A review of the relationship between environmental change and internal-human mobility

Climate change is significantly influencing worldwide habitability and liveability, with profound implications for human mobility. Most people move within national borders, yet the effects of climate hazards remain insufficiently investigated and integrated into existing theoretical frameworks. This review examines quantitative peer-reviewed literature on climate hazard-internal mobility linkages, revealing key conceptual and empirical gaps in how mobility is defined, theorized, and modeled. The literature remains geographically uneven and predominantly micro-level, with stronger evidence for short-term weather shocks than for long-term climatic processes, while often failing to distinguish between voluntary and involuntary mobility or to account for agency and timing dimensions. Climate-mobility relationships are often conceptualized as contingent, meaning hazards are seen as altering opportunity structures, rather than being framed as hazard-driven, or reflecting mediated and constructed assumptions. This is also reflected in dominant theoretical approaches often conceptualizing mobility as a rational economic response, overlooking aspirations, capabilities, governance, and institutional dynamics. The literature further suggests that climate hazards operate across multiple scales and interact with a broad range of factors to shape mobility and immobility outcomes, posing particular attention to the agricultural, asset exposure and social network channels. Overall, compounding and cumulative risks, socio-ecological tipping points, and institutional and governance dynamics remain insufficiently explored. https://www.scopus.com/pages/home#basic, https://www.webofscience.com/wos/woscc/smart-search.

Throughout history, people have always moved in search of a higher quality of life or to escape from imminent and perceived threats (D’Andrea et al., 2011). Human movements spans a spectrum from voluntary to involuntary movements, driven by a complex interplay of societal, political, economic, demographic, and environmental factors (Black et al., 2011b; Hunter et al., 2015). While the majority of those who move remain within national borders (Clement et al., 2021), due to lower cultural and economic barriers (763 million internal vs. 281 million international migrants in 2020) (Black et al., 2011a; Koubi et al., 2016), this internal dimension receives disproportionately less academic attention (Bell et al., 2015; International Organization for Migration, 2020; Otoiu et al., 2014). Many countries lack the technological capacity and institutional expertise to systematically monitor domestic movements (Skeldon, 2017; Ukaro and Davina, 2022) or treat such data as a matter of national sovereignty. This hampers a comprehensive understanding of the social and economic implications of internal mobility, such as those related to human well-being, poverty, inequality and health, and it constrains policymakers’ ability to evaluate and invest in preventive and early-response strategies.

Concurrently, climate change and unsustainable economic growth are reshaping the habitability of many regions globally. The increasing frequency of extreme events accelerates environmental degradation, placing severe pressure on ecosystems and eroding the resilience of communities dependent on natural resources (The Food and Agriculture Organization (FAO) of the United Nations, 2021). These dynamics generate highly unequal coping strategies, with some individuals compelled or able to move in search of safer or more viable conditions (Ayeb-Karlsson et al., 2020; Nawrotzki and DeWaard, 2018; Zickgraf, 2021b), and others lacking the necessary resources, capacities, or rights to remain in place (Urpilainen and Lenarčič, 2022). This topic has captured the attention of the academic community since the 1990s but, despite this, it still lacks sufficient prominence in global policy agendas. Bridging this divide calls for critical engagement with the narratives that frame mobility and immobility in the context of environmental change, as well as more robust integration of empirical evidence into policy design to ensure responses reflect the diverse realities of affected populations.

Building on previous reviews (Almulhim et al., 2024; Beine and Jeusette, 2019; Borderon et al., 2019; Cattaneo et al., 2019; De Sherbinin et al., 2022; Flavell et al., 2020; Hoffmann et al., 2020, 2021; Hunter and Simon, 2022; Kaczan and Orgill-Meyer, 2020; Mukherjee and Fransen, 2024; Piguet et al., 2018) and drawing on a large set of peer-reviewed studies, this work investigates the assumptions, definitions and theoretical frameworks used to study the relationship between environmental change and internal mobility. This article aims to deepen insight into environmental change and mobility, providing a perspective that contribute to address important gaps in the literature. It clarifies the often blurred distinctions between different forms of mobility, showing how terms such as migration, displacement, and trapped populations are varyingly defined and operationalized (Ayeb-Karlsson et al., 2020; Black et al., 2011b; Black et al., 2011a; Clech et al., 2026; Fussell et al., 2022; Kuhn, 2015). It also develops a conceptual bridge between high-level syntheses and empirical research by examining how hazard-mobility linkages are theorized, whether through deterministic, contingent, mediated, or socially constructed frameworks (Borderon et al., 2019; Cottier et al., 2022; Hunter and Simon, 2022; Turner et al., 2025). In addition, the review identifies the main channels through which climatic impacts influence mobility and highlights where conceptual and empirical blind spots persist (Cattaneo et al., 2019; Hoffmann et al., 2021).

