Bridging the gap between seasonal climate forecasts and decision makers in agriculture

The first objective of the project was to work with PAGASA (Philippine Meteorological Bureau) to deliver and communicate the basis of seasonal climate forecasts. This is foundational to any future work on valuing seasonal climate forecasts. The communication between climate science and agricultural decision makers has been possible through the close interaction between PAGASA, PIDS and Leyte State University (LSU). Workshops have been held at PIDS office in Manila (April 21), Leyte State University, Baybay (June 30), and Malaybalay City, Bukidon (December).

One of the tools that PAGASA has used to explore the relationship between tropical sea surface temperatures and Philippine rainfall has been the International Version of RAINMAN, a product of a previous ACIAR project. This has been valuable to show where and when the forecasts have statistical skill. The interaction between PAGASA and PIDS has produced a number of publications suitable for non- technical audiences which explain the scientific basis behind climate outlooks- for example how the El Nino Southern Oscillation influences the Philippines and how this makes forecasting the season more reliable in certain parts of the country at certain times of the year.

How to match the information climate science can offer to decision making has been much discussed within regular project team meetings between the Philippine partners and during trips from Australian partners to the Philippines in March (project initiation), July (project launch) and October. Further opportunities were available when Philippine partners from PIDS and LSU visited Australia to take part in the economist planning meeting of NSW DPI economists held in Trangie, NSW in November 2005..

An encouraging aspect of the project has been the high level Philippine government support of the project through membership of the steering committee. This support was evident at the project launch by the Australian ambassador Anthony Hely on July 27th . Over 70 representatives attended the launch from government, farmer organisations, universities and industry. The main organisations represented (in addition to PAGASA, PIDS and LSU) were the National Irrigation Administration (NIA), The National Food Authority (NFA), the National Economic and Development Authority (NEDA) the Southeast Asia Research Council for Agriculture (SEARCA), the International Rice Research Institute (IRRI) and the University of the Philippines Los Banos. The launch received a high level of media coverage and provided an opportunity to draw attention to a website for the project http://dirp3.pids.gov.ph/ACIAR

Fundamental to the project in both Australia and the Philippines is the notion that economic frameworks for valuing uncertain seasonal forecasts will assist the communication process.
To value SCF, a particular decision must be analysed and the benefits of a forecast calculated over time. Four case studies were started in the first year of the project as a means of ensuring that the theory and modelling is grounded in real decisions. From these case studies we hope to determine a potential value of SCF (desktop analysis) and the actual value when all the complications and restrictions are taken into account.

In the Philippines, the first case study was set up to examine the importation of rice by the National Food Authority. In discussion with decision makers this was expanded to include the purchase, storage and distribution of rice. The second case study is at a farm level and examines decisions dryland corn farmers may make about cropping intensity, variety, fertiliser and sowing density.

The documentation from PAGASA of the status of climate data in the Philippines is a valuable resource for both this project and other work on managing climate risk. The information on the skill of the forecasts using RAINMAN provides a sound basis for valuing SCF in the project. The workshops in the provinces (Leyte and Bukidon) and the accompanying material explaining ENSO are important background for later work on valuing and using SCF within the project.

In Australia one case study has examined crop choice on the Liverpool Plains in northern NSW where farmers are presented with a climatically risky decision whether to sow a wheat crop in winter or continue the fallow to store more moisture for a subsequent sorghum crop. The second case study examines decisions about stocking rate on large wool growing properties in the rangelands of western NSW.

Although we are using case studies, we also aim to provide generic guidelines on valuing seasonal climate forecasts. To that end we have produced a number of Excel based spreadsheets with accompanying documentation. Two versions have been produced in Australia and one in the Philippines. These are not aimed at decision support systems for growers but rather frameworks for thinking through the value of SCF. An example of such a simple decision tree for nitrogen decisions on wheat was presented to the WMO Expert team on Weather Climate and Farming at Geneva, November 2004

The ongoing drought in Australia and typhoons in the Philippines combined with a high level of media coverage regarding climate change has ensured interest in the results of this project on bridging the gap between climate science and decision makers.

The project has provided resources to improve the capacity of PAGASA to deliver Seasonal Climate Forecasts for the regions in the case studies (objective 1). More importantly it has enabled PIDS and LSU to work with PAGASA to better link the climate science with the user community.

