Item

Abstract

Many studies now provide evidence of weather and climate effects on human migration, but only few have attempted to project the impact of future climate change, or attribute the impact of past climate change, on global migration patterns. Here we compare the existing projections, and find that for international migration from African countries, they differ by about two orders of magnitude, while for internal migration even the sign is uncertain. None of the various models used have been shown to explain historical migration changes, limiting the confidence one may have in their projections. We then discuss prospects for two types of models. Econometric models have been used to identify the marginal effects of climate on migration. Their utility for projections is limited, but they may lend themselves to specific questions of attributing current migration patterns to climate change, which has rarely been done so far. On the other hand, models of total migration can better account for the complex dynamics likely important for long-term projections, but constraining them is a challenge given the current understanding of these dynamics. Improvements may come from closer investigation of potential nonlinearities in the response to increasingly extreme climatic conditions.