Using the Social Vulnerability Index to Forecast Disaster Migration

Abstract

This report analyzes disaster-driven internal migration, its effects on residence layout by social vulnerability, and demonstrates a modeling procedure for this task. Since the processes underlying disaster-driven migration may unfold at scales below those in the data, especially in vulnerable areas where these models are especially useful, this report focuses on South Africa, which had key flooding events at urban centers over 15 years before the 2011 census. Since only low spatial resolution census data are obtainable for research, we use universal kriging (UK) with a pre-fit model from the Philippines (which used more detailed data) to estimate social vulnerability in South Africa at a higher spatial resolution. With the UK model, we estimate and reaggregate a social vulnerability index (SVI) from South African provincial to municipal boundaries, then fit a model testing the relationship between internal migration, SVI, and flooding risk within each municipality. Results show that, when controlling for flood risk, within-province migration is inversely related to SVI, while neither SVI nor flooding affect migration between provinces. This is consistent with our hypothesis that the socially vulnerable are less likely to leave flood-prone areas, and over time, a positive spatial correlation emerges between SVI and flood risk.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2019
Accession Number
AD1090876

Entities

People

  • Chandler M. Armstrong
  • Lance L. Larkin

Organizations

  • Construction Engineering Research Laboratory

Tags

Communities of Interest

  • Biomedical
  • Engineered Resilient Systems

DTIC Thesaurus Topics

  • Africa
  • Climate Change
  • Communities
  • Databases
  • Demography
  • Disasters
  • Emergencies
  • Flood Hazards
  • Floods
  • Geography
  • Groundwater
  • Human Population
  • Information Science
  • Risk
  • South Africa
  • Urban Areas
  • Vulnerability

Readers

  • Emergency Management and Homeland Security.
  • Marine Ecological Systems Migration
  • Regression Analysis.