Some of the climatically risky decisions we are examining (objective 3) are policy decisions of how much rice to import into the Philippines; what variety of corn and how much fertiliser to use in the central Philippines and the Isabella north of Manila; variety choice and sowing time to maximise yield but avoid frost in southern Australia; the optimum amount of cropping and pasture on a farm in central west NSW. By engaging with decision makers in these diverse contexts we not only are having an immediate impact on the way that SCF are used, but we are also learning about the challenges of using climate information and ways of better placing a value on seasonal climate forecasts.

Surveys of farmers in the Philippines (objective 4) show a high degree of concern about climate risk and a high level of awareness of El Nino but only moderate use of the information in decision making. A paper that reviewed much of the survey work in Australia also highlighted difficulties in applying probabilistic SCFs to decisions. As part of the project (objective 2) we have further refined an Excel based game that allows participants to consider how they would use forecasts that are much better than guessing, but a long way short of perfect information in decision making.

The project team were successful in having a symposium accepted for the 26th Conference of the International Association of Agricultural Economists (IAAE) which was held on the Gold Coast, Queensland, in August 2006. The title of the symposium was 'Valuing Seasonal Climate Forecasts'. This conference along with a workshop on the Science of SCF organised by the Australian Academy of Science on 2nd and 3rd of August and an International Conference on climate impacts and adaptation in Cairns in July provided excellent opportunities for the Philippine and Australian team to learn and discuss the theory and practice of valuing SCFs.

This project was developed on the broad premises that a) Philippines and Australia were both significantly affected by El Nino Southern Oscillation (ENSO), b) there is a lot known in the climate science community about the onset and impacts of ENSO events and c) there was a gap between the knowledge in the scientific community and the application of this knowledge in decision making.
The El Nino event of 2006/07 had an impact in both countries. Although this event developed quickly it was generally well communicated, nevertheless the challenge of how to use this information in decision making still remains. It is nave to assume that a modest project such as this can solve the challenge of developing actionable knowledge from climate science, however through a range of case studies and thoughtful analysis, we can contribute to the understanding of barriers and suggestions for success.
We have looked at the use and potential value of seasonal climate forecasts in the following case studies: Corn farm level study in Leyte; Corn farm level study in Isabela; Corn farm level study in Panglao, Bohol; Rice farm and household level study in Nueva Ecija; Rice policy case study in Philippines; Use of seasonal climate forecasts (SCFs) in mixed farm context in Central West NSW; SCFs and opportunity cropping in the Liverpool Plains, NSW; Economic analysis of farming systems in Western NSW; Nitrogen on wheat; Low rainfall farming in South Australia; and Use of SCFs to manage frost risk in southern Australia. In August 2008 a number of these case studies will be formerly presented as part of the Asian Society of Agricultural Economists (ASAE) conference in Manila, Philippines.
The rice policy study, when completed, can provide a very useful input to policymakers since it takes into account not just domestic rice production but also world prices. The rice simulation model highlights the importance of SCFs, domestic production, and external factors as bases for decision to import or not, and the corresponding policy costs of such decisions.
From these case studies we have started to draw together information on the potential value of seasonal climate forecasts. In some cases there is an opportunity to develop operational applications of forecasts, however the case studies were always intended as a means to the end of generalising on the application of seasonal climate forecasts in a range of decision contexts.
Part of this generalising came from the paper comparing seasonal climate forecasts to other innovations that Australian farmers were encouraged to adopt. The use of adoption theory helped us identify characteristics of seasonal climate forecasts as an innovation. We identified that it is difficult to communicate the comparative advantage of probabilistic SCFs and there is a significant cost in terms of time rather than money in developing a working knowledge of SCFs. However, compared to say an innovation such as a new wheat variety, SCFs have economies of scale (can be used across one paddock or a number of farms) and economies of scope (can be used for different enterprises such as cropping or livestock and for different levels of decisions such as fertiliser rates or buying extra land).
In the final period of the project we are looking at ways of summarising and producing guidance material to bridge the gap between the current use and potential value of SCFs in farming systems. It is encouraging that a number of communication tools and frameworks developed in this project have already been used by other groups in the Philippines and Australia.