As the planet continues to warm, many decision makers are grappling with how best to adapt to our changing climate. But for these decision makers, the most advanced global climate models are also perceived as being too ‘wrong’ to be useful for planning and adaptation. In response, an explosion of higher resolution ‘downscaled’ climate models have emerged that could have increased value. This talk presents work being done at the Southeast Climate Science Center to develop decision-relevant high resolution climate model simulations for the US Caribbean and new strategies for evaluating investments in downscaling in the context of real-world decisions.