Impacts of climate change on internal migration in South Africa through an inferential network analysis approach
Abstract
Climate variability and climate change influence human migration both directly and indirectly through a variety of channels that are controlled by socioeconomic, cultural, and psychological processes. Modeling of human migration flows is thus complex and challenging, leading to development of a range of approaches including econometric frameworks. Such models often assume that a migration flow directing from one origin to one destination is independent of any migration flow from a different origin and/or to a different destination, ignoring network autocorrelation. Here we apply a recently developed inferential network analysis (INA) approach to model internal migration flows in South Africa. This country has high levels of internal migration and is projected to experience dramatic climate change in the 21th century. Preliminary studies indicate that internal migration there is expected to be sensitive to climate extremes such as heat waves and droughts. We calculate monthly SPI (standard precipitation index) and soil moisture percentile from 1950 to 2016 to quantify meteorological and agricultural drought separately using an updated land surface model (VIC) simulation driven by the latest Princeton Global Forcing (PGF) datasets. We examine the interaction between migration and water scarcity, through water withdrawal and consumption for different sectors (e.g., domestic, industrial, irrigation, and livestock). Using a network modeling approach, we not only naturally account for dependencies in the migration flows, but also display the influence of network structures on the migration pattern. Our results show "push-pull-mooring" effects of the climate variability on migration flows and found rarely addressed destination climate dependence of the migration flows. We also compare the difference in the climate impact mechanisms behind urban oriented and non-urban oriented migration flows, demonstrating that urban oriented migration flows are more sensitive to climate variability.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMGC31F1313X
- Keywords:
-
- 1630 Impacts of global change;
- GLOBAL CHANGEDE: 1986 Statistical methods: Inferential;
- INFORMATICSDE: 6304 Benefit-cost analysis;
- POLICY SCIENCESDE: 6615 Legislation and regulations;
- PUBLIC ISSUES