6

Regional Versus Global Perspectives for Modeling Climate Change–Related Migration Impacts

This session explored how evaluating climate change and migration at specific geographic locations affects the understanding and perception of migration flows and drivers. Speakers discussed how they incorporate geography at varying levels into their research and the value of incorporating various spatial analyses into study designs and research.

A GLOBAL LOOK AT MIGRATION

Paolo D’Odorico (University of California, Berkeley) highlighted insights from global studies on migration. Although migration is likely underreported in official records, he said, international migrants account for 2–3 percent of the global population according to census data. An analysis of the the global human migration network showed a pattern of increasing globalization in migration, reflected in a growing total number of migrants and an increasing level of interconnection among countries (Davis et al. 2013).

Overall, D’Odorico said that studies have revealed that the nexus between migration and environmental conditions can be extremely complex, with multiple causal factors and not driven solely by environmental conditions. There are also time lapses between environmental migration and its causes and drivers. He noted that where people go also depends partly on where they come from; the community structure of the global network of migrations shows modularity in that there is a community of countries within which people tend to move based on identities of language and religion.

A recent meta-analysis examined factors such as migration timing, volume, and geographic patterns in sub-Saharan Africa based on research

in the environmental sciences and social sciences (Wolde et al. 2023). The study identified two major environmental drivers. One was excessive water events such as heavy rainfall, cyclones, and riverine flooding. At the other extreme was drought or water scarcity, causing poverty and famine. Each of these drivers was associated with the migration of about 6 million people.

Migration decisions are driven by complex arrays of pushes and pulls, direct and indirect factors, and environmental and non-environmental reasons. D’Odorico highlighted how radiation models can be used to examine the relationship between migration and environmental drivers by accounting for different pathways through which migrants are induced to move (Davis et al. 2018; Simini et al. 2012). This requires an analysis of the environmental drivers, how these drivers affect their livelihoods, and exposures to environmental risk and crop failure, he said.

HABITABILITY, SOCIAL DRIVERS, AND CLIMATE MOBILITY MODELING

Alexander de Sherbinin (Columbia University) focused on the role of habitability in climate mobility modeling, the importance of social drivers, and methods for modeling migration. In 2021, de Sherbinin and colleagues defined habitability as the conditions in a particular setting that support healthy life, productive livelihoods, and sustainable, intergenerational development (Horton et al. 2021), and de Sherbinin noted that “at a reasonable cost” it could be added to this list. He said that climate change can undermine multiple dimensions of habitability, including basic survivability, livelihood, security, and society’s capacity to manage environmental risk. “This has to do with essentially climate change undermining the very foundations of resilience,” he said. “If your governance structures are being affected by climate impacts, if your GDP [gross domestic product] declines by 50 percent because of a storm, all of these things are going to undermine the capacity of society to respond.”

Although top-down, process-based models driven by global climate models can help identify hotspots and warn of potential dire consequences, de Sherbinin cautioned that they are often overly deterministic. As other speakers noted, he said that distress migration will likely rise in response to climate stressors and some of this will be involuntary mobility. From a policy perspective, the Global Compact for Migration calls on governments to strengthen joint analysis and sharing of information to better map, understand, predict, and address migration (Global Compact for Migration n.d.). Toward this goal, de Sherbinin and colleagues proposed combining top-down assessments with bottom-up co-production of knowledge with local communities and stakeholders as well as bottom-up modeling

approaches such as agent-based modeling, which are more suited to local circumstances (Horton et al. 2021).

There have been various hotspot mapping efforts based on climate factors such as sea-level rise, global temperature, heat stress, droughts, and floods (Giorgi 2006; Horton et al. 2018). It is also possible to produce maps using an ecological perspective by showing where humans tend to preferentially inhabit certain types of climatological niches and looking at climate change shifts that would change the optimal conditions for human habitability (Samson et al. 2011; Xu et al. 2020). Overlaying this with demographic pressure can reveal joint risks from future climate impacts and population issues.

Highlighting presentations from the March 2023 Population-Environment Research Network cyberseminar, “The Habitability Concept in the Field of Population-Environment Studies: Relevance and Research Implications,”¹ de Sherbinin underscored the multifaceted nature of habitability. Overall, researchers have described three dimensions of habitability: the collective ability to respond to risk (governance), livelihood resilience (e.g., food security), and physical and psychological safety. To understand habitability, he emphasized that it is important to consider both physical dimensions and socially constructed dimensions, which are often overlooked in research. For example, social dimensions can relate to capabilities, such as what people can do with the environmental resources they have or what power they have over their local environment. Emotive factors, such as a person’s sense of attachment to the area they consider home is another social dimension. If populations are declining in an area, it may signal that habitability has declined, he added.

de Sherbinin stressed the importance of looking at the underlying social factors driving migration. As an example, he pointed to an analysis of migration in Senegal, which described how agricultural market structures have put smallholder farmers in a precarious financial position, leaving many indebted, with a limited economic future (Ribot et al. 2020). Although climate-related events may create the final push that drives some of these farmers to migrate, there are also other factors at play. To understand such issues, he added that it is key for academics in the West to consider the experiences of people categorized as migrants or displaced. He highlighted two studies examining different strategies for climate-migration modeling that does or does not incorporate migrants’ experiences alongside environmental factors (Khosravi 2024; Tschakert and Neef 2022). For example, an examination of three surveys in West Africa found that the

___________________

¹ See https://populationenvironmentresearch.org/cyberseminars (accessed June 24, 2024); see also https://iussp.org/en/pern-cyberseminar-habitability-concept-field-population-environment-studies (accessed May 28, 2024).

top reason migrants cited for leaving their homes was financially related, followed by violence, family and personal reasons, and marriage and education. Environmental factors were less important. Even though environmental issues likely influenced some of the other factors cited, these findings demonstrate how the actual perceptions of migrants may differ from what scientists consider to be important drivers, de Sherbinin said.

