The Bigger Picture Applying Climate Forecasts Across the Value Chain

To be useful and valuable, climate forecasting must not only individually address relevant problems at farm or policy level, but be flexible enough to take a whole value chain perspective to ensure that benefits achieved at one level are not undone at the next (IRI, 2000).

An example of this approach is an assessment of the value of seasonal climate forecasting for the Australian sugar industry (Everingham et al., 2002). Australian sugarcane industries go across an integrated value chain comprising cane growing, harvesting, transport, milling, marketing and shipping. Sugarcane industries worldwide are exposed to uncertain variable climatic conditions, which have large impacts across all industry sectors. It is believed seasonal climate forecast systems offer the potential for improved risk management and decision-making across all these sectors, leading to enhanced profitability and international competitiveness.

Such a 'whole value chain approach' can:

• identify, in partnership with all relevant sectors of that industry, the key decisions influencing sustainability and profitability that are impacted by climate;

• identify the key vulnerabilities within the value chain related to CV and CC;

• develop the necessary and appropriate databases of climate and industry sector performance;

• establish the role of climate forecast systems for different geographical regions and key industry decisions;

• assess the benefits and costs of tactical decision-making based on climate forecasting across all the different components of the sugar industry value chain; and

• effectively facilitate the appropriate implementation and delivery of climate systems for enhanced risk management and decision-making (Everingham et al., 2002).

The application of the above approach is provided for those decisions relating to yield forecasting, harvest management, and the use of irrigation. There are key lessons to be learnt from this approach that can be considered generic in terms of preparedness for all agricultural industries. These include

• the absolute need for a participative R&D approach with stakeholders,

• the need to consider the whole industry value chain,

• the need for climate forecast systems with appropriate skill and underlying mechanistic foundation appropriate to different regions and different decisions (Stone et al., 2000b; Everingham et al., 2002).

In a recent example of an assessment of the economic value of seasonal climate forecasting, Antony et al. (2002) analysed the value of climate forecasting to management systems in just one sugar milling region in Australia. They determined that in one case study season (austral winter, 1998) the value of a probabilistic climate forecasting system amounted to in excess of AUD $1.9 million for one relatively small cane growing region. The assessment was made incorporating decisions at both the farm and mill scale. However, they point out that if prior 'perfect knowledge' of rainfall patterns for that season had been possible, the value to industry would have amounted to AUD $20 million, 10 times the value achieved through existing climate forecasting technology.

There is compelling evidence that during recent El Niño events media reports (often factually wrong or distorted) influenced factors such as bank lending policy and agribusiness advice in Australia. Better contextualised information appears to be required. Brennan et al. (2000) discussed how agribusiness might be able to use information about CV. The same issue exists to an even greater degree for CC issues. In 2001, a major Australian insurance company withdrew its insurance coverage for cyclone damage as a result of indications of more intense cyclones in a warmer world. Brennan et al. (2000) concluded that there is considerable potential to use simulation approaches to improve bank lending policies, crop insurance policies, product inventories and marketing advice. Specifically, they found that model predictions can reduce claimant disputes and cut legal costs. They also provide the option for individually tailored financial packages. They stated that... while discussions with agribusiness indicate keen interest in such tools and information, they have yet to have impact on policy and operations. While clear outcomes have been seen in some areas of engagement (e.g., insurance/loss assessment), other collaborative efforts have only progressed some way to exploring the role for seasonal climate forecasts and simulation models in their business operations (e.g., banking, portfolio analysis) (sic).

Chapman et al. (2000a) provided a further example of the powerful combination of simulation modelling and climate forecasting: Timing and severity of water limitation affect crop growth and yield differently. In order to screen germplasm for broad adaptation to drought, plant breeders conduct large-scale multi-environment trails for several seasons, where seasonal conditions vary spatially and temporally. These trials are conventionally analysed by assuming that each location is representative for an environment-type. However, it is unlikely that the few seasons encountered at these locations during a trial represent the true frequency of all possible season types at this location (particularly when considering some of the low-frequency variability discussed in Table I). Using a simulation model for sorghum, Chapman et al. (2000a) characterised the stress environments for each location and season (e.g., early, mid or late stress) and then analysed the results by environment type rather than location. This resulted in a considerable amount of additional information that was previously attributed to 'environmental noise'. Hence, this environment-type classification can lead to more rapid improvement in the selection process, thus improving the efficiency of a breeding program leading to a more rapid development of better-adapted cultivars. Once such improved cultivars are available, seasonal climate forecasting can help farmers to select the appropriate cultivars for the most likely environment type at a location.

Was this article helpful?

0 0

Post a comment