The paper summarizes an end-to-end activity connecting the global climate modeling enterprise with users of climate information in Alaska. The effort included retrieval of the requisite observational datasets and model output, a model evaluation and selection procedure, the actual downscaling by the delta method with its inherent bias-adjustment, and the provision of products to a range of users through visualization software that empowers users to explore the downscaled output and its sensitivities. An additional software tool enables users to examine skill metrics and relative rankings of 21 global models for Alaska and six other domains in the Northern Hemisphere. The downscaled temperatures and precipitation are made available as calendar-month decadal means under three different greenhouse forcing scenarios through 2100 for more than 4000 communities in Alaska and western Canada. The visualization package displays the uncertainties inherent in the multi-model ensemble projections. These uncertainties are often larger than the projected changes.

本文总结了一项端到端的活动，该活动将全球气候建模企业与阿拉斯加的气候信息用户联系起来。这项工作包括检索必要的观测数据集和模型输出，模型评估和选择程序，通过带有固有偏差调整的delta方法进行的实际缩减以及通过可视化软件向用户提供产品，从而使用户能够使用探索缩减规模的产出及其敏感性。附加的软件工具使用户可以检查技能指标和针对阿拉斯加和北半球其他六个领域的21个全球模型的相对排名。在阿拉斯加和加拿大西部的4000多个社区中，到3100年，在三个不同的温室强迫情景下，降级的温度和降水可以作为历月的十年平均值来使用。可视化软件包显示了多模型集合投影中固有的不确定性。这些不确定性通常大于预计的变化。