of these adaptations was about $225M p.a. but with a range of $100M to over $500M p.a. Identifying further adaptation strategies and building the capacity in the farming community to implement them will be a key pathway to deal with uncertain climate and atmospheric changes in the future.

Media summary

Increases in atmospheric CO2 and associated climate changes pose a significant risk for the Australian wheat industry. Adaptation strategies could substantially reduce this risk.

Key Words

Wheat, climate change, adaptation, Australia, risk


Wheat is the major crop in Australia in terms of value ($4.2 billion), volume (22Mt) and area (11Mha). Yields are generally low due to low rainfall, high vapour pressure deficit and low physical and chemical soil fertility. High climate variability forces low input management to limit financial risk. Thus the Australian wheat industry is highly sensitive to climatic influences. Increases in levels of atmospheric CO2 and other greenhouse gases are considered likely to significantly change global climate, increasing temperature and changing regional rainfall patterns, with consequent impacts on the wheat industry. However, there is considerable uncertainty in scenarios of CO2 increase and related climate change - and wheat responds to both factors, with raised CO2 levels enhancing crop growth through increased photosynthetic rates and water use efficiencies, but reducing grain protein content (e.g. Howden et al. 1999). Atmospheric CO2 levels may rise from current levels (370ppm) to between 525ppm and 716ppm by the year 2070. In the same time frames, temperatures across Australia may increase by a range of 1C to almost 7C by the year 2070. Large changes in rainfall are possible with changes of up to +60% by 2070 – noting that there is marked variation between regions and seasons. Such changes in climate and CO2 levels would have potentially significant impacts on wheat yields in Australia as well as areas suitable for cropping, changes in salinity and erosion risk (e.g. Reyenga et al. 1999, van Ittersum et al. 2003).

Earlier site-based analyses of the impact of combined changes in atmospheric CO2 concentration and regional climate change (e.g. Howden et al. 1999) did not account for the large range of uncertainty in CO2 and climate changes nor scale this up to the national level. This study attempts to incorporate uncertainty in CO2 levels and consequent climate changes into one analysis using Monte Carlo (random number) sampling across the specified ranges of uncertainty at each level. The output of such analyses is a probability distribution – it indicates the likelihood of different outcomes for input into risk analyses.


The project focuses on ten sites in the major Australian wheat growing districts as a pathway to scaling up results for the whole of the industry. The sites were Geraldton, Wongan Hills, Kattanning, Minnippa, Horsham, Wagga Wagga, Dubbo, Moree, Dalby and Emerald. Response surfaces of mean wheat yields to CO2, rainfall and temperature were developed for each site using I-Wheat (Meinke et al. 1998) which is a module of the APSIM systems modelling framework (Keating et al. 2003). The approach used to model CO2 response (Reyenga et al. 1999) has been validated (Asseng et al. 2004). I-Wheat was run for a factorial combination of CO2 increase, temperature and rainfall change using modified 100-year climate records (Reyenga et al. 1999) to provide response surfaces (or a summary model) of the general form:

Yield change (% from historical mean) = aCO2 + bT + cR + ε Equation 1

Where T (C) and R (% change) are temperature and rainfall change respectively from the 100-year average, CO2 is in parts per million and ε is the residual error. Most regressions had non-linear terms and interaction terms in the above equation. A separate response surface was developed for simulations which included optimised management adaptations of change in variety and change in planting window (Howden et al. 1999). Change in grain nitrogen (%) was calculated as a function of yield change using similar regression techniques. These response surfaces were sampled using Monte Carlo methods using a proprietary package (@RISK). Uncertainty in the input variables (i.e. CO2, temperature, rainfall) for this sampling was incorporated using the following approach.

Carbon dioxide concentrations for 2070 were sampled using a uniform distribution across the range of values provided by the Intergovernmental Panel on Climate Change scenarios (525ppm to 716ppm: IPCC 2000) which relate CO2 levels to global temperature change. For the sampled CO2 level, global temperature change was sampled using a uniform distribution from within the range of values provided. Uniform distributions were assumed as there is no a priori reason for CO2 concentrations to have a greater likelihood of occurrence at any point within this range. The global temperature changes were then used to calculate monthly temperature and rainfall changes (Figure 1) at the ten sites using probability distributions derived from the results of nine Global Climate Models (Howden and Jones 2001). Means of the monthly values across the growing season (typically May to October) were calculated to sample the response surfaces from which yield changes were subsequently calculated (Equation 1). Correlation matrices were calculated between all sites for both precipitation and temperature changes as adjacent sites are likely to have similar climate changes within a sampling whilst sites distant from each other may be largely independent (Howden and Jones 2001).

Figure 1. Change in a) temperature (C) and b) rainfall (%) for Wongan Hills for the year 2070.

Changes in site yield (t/ha) were scaled to regional productivity (tonnes) using the average regional Australian Bureau of Statistics (ABS) production statistics for the past decade and the change in yield under a given global change scenario. These regional values were then aggregated to give national production. Crop value ($/tonne) was calculated as a function of grain N concentration (%N) based on several years’ data (Howden et al. 1999):

Value ($/tonne) = -66.395x3+435.6x2-851.36x+656.81 Equation 2

Where x is calculated percent nitrogen (N%) in the grain. For simplicity, the same equation was used for all States although minor differences occur. Regional crop value was then re-calculated using the yield changes and the revised crop value and then aggregated to national values. We did not incorporate changes in cropping areas for the global change scenarios as a response of change in yield potential and risk as there may be buffering responses via landuse changes (e.g. Reyenga et al. 1999). These require further analysis.


