Climate services – are they for everyone?

Information provided by climate services serves as guidance for stakeholders was the take-home message from the four-day British Council Researcher Links workshop which I attended in Cape Town on 3 March 2014.

Perspective of a climate researcher

As a climate impacts researcher, I understand why climate services play a guidance role, and not that of decision-making. Projections of future climates and/or impacts, which largely form the basis of climate services, are derived from modeling work using climate models which themselves have uncertainties [1]. This reminds me of a popular term in computer science: GIGO – garbage in, garbage out, i.e., what you produce is as good as what you put in. I am not suggesting that climate modeling activities are garbage and that we should not trust their results. Rather, I am emphasizing the fact that when you have uncertainty in the input, you will produce results which contain a level of uncertainty as well. It could be because of this uncertainty that climate scientists are unable to play a decision-making role.

Assuming the role of the stakeholder

From the perspective of a stakeholder, the notion of “guidance” can be problematic. Consider this analogy: Person A wants to build a house and so publishes an advertisement seeking a construction company. A company responds, offering its service. As part of its policy, the company is not allowed to engage in actual construction of the house but can provide a comprehensive manual titled, "Everything you need to know to build a house". The manual has been compiled by some of the best minds in the industry and contains scientifically accurate information based on state of the art technology. Printed on the first page of the manual are the following words: "USE AT OWN RISK! ALTHOUGH STEPS HAVE BEEN TAKEN TO ENSURE CONTENT INTEGRITY AND ACCURACY, THE COMPANY CANNOT GUARANTEE THAT THE EVENTUAL PRODUCT (I.E. THE HOUSE) WILL BE FREE FROM STRUCTURAL FAILURE ORIGINATING FROM THE USE OF THE MANUAL. THE COMPANY BEARS NO RESPONSIBILITY IN THE LIKELIHOOD OF SUCH AN EVENT.". How do you think person A would respond?

Drawing lessons

The moral of the story is that uncertainty is not helpful for stakeholders. Talk to any stakeholder – policymakers, farmers, etc. – and it is highly likely that the person requires a single, deterministic answer to help them decide on the next course of action. In this regard, he/she is keen to know how WILL the climate change? Our experience [2] with farmers shows that they are most concerned with the predictability of rainfall for the following week so that they can decide to plant or to delay the process. To provide that level of prediction is still beyond the reach of the modeling community. Such will happen but not overnight [3].

The way forward?

Recognizing the gap between the information which can be provided (climate services) and that which is actually needed (stakeholders), how then can climate services be made useful? On the side of climate services, the scope for narrowing that gap is largely limited apart from the quest to reduce the uncertainties as much as possible. The scope for reducing the gap on the part of the stakeholders is wide. For a start, I believe that stakeholders need to recognize that climate services are not for everyone. Climate services would be useful for those who can see the benefit of uncertainties in climate projections, as in the case of, for example, an engineer who would be able to utilize the range of rainfall projections to test system failure such as identifying thresholds over which a dam will fail. A farmer concerned with seasonal produce is least likely to find the usefulness in uncertainties in climate projections. However, such information may be useful if the farmer is interested in identifying climate-suitable growing regions in the near future.

I believe the way forward is for both parties to work together. Already such engagements are present in the form of scoping workshops which have been a key feature in many projects. These engagements are less likely to reduce the uncertainties related to modeling of future climates, but understanding the needs of the stakeholders would enable climate services to refine (or even rethink) the information they provide. After all, if you are providing a service, you might want to aim for that “100% customer satisfaction guaranteed”!

[1] There are three types of uncertainties related to projections of future climates: 1) modeling uncertainty - that our knowledge of the Earth system processes is imperfect, 2) emissions uncertainty - that we are not sure how the future is going to evolve and can only guess using a set of future pathways, and 3) the role of natural climate variability - that natural factors play a role in influencing climates, and that it is difficult to separate these factors from anthropogenic sources (e.g., greenhouse gases).

[2] Based on the work of Mariko Fujisawa on the use of climate information in agricultural decision-making as part of CCAFS project on Emerging Climate Risk, Agriculture and Food Security at ACDI.

[3] I recall from a keynote presentation at a workshop at the ECMWF in Reading, UK in 2010 that weather forecast has improved by three days over the last three decades.