4.1 Physical Vulnerability
Long-term trends (Table 1, Fig. 2) revealed increased precipitation in the southwest, especially in Mississippi. Counties gradually grew drier moving east, with negative tau values peaking in the Carolinas. A total of 62.6% (263 of 420 counties) reported significant precipitation (positive) trends and 6.2% (n26) with significant drought (negative) trends, while the remaining 31.2% (n131) showed no significant movement in either direction. The strongest drought trends had a tau value of -0.08 (pvalue0.001) in Alexander and Davie counties in North Carolina, while the strongest precipitations trends had taus that reached up to 0.19 in Tompkins and Schuyler counties in New York (pvalue0.001).
Trends at 30-year intervals (Table 2, Fig. 3) revealed patterns of alternating wet and dry periods. Period A (1895-1924) was a predominantly wet period, especially in the central and southern subregions, with 70.5% (n297) showing positive trends. Counties were drier in the north, though few of these (14 counties) were statistically significant. Period B (1925-1954) was notably drier, with negative tau values in 182 counties, though only 9.7% of these were statistically significant. We observed statistically significant positive trends in 22.5%, or 95 counties. Similarly, Period C (1955-1984) was an extremely wet period, with almost exclusively positive trends (e.g., less drought periods). Only 2 counties reported negative, but insignificant, tau values, while 85.2% of counties were statistically significant in the positive direction. Period D (1985-2016) follows with another dry period, though only 5 counties reported significant drought trends. The majority of the region displayed much weaker and insignificant tau values, with only 40.9% (or 172 counties) showing significantly positive trends.
4.2 Social Vulnerability
Our principal component analysis resulted in five components that explained a total of 72.5% of variance across social indicators. These components were totaled and mapped in ArcGIS Pro to visualize overall social vulnerability across the region (Fig. 4). This map did not reveal many obvious clusters of vulnerability, though Mississippi had the highest proportion of vulnerable counties than any other state (70.9%). The southern, south-central, and central subregions exceeded 30% when the three highest vulnerability classification were compiled (Extreme-High, High, Moderate-High) (Table 4). The northern and north-central subregions were deemed least vulnerable, as well as counties that surrounded metropolitan areas.
Component 1 contained 26.7% of explained variance, where most prominent variables included Disability, No Internet, Education Attainment, Poverty Status, and Unemployment (Table 3, Fig. 5). These were also variables which primarily affected rural Appalachia according to past demographic studies (e.g., Pollard and Jacobsen 2020b), demonstrating that this component most efficiently displayed vulnerability associated with rurality. There were two clusters with high vulnerability: 1) extremely high values in Kentucky, and parts of Virginia and West Virginia; and 2) the southernmost counties in Mississippi and western Alabama. The northern subregion and counties surrounding metropolitan areas showed low vulnerability. Component 2 contained 12.4% of total explained variance, and most prominent social indicators were No Health Insurance and Low English Proficiency. While this component contained few counties in the highest vulnerability ranking, there were large clusters with high and moderately-high levels in the southern and south-central subregions. The remaining subregions had rather low vulnerability. Component 3 reported 10.9% of explained variance, and contained the highest variance for Female Population, Single-Parent Households, and moderate variance for Nonwhite Minorities. There were two small clusters of elevated vulnerability in Mississippi/Alabama and in the Carolinas. Where New York and Pennsylvania had previously been presented as least vulnerable, this subregion notably shifted to having higher vulnerability. Metropolitan areas also exhibited higher rates of vulnerability. Component 4 contained 10.2% of total explained variance, and served primarily as an indicator for Elder Populations. Most of Virginia was deemed moderately vulnerable, and there was also a small cluster of vulnerable counties along the North Carolina/Georgia border. Metropolitan areas, eastern Kentucky, and a small handful of counties in New York and Pennsylvania were least vulnerable. Component 5 contained 9.373% of total explained variance, and highlighted variables for lacking access to phone services and vehicles. This component displayed clusters of elevated vulnerability in the central, north-central, and northern subregions, while the southern and south-central subregions were least vulnerable.
4.3 Bivariate Mapping
The bivariate map (Fig. 6) displayed counties that are most vulnerable to both hydroclimate extremes and socioeconomic factors. The map identified the central-southwestern portion of the study area as most vulnerable to increased precipitation, especially in Mississippi, Alabama, Tennessee, and western Kentucky. A handful of counties in northern Pennsylvania and New York also showed high increases in precipitation, but generally low social vulnerability. The central-eastern portion of the region (i.e., Ohio, West Virginia, Virginia, and North Carolina) had high socioeconomic vulnerability and low PDSI values, indicating that this area is most susceptible to increases in drought or drought-precipitation variability.