Historical and future wind events for Alaskan communities, Alaska, 1980-2099

Warning to data users - In versions before doi:10.18739/A2QB9V66G, errors have been discovered in the following variables of this dataset: mean_direction. If you have previously downloaded this dataset which included those variables please download the new version of the dataset. Hourly wind data from 67 Alaska communities with observing stations, one atmospheric reanalysis, and two global climate models were processed to identify wind events exceeding location-specific thresholds. The ERA (ECMWF (European Centre for Medium-Range Weather Forecasts) Re-Analysis)-Interim reanalysis and output from the GFDL-CM3 (Geophysical Fluid Dynamics Laboratory) and NCAR-CCSMM4 (National Center for Atmospheric Research) global climate models were downscaled to Alaska and the surrounding regions using the Weather Research and Forecasting model. Hourly wind speed measurements collected during 1980-2015 at 67 Alaskan communities via ASOS/AWOS (Automated Surface Observing System/Automated Weather Observing System) stations were used to bias-correct the WRF-downscaled (Weather Research and Forecasting) ERA-Interim hourly wind speed output. These were subsequently used to bias-correct the historical (1979-2005) and future (2006-2100) WRF-downscaled CM3/CCSM4 wind speeds. These wind speeds were then processed to identify wind events meeting a set of speed and duration thresholds. A wind event was defined as the wind speed being greater than the wind speed threshold, for some duration in hours. Five wind speed thresholds were used for each community. The thresholds were community-specific, and were calculated as the 50th, 75th, 85th, 95th, and 99th percentiles of the ERA-Interim wind speed distributions for the particular community. The five duration thresholds used were 1, 6, 12, 24, and 48 hours. This dataset contains all wind events identified for each of the 67 communities, with each row of the data file consisting of a single wind event. This data drives some of the visualizations in the webtool found here.

致数据使用者的警告：在doi:10.18739/A2QB9V66G之前的版本中，本数据集的mean_direction变量存在已发现的错误。若您此前下载的数据集包含此类变量，请下载数据集的最新版本。

本数据集整合了67个设有观测站点的阿拉斯加社区的逐小时风速数据，结合一套大气再分析数据与两套全球气候模式数据，经处理后用于识别超出站点专属阈值的大风事件。

本次研究使用的ERA（欧洲中期天气预报中心（European Centre for Medium-Range Weather Forecasts）再分析）-Interim再分析数据，以及GFDL-CM3（地球物理流体动力学实验室（Geophysical Fluid Dynamics Laboratory））与NCAR-CCSMM4（美国国家大气研究中心（National Center for Atmospheric Research））全球气候模式的输出结果，均通过WRF（Weather Research and Forecasting，天气研究与预报）模型降尺度至阿拉斯加及周边区域。

研究人员采用1980-2015年间，通过ASOS/AWOS（自动地面观测系统/自动气象观测系统）站点在67个阿拉斯加社区采集的逐小时风速实测数据，对经WRF降尺度后的ERA-Interim逐小时风速结果开展偏差校正。基于校正后的ERA-Interim数据，研究团队进一步对经WRF降尺度的CM3与CCSM4历史时段（1979-2005年）及未来时段（2006-2100年）的风速结果进行偏差校正。

随后对上述校正后的风速数据进行处理，以识别同时满足风速与持续时间阈值的大风事件。大风事件的定义为：风速在至少指定时长内超过对应社区的专属风速阈值。本研究为每个社区设置了5类风速阈值与5类持续时间阈值：其中风速阈值为该社区ERA-Interim风速分布的第50、75、85、95、99百分位数，持续时间阈值分别为1、6、12、24、48小时。

本数据集包含为67个社区识别出的全部大风事件，数据文件的每一行对应单一场大风事件。该数据集为对应网页工具中的部分可视化内容提供数据支撑。