BAMOS Vol 30 No. 4 2017 | Page 26

26 BAMOS Dec 2017 Article Linking forecasts and end users: perspectives from a Pacific Aid Program Roan D. Plotz and Lynda E. Chambers Climate and Oceans Support Program in the Pacific (COSPPac), Community Forecasts, National Forecast Services Group, Bureau of Meteorology, Melbourne, Victoria, Australia. Email: [email protected] Introduction In recent decades there have been significant advances in the skill of climate predictions (Bauer et al., 2015). While numerical weather and climate forecasts are pervasive throughout modern society, many of the world’s regional and remote communities do not utilise these forecasts to the degree expected (Gilles and Valdivia, 2009; Pennesi, 2007, 2011). There are many reasons for this, including: forecast outputs covering too large an area to be relevant to local communities; an incomplete understanding or trust of forecasts; and limited access because they are not disseminated via appropriate media or time scales (Gilles and Valdivia, 2009; Plotz et al., 2017). As a consequence, many regional communities remain detached from the recent technological advances in forecast accuracy, potentially reducing their adaptive capacity to an increasingly variable climate. The impacts of increased climate variability and extremes on regional communities are already severe across the Pacific Islands due to their location in vulnerable environments (e.g., small low lying islands) and reliance upon resource- based livelihoods (Nakashima et al., 2012). If the benefits of contemporary forecasts are to reach more vulnerable end users, forecast accuracy needs to be balanced with timeliness and relevance and expressed using language understandable to the end users (Pennesi, 2007, 2011; Plotz et al., 2017). For example, an urban professional able to access climate information via multiple media options will likely have very different forecast needs to that of a rural subsistence farmer on a remote Pacific Island. For these farmers, a less accurate forecast, arriving on an accessible medium such as radio or via a community meeting with sufficient lead time would be more valuable than a highly accurate but overly technical forecast delivered online after irrevocable decisions have already been made. These are some of the reasons why many local and indigenous communities continue to rely solely on traditional forecasting methods, such as observations of biological and physical indicators, to interpret meteorological phenomena even though seasonal climate forecasts (SCFs) are provided by the National Meteorological Services (NMSs) in most countries (Gilles and Valdivia, 2009; Plotz et al., 2017). Combining seasonal forecasts with traditional weather and climate knowledge Evidence suggests that the uptake of contemporary SCFs by many local communities can be significantly improved when combined with traditional knowledge (TK) forecasts (Pennesi 2007, 2011; Plotz et al., 2017). Benefits of this approach include not only the potential to increase understanding through use of familiar terminology, but also improved spatial and temporal resolution of SCFs because TK forecasts are typically more localized and have shorter outlook periods than the SCFs currently produced by NMSs (Plotz et al., 2017). However, as TK forecasting methods remain largely undocumented (i.e., oral; Figure 1), any attempt to formally combine traditional and contemporary forecast systems requires an improved understanding of the communities where TK forecast usage still remains strong, including which TK indicators are being used and how widely they apply. Although many methods for combining TK and contemporary seasonal forecasts have been proposed, they can be broadly categorised into two main approaches: Consensus and Science Integration (see Plotz et al., 2017). The Consensus approach brings together groups of experts, usually representatives from the indigenous group who holds the forecasting knowledge and representatives from the NMS. These two groups discuss their respective forecasts and form an agreed or consensus forecast. In the Science Integration approach the TK forecast is formally (mathematically) combined with a dynamical or statistical climate model. Formal combination of traditional and contemporary SCFs has historically been practised in Africa, with regular meetings occurring between NMSs and indigenous TK experts (e.g., Kenya, Tanzania, and Uganda; see Plotz et al. 2017). More recently, Pacific Island NMSs have recognised the value that TK could bring to the communication and uptake of their climate products after community feedback indicated a strong regional preference for traditional forecasting m ethods (Chambers et al., 2017; Chambers and Plotz, 2017).