Harmonizing the supply of climate information with the type of information needed by next-users is crucial for effective weather and climate services (CS). Understanding of information demand could help reshape supply-side based CS that have dominated the field over the last few decades. Most CS have been developed using a ‘loading dock’ model, whereby products are designed by information suppliers with little input from or consultation with users of climate services. Notably, a focus on climate modelling and prediction has largely resulted in a lack of consideration of the demand-side when producing climate services. Here, we contribute to understanding of CS demand by presenting a global meta-analysis – a ‘decision matrix’ - of farmers’ climate-influenced decisions. We identify 41 studies that encompass 186 decisions, three forecast timescales (weather, dekadal, seasonal), and five forecast variables (precipitation, temperature, wind, soil moisture and soil temperature). Several insights were offered by this literature review into the value of climate services and the way forward in considering users’ needs. We find that the seasonal precipitation is the most frequently used forecast variable for decision-making, particularly of crop sowing date. Forecasts such as temperature, soil moisture and soil temperature appeared to be less used by farmers, according to the decision matrix. It is apparent that more investigation is necessary into how farmers use climate information in their decision-making to better establish the value of CS. We suggest that different sectors should make their respective decision matrices to explore decision spaces and engage with users of climate information in various sectors.

Natural resource managers model plant and animal climate vulnerability to efficiently plan and prioritize conservation actions. We assessed a commonly used online decision support tool, NatureServe’s Climate Change Vulnerability Index (CCVI) and its sensitivity to different generations of climate models, socioeconomic climate scenarios, and different types and scales of spatial data. Our goal was to help mangers better interpret CCVI species vulnerability rankings and provide a workflow for generating the required climate metrics when those data are not already calculated. We found vulnerability scores were much larger when using an older generation as compared to a newer generation of climate projections, especially for water associated species. Socioeconomic climate scenarios were very important and resulted in differences of 130% between optimistic and pessimistic scenarios. Hence, we caution when comparing across studies to consider the generation of climate models and scenarios because the scores may not be directly comparable. The online tool was robust to the type and scale of spatial data. However, this data is only applied in calculating exposure and is not included in species specific sensitivity via adaptive capacity. Consequently, we suggest that managers use range size as an index of adaptive capacity when interpreting similar scores. Overall, we find the species climate vulnerability model supported by NatureServe a valuable and robust assessment that is easy to use and provides an objective method for prioritizing conservation action and justifying these actions to policy makers and the general public.