The work proposes to go beyond the traditional dichotomy of voluntary or forced migration, by adopting the term internal human mobility to capture a wide range of intra-national movements and population flows (Kuhn, 2015). It investigates the forms of mobility identified by the Cancun Adaptation Framework, which recognizes displacement, migration, and planned relocation as forms of human mobility in the context of climate change. At the same time, the study aligns with recent literature emphasizing the full spectrum of mobility and immobility (Schewel, 2020; Simpson et al., 2024; Smirnov et al., 2022; Zickgraf, 2021a), including evacuation as well as both voluntary and involuntary immobility. The work also moves beyond climate-change-specific mobility to consider wider environment-mobility interactions. It refers to the concept of environmental change to denote both natural and human processes triggering changes in the natural environment and encompassing natural climatic variability, climatic extreme events, and long-term changes in temperatures, precipitation and sea-levels (Selby and Daoust, 2021). This distinction is particularly relevant in human mobility research since not all climate-related hazards can be attributed to anthropogenic climate change (Sengupta, 2025).

The paper is organized into three main parts. First, it describes the data and methodological approach used in the review, by describing sources and frameworks used to conduct the analysis as well as the main features of the corpus of studies taken into consideration. Second, it presents the outcomes of the scoping review, progressing from foundational terminology to theoretical interpretations and, finally, to the applied mechanisms linking climate hazards and (im)mobility. Third, it provides an in-depth discussion of the main trends and gaps in the literature, stressing the main avenues for future research. Finally, the paper concludes by reflecting on the broader implications of these findings.

This literature review adopts a scoping review approach and follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Figure 1). PRISMA is presented to predominantly guide the reporting of a systematic review and meta-analyses by providing the replicability, transparency, and completeness of the process of document selection (Page et al., 2021). The literature search was completed between September 2023 and May 2026 using Climig, Scopus and Web of Science databases. These databases were selected to balance domain specificity with broad interdisciplinary coverage. Climig provides a specialized repository focused on climate and environmental migration research, while Scopus and Web of Science offer extensive international coverage of peer-reviewed literature while ensuring broad interdisciplinary coverage. This approach enabled a systematic and reproducible search strategy even if may underrepresent gray literature, locally published and non-indexed regional research, policy reports, and non-English publications.

The literature query combined a broad set of human-mobility terms (e.g., displacement, migration, planned relocation, diaspora, immobility and evacuation etc.) with climate-and environment-related terms (climate change, environmental change, natural hazards, and disasters) (Figure 1) to capture the full spectrum of climate-related mobility in line with prior literature reviews (Almulhim et al., 2024; Ghosh and Orchiston, 2022; Hoffmann et al., 2020). The initial screening was aimed to identify the references containing the selected terms and it resulted in an initial corpus of 44, 243 records. A first filtering stage excluded documents written in languages other than English, publication types outside the scope of the review (e.g., editorials, notes, conference abstracts, and non-scientific reports), and studies belonging to disciplinary areas not directly related to the research objectives. The analysis was therefore restricted to journal articles, books, book chapters, and within social sciences, environmental studies, economics, geography, development studies, and multidisciplinary fields. This process reduced the corpus to 12,621 records. The second screening stage focused on duplicate removal and title screening to exclude studies unrelated to climate–mobility interactions, such as tourism mobility or international migration unrelated to environmental change. After this stage, 9,868 records remained for abstract review. Abstract screening represented the largest exclusion stage. A large number of documents were excluded because they addressed environmental change without considering mobility outcomes, focused on mobility without environmental or climatic drivers, relied exclusively on conceptual or normative discussions, or did not contain empirical evidence. This stage reduced the corpus to 2,111 records eligible for full-text assessment. The eligibility assessment involved a full-text review of the remaining studies according to two main inclusion criteria: (1) empirical studies investigating the effects of climatic events on human (im)mobility occurring within the boundaries of nation-states and not vice-versa; (2) documents explicitly or implicitly discussing climate change or variability in relation to human movements. At this stage, conceptual papers, purely discursive contributions, policy analyses without empirical evidence were excluded. Following this final eligibility assessment, 309 studies published between 1978 and 2026 were retained for the review.

