SCIENCE OF SEASONAL CLIMATE PREDICTION

Applications of seasonal predictions in Australia
by Holger Meinke, Queensland Department of Primary Industries and Fisheries/APRSU

Dr Holger Meinke is a Principal Scientist with DPI&F who manages an interdisciplinary research team. He has a Masters Degree in International Agricultural Development (TU Berlin, Germany) and PhD in Agriculture and Environmental Sciences (Wageningen University, The Netherlands). His work covers two major disciplines: agricultural systems sciences and climate sciences. He is a founding member of the Agricultural Production Systems Research Unit (APSRU), a joint venture between CSIRO, Qld Govt and UQ. He and his team focus on the development and application of agricultural systems models to deliver climate risk technologies for rural industries. They conduct climate variability/climate change scenarios analyses for risk assessments at field and farm levels but also as input into policy decisions. Their work includes strong national and international collaboration. Dr Meinke is a member of CLIVAR’s Asian-Australian Monsoon Panel and a part of a WMO Expert Teams on forecast verification. He has 20 years
of international research experience in agriculture, natural resource management, systems analysis
and climatology.

‘Seasonal prediction (or forecast)’ = ex ante assessment of
likely climatic conditions for the season ahead
‘Application’ = Action taken in response to a seasonal prediction

Unless a seasonal prediction has ‘relevance’, that is, it addresses issues in ways that influence decisionmaking, the prediction will remain without impact and will thus be without value. Perception of forecast users, rather than just scientific precision, strongly influences the relevance of a forecast. Although farmers are one obvious client group, the value of a forecast might not depend on their response alone, but rather on a series of responses by interrelated decision makers at different scales. This responsiveness across multiple tiers of governance depends very much on the socio-economic and political context, local infrastructure including level of capacity, nature of scientific institutions, past experiences with climate forecasts and the agricultural system in question. To identify clearly clients and their decision points, it is helpful to classify them according to geographic and governance scale and the types of decisions that they make. Using economic decision analysis and adaptive governance as conceptual frameworks assists in identifying decision makers, the questions they need to answer, and therefore the types of climate information most useful to them. The needs of decision makers then become design criteria for applied climate science, assisting in the selection of the most appropriate and efficient data and tools to use. This implies that we need at least two pathways for effective applications. Firstly, a technically and scientifically sound prediction scheme that narrows the possible outcomes of the decision variable of interest to the decision maker. Such a scheme must therefore allow probabilistic assessments of alternative management options on secondary or even tertiary decision variables such as production, incomes and wellbeing. Secondly, we need institutional arrangements, structures and relationships that allow the use of a scientifically robust approach in a decision-making environment that goes beyond science. These institutional and social pathways are about engagement between scientists from diverse disciplines and decision makers across multiple tiers of governance. These pathways need to transcend the limitations posed by traditional institutional arrangements, structures and the vested interests of science institutions and decision makers. Such institutions need to embrace and foster pluralistic approaches that create an environment that values scientific knowledge (quantitative approaches) as well as qualitative methods. The combination of both is likely to yield more knowledge that either alone. In this presentation we will explore these issues. We will also provide some specific Australian examples.