There was considerable variance in the responses among the ten sites. Wongan Hills (Western Australia) and Dalby (Queensland) were chosen to represent two extremes of response. For Wongan Hills, the climate change and CO2 scenarios resulted in marked reductions in yield in the absence of adaptation (Figure 2a) with extreme reductions of 80%. There was a small (6%) likelihood of increased yields. Median yields were 18% lower than historical values. In contrast, for Dalby in the year 2070, there was an 82% likelihood of increases in yield (Figure 2b) with these being up to 35% higher than historical values. Nevertheless, there was significant chance of lower yields with reductions up to 42% compared with historical yields. Median yields were 12% higher than the historical baseline.

Figure 2. Changes in grain yield (%) for a) Wongan Hills (WA) and b) Dalby (Qld) in the year 2070 with no adaptations practiced.

When aggregated to a national basis, with no adaptation simulated, the yield changes in addition to changes in price associated with altered grain protein content resulted in effectively no change (-0.3% or -$15M p.a.) in median crop value (Figure 3a). However, there is a marked skewness of possible results with national gross crop value possibly falling by as much as 49% but increasing by only 10%.

Figure 3. Change in national gross value of the wheat crop from historical baseline values (%) for the year 2070 as a result of increase in CO2 and change in temperature and rainfall a) without adaptation and b) with adaptations of changed planting windows and varieties.

When adaptations of changing varieties and changing planting windows (to take advantage of reduced frost risk) were simulated across the Australian wheat growing industry for the year 2070, there was a marked offsetting of the negative impacts of global change and an enhancement of the positive aspects (Figure 3b), resulting in the median value the national wheat crop increasing by 5% ($190M p.a.) with a range from –25 to +16%. The value of adaptation at a national level (i.e. the difference between Figures 3b and 3a) ranged from $100M p.a. to $550M p.a. with a median of $225M.


The approach described here allows a risk assessment of the combined effects of possible increases in atmospheric carbon dioxide concentrations, and associated temperature increases and rainfall changes on the Australian wheat industry. The approach enables a systematic treatment of the high levels of uncertainty in: 1) scenarios of global CO2 emissions, 2) the implications of these emissions for global temperature change and 3) the implications of global temperature change for regional temperature and rainfall change in Australia. It provides indications of both the median and extreme ranges of change likely for regional and national wheat yields and crop value.

Increases in atmospheric concentration of carbon dioxide and associated climate changes are likely to have significant effects on the Australian wheat industry, impacting on yields and on gross value of production. Generally, increases in CO2 concentrations had a significant positive impact on wheat yields and gross crop value but these effects were over-ridden in some scenarios of large reductions in rainfall and large increases in temperature, leading to significant risk to the industry at both regional and national levels. The likely impacts varied markedly among regions with regions in Western Australia having a high likelihood of significant reductions in yield, but those in north-eastern Australia having a high likelihood of moderate increases in yield but also a small probability of substantial yield reductions. At the national level, whilst median yields are little changed (without adaptation) there is a significant risk to the industry as the scope of increases in crop value is limited (to about 10% or $0.4B p.a.) but the scope of potential losses is large (about 50% or $2B p.a.)

Investment in developing adaptation strategies could be highly effective, offsetting the negative impacts of global change (particularly in the WA sites) and enhancing the positive aspects (particularly in eastern Australia). Use of these readily implemented adaptation strategies (changing planting windows and varieties) fundamentally alters the risk profile of climate change, changing the median result from a negative to a positive with a median gain of about $225M p.a. but with a range of $100M to over 500M p.a. Clearly, there is a sound investment in identifying suitable adaptation strategies and building the capacity in the farming community to implement them so as to deal with uncertain climate and atmospheric changes in the future.


We would like to thank Neil Huth and Brian Keating (CSIRO Sustainable Ecosystems), Sarah Lennox (Qld Dept Primary Industries) and Penny Whetton (CSIRO Atmospheric Research) for reviewing various versions of this paper. Some of the work reported here was supported by the Australian Greenhouse Office.


Asseng S, Jamieson PD, Kimball B, Pinter P, Sayre K, Bowden JW, Howden SM (2004) Simulated wheat growth affected by rising temperature, increased water deficit and elevated atmospheric CO2. Field Crops Research 85, 85–102.

Howden SM, Reyenga PJ and Meinke H (1999). Global change impacts on Australian wheat cropping. Report to the Australian Greenhouse Office. CSIRO Wildlife and Ecology, Working Paper 99/04, Canberra.

Howden SM and Jones RN (2001). Costs and benefits of CO2 increase and climate change on the Australian wheat industry. Report to the Australian Greenhouse Office. pp 25.

IPCC (2000). Special Report on Emissions Scenarios. Nakicenovic, N. and Swart, R. (Eds.) Intergovernmental Panel on Climate Change, Cambridge University Press, UK. pp 570

Keating BA, Carberry PS, Hammer GL, Probert ME, Robertson MJ, Holzworth D, Huth NI, Hargreaves JNG, Meinke H, Hochman Z, McLean G, Verburg K, Snow V, Dimes JP, Silburn M, Wang E, Brown S, Bristow KL, Asseng S, Chapman S, McCown RL, Freebairn DM and Smith CJ (2003). An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy 18, 267-288.

Meinke H, Hammer GL, van Keulen H and Rabbinge R (1998). Improving wheat simulation capabilities in Australia from a cropping systems perspective III. The integrated wheat model (I_WHEAT). European Journal of Agronomy 8, 101-116.

Reyenga PJ, Howden SM, Meinke H and McKeon GM (1999). Impacts of global change on cropping in SW Queensland. Environmental Modelling and Software 14, 297-306.

van Ittersum MK, Howden SM and Asseng S (2003) Sensitivity of productivity and deep drainage of wheat cropping systems in a Mediterranean environment to changes in CO2, temperature and precipitation. Agriculture, Ecosystems and Environment 97, 255-273.

Previous PageTop Of PageNext Page