This work relies on the information collected from the quantitative empirical literature since it requires the explicit specification of hypotheses, variables, and causal mechanisms, making the underlying conceptual assumptions more transparent and comparable across contexts. The information is organized in a dataset compiled by following the Peer Review Framework for Predictive Analytics (PRFPA) recommendations developed by the United Nations Office for the Coordination of Humanitarian Affairs (OCHA) Centre for Humanitarian Data (“the Centre”). PRFPA recommendations are designed to explore the conceptual and technical specificities of scientific analyses, gathering information about the analytical models, theoretical and conceptual frameworks, indicators and data used by researchers in their analyses (Habas et al., 2022). In this way, it is provided a standardized framework that enhances transparency and reproducibility in systematic reviews. While hand-coding introduces potential subjectivity and researcher bias, the approach supports the structured compilation of data, which will be publicly released to promote openness and transparency of information collected and processed.

This section presents the outcomes of the scoping review, progressing from foundational terminology to theoretical interpretations and, finally, to the applied mechanisms linking climate hazards and (im)mobility. First, it provides an overview of the geographical distribution of studies across the different hazards. Second, it maps the main concepts and definitions of human mobility found in the literature, while examining the assumptions and principles guiding empirical research. Third, it analyses the theoretical frameworks applied in the literature from a hazard perspective, highlighting their core causal assumptions about climate hazards and mobility dynamics. Fourth, it provides an overview of the principal climate transmission channels and vulnerability factors identified across studies.

Geographical trends

Most existing research on climate-human mobility linkages focuses on short-term weather shocks compared to long term climatic changes (Table 1). More specifically, 51 studies examine short-term temperature and precipitation variability, focusing predominantly on Northern America, Eastern and Western Africa, and Southern Asia. Flood events are analyzed in 41 studies, largely in Southern Asia, Western Africa, and Northern America, addressing both forced and voluntary mobility. Drought has been the focus of 28 studies, particularly in arid and agriculturally dependent regions of Africa and Southeast and Eastern Asia. Tropical storms, including cyclones, typhoons, and tropical depressions, are explored in 46 studies, mainly in North America and Southern and Southeastern Asia, with attention to both voluntary and involuntary mobility at different stages of the events. On the other hand, long-term fluctuations in temperatures and precipitation levels are examined in 40 studies across multiple world regions and sea-level rise is analyzed in 25 studies, especially in Southern Asia and Northern America. The remaining 78 studies either consider multiple hazards or refer to umbrella terms such as “natural disasters,” “climate change” etc. Overall, the evidence reveals an expansive but uneven global research landscape, with clear concentrations in highly studied regions and climatic phenomena, with a prevailing focus on micro-level rather than macro-level case.

Delving into conceptualizations and theoretical frameworks of internal human mobility in the context of environmental change

Existing conceptualizations of internal human mobility result in varied perspectives and classifications of movements that do not allow for a clear overview of the whole spectrum of voluntary and involuntary mobilities (Zickgraf, 2021b). Useful entry points in human mobility theory have been advanced by introducing paradigmatic distinctions in function of the level of analysis (micro—meso—macro level) and purpose of movements (functionalist vs. constructivist). The micro (individual), meso (household or community) or macro (structural) level classification allows to disentangle the complex set of factors driving human mobility (De Haas, 2021; Hagen-Zanker, 2011), yet it overlooks cross-level interactions and influences. Functionalist - constructivist paradigm distinguishes between theories treating mobility as a response to structural constraints (e.g., economic push-pull factors) from those that foreground subjective interpretations, norms, and meaning-making processes (e.g., behavioral aspects), but it often imposes overly rigid categories. This work contributes to the scientific debate by examining how human mobility is conceptualized across dimensions of agency and temporality, and by exploring how climate hazards are linked to mobility outcomes. The proposed classification moves beyond binary distinctions, such as voluntary versus involuntary migration or forced versus non-forced mobility, by: (i) developing an original typology of human (im)mobility across agency (types of mobility) and temporality (forms of mobility) dimensions; and (ii) proposing a framework to analyze mobility theories according to their underlying assumptions about the relationship between climate hazards and human mobility; and (iii) shedding light on the main transmission channels and vulnerability factors used to investigate how climate hazards shape mobility and immobility outcomes. These aspects are further discussed in the following subsections.