Climate mobility models have varying capabilities, de Sherbinin said, noting that some are better at causal inference and other models are better in prediction. Modeling can be used at different temporal and spatial scales, ranging from local and very immediate scales of hours to days for evacuation modeling to longer-term migration trends. To use models effectively, de Sherbinin emphasized the need to ground modeling in migration theory. He noted that an analysis of about 75 climate migration papers found that the neoclassical theory (the basic argument that people move for rational reasons based on whether they can earn a higher wage somewhere else or due to wage differentials) dominated among the theories represented (de Sherbinin et al. 2022).

de Sherbinin summarized various modeling approaches that have been used to study climate migration and its impacts. For example, the exposure model focuses on the metrics or threshold of environmental factors such as sea-level rise or drought to determine when an area becomes uninhabitable. Agent-based models offer a bottom-up approach, exploring internal dynamics and policy implications, but require extensive data, de Sherbinin noted. Integrated assessment models, which are more complex and less common, analyze migration between countries. Finally, de Sherbinin noted that intercomparisons of model outputs to date show widely varying projections of migration patterns in response to climate change, emphasizing the nonlinear nature of these dynamics. He is working with colleagues at Columbia, Princeton, and Cornell Universities to develop a formal inter-comparison process that would put climate mobility modeling on a more solid footing. Expressing agreement with other speakers at the workshop, he also drew attention to the need for more transdisciplinary collaborations. There are funding initiatives such as the NSF Global Convergence Research² program that support these collaborations. de Sherbinin also acknowledged that working across disciplines may be challenging and that there are barriers such as the lack of shared vocabulary and training and limited resources that need to be overcome.

___________________

² See https://new.nsf.gov/funding/opportunities/growing-convergence-research-gcr (accessed July 1, 2024).

DRAWING ON TRADITIONAL ECOLOGICAL KNOWLEDGE

Meda DeWitt (The Wilderness Society) discussed how sharing traditional ecological knowledge from Indigenous cultures can help guide communities in their response to climate change by sharing a story that has been recounted through thousands of years. During a time when the climate was changing rapidly, the Tlingit and Dene peoples who lived inland faced changes that disrupted traditional food sources and led to starvation, illness, and increased mortality. Recognizing the threat to their community’s existence, they came together and decided to migrate in order to survive.

DeWitt relayed the story of their journey that included traveling through a river that went under a glacier. Two elder women decided to travel in a canoe under the glacier while two young men hiked over the glacier to map out the route and see if it was safe for their community. The community sang a grieving song to recognize the women who might perish and also the fact that their community was entirely dependent on their success. The women ultimately made it safely out to the coast and the rest of the community were able to follow the same route they had taken. They traveled to the coast, an area with an abundance of food, but grieved for their homeland. After a time, they adapted to living on the coast. DeWitt added that the community also has a song that helps repair grief and encourages people to enjoy living again.

DeWitt said that Indigenous people from all over Alaska have stories and traditional ecological knowledge relevant to climate change, often transmitted through stories like the one she shared. They know that Alaska was warm once and will be warm again. They have experienced the sea level rising and falling. While migration and climate change are not new experiences for them historically, she said, today’s experiences are new due to geopolitical boundaries and policies that affect people’s ability to migrate, and also because the human contribution has accelerated climate change. “Traditional ecological knowledge help[s] give us guidelines,” she said. “We are challenged to grow, to evolve, to take responsibility for ourselves and our behavior, and to work together so that way we can thrive as we move through this.”

During the discussion, it was noted that some communities are facing full relocation due to rising sea levels and erosion. DeWitt said that it is important to understand that the relocation process is not just about physical displacement but also about addressing longstanding inequities perpetuated by historical contracts and governance structures. These communities were historically forced to settle in vulnerable areas and have complex governance structures involving federal, tribal, and municipal entities. The fear of losing relationships, ties, and stories is a significant concern; DeWitt suggested that ideal solutions would involve moving entire communities together to preserve that sense of community.

Another problem is that funding for relocation comes from various sources, leading to fragmented efforts and long processes for relocation. DeWitt said that many Alaskan communities facing relocation due to climate change would welcome government assistance, but promises have fallen short. She added that communities feel both underresearched and overresearched, with little of the research benefiting them directly. It was also pointed out that various tribes have different cultures and different climates, so when developing solutions, it is important not to lump all tribal people together and to recognize that each tribe has its own history and knowledge. With increasing scientific recognition of climate change impacts and the value of traditional knowledge, DeWitt expressed hope for advancing collaborative solutions and a smoother transition. Schaeffer, the moderator for the session, similarly noted that it is challenging to make informed decisions when there is a lack of climate data and relationships with scientists and researchers.