Social Vulnerability in Watershed Planning - Milwaukee River Basin 1980 to 2010

The dataset supports the social vulnerability index (SoVI) in the Milwaukee River Basin (Wisconsin, USA). We downloaded data for the watershed from 1980, 1990, 2000, and 2010. These data came from the Neighborhood Change Database maintained by Geolytics, which apportions historical census data into 2010 boundaries. Seven of the 32 variables used by Cutter et al. [1] were not available. The loss did not alter vulnerability designations (authors, unpublished data). We performed Z-score transformations on the remaining 25 variables and completed principle components analysis (PCA) with varimax rotation. Such standardization and rescaling limits the influence of indicators with different units or ranges. Before summing PCA coefficients, each is assigned a direction so that higher scores indicate higher vulnerability.

本数据集可为美国威斯康星州密尔沃基河流域的社会脆弱性指数（Social Vulnerability Index, SoVI）构建提供支撑。我们从Geolytics公司维护的社区变迁数据库（Neighborhood Change Database）中下载了该流域1980年、1990年、2000年及2010年的数据集，该数据库可将历史人口普查数据匹配至2010年的普查边界体系中。Cutter等人[1]所采用的32项变量中，有7项无法获取，但该变量缺失并未改变区域脆弱性等级的判定结果（作者未发表数据）。我们对剩余的25项变量进行了Z分数标准化变换，并采用方差极大旋转法完成主成分分析（PCA）。此类标准化与重缩放操作可限制不同量纲或取值范围的指标对分析结果的影响。在对PCA系数进行求和前，我们为每一项系数赋予了特定方向，以使最终得分越高，对应区域的社会脆弱性水平越高。