Conceptualizations and definitions of internal human mobility

The identification of human mobility types used in this paper is grounded in the Cancun Adaptation Framework. It represents one of the first internationally recognized policy frameworks to explicitly recognizes displacement, migration, and planned relocation as key forms of climate-related mobility. Adopting this framework provides a consistent and policy-relevant basis for classifying mobility responses across the literature and for facilitating alignment with United Nations Framework Convention on Climate Change (UNFCCC) and Intergovernmental Panel on Climate Change (IPCC) discussions on climate mobility adaptation, and loss and damage (Jiang et al., 2025). Building on this framework, the study further expands the analysis to include evacuation and immobility as additional forms of climate-related responses. To clarify the conceptual distinctions adopted in the review, each form of mobility and immobility is defined below. Displacement denotes involuntary movement from homes or habitual residences triggered by disasters or to avert imminent natural hazards. Migration refers primarily to voluntary movement in which environmental factors contribute to the decision to move. Planned relocation is the intentional resettlement of people (temporarily or permanently) to a new site. It encompasses two operational forms: evacuation, a short-term move in response to immediate threats, and resettlement, a move to a new location typically with assistance. Finally, immobility captures cases where people remain in place-by constraint or choice-despite environmental change. These categories of movements have been taken into consideration to quantify the distribution of movement types and highlight the main trends in the literature.

There is no consensus on the definition of the term “migration” in the environmental change literature. This word is used rather indiscriminately (55%), referring to both voluntary and involuntary movements, and is often defined in relation to administrative boundaries (internal vs. international) or the duration of movement (temporary, seasonal, or permanent). What clearly emerges from this review is that “internal migration” is also examined alongside “international migration” (16%) highlighting the interest in understanding the interconnections and the differences between the two types of movements. Terms stressing the lack of choice and immediate risk, such as disaster displacement (each 9%), planned relocation (4%), and evacuation (11%), are instead not widely present in the literature even if this trend is inverting due to the increasing availability and improved resolution of data. This finding echo prior reviews in noting a striking evidence gap on immobility (1%) and the need to further investigate the mechanisms and dynamics influencing the involuntary and voluntary nature (Schewel, 2020; Tarun et al., 2025).

To account for the agency-timing perspective, the above forms of mobility have been reclassified into five categories of movements (Figure 2): anticipatory mobility, proactive mobility, reactive mobility, administered mobility and immobility. These are defined in function of the capacity of people to or not to make intentional, informed, and self-directed choices regarding whether, when, and how to move (or not move) in response to environmental risks and climate-related threat. Anticipatory mobility refers to movements taken in advance of expected risks, guided by early warnings or perceived future threats (e.g., looming climate events, environmental degradation, anticipated livelihood loss). Proactive mobility captures strategic and voluntary movements aimed at improving long-term resilience or adapting to changing conditions (e.g., livelihood diversification, planned climate adaptation, development goals). Reactive mobility denotes movements triggered by disruptive hazard (e.g., floods, cyclones), typically assuming urgent, unplanned, and often involuntary traits. Immobility refers to situations in which individuals or groups remain in place despite exposure to environmental or climatic risks. This can either reflect constrained mobility and limited access to resources, but it can also represent a deliberate and meaningful choice shaped by place attachment, social obligations, cultural values, or perceptions of risk and opportunity (Adams, 2016). Administered mobility encompasses movements organized or facilitated by institutions or governments, which may or may not involve voluntariness. Together, these categories foreground the role of decision-making power and constraint in shaping environmental (im)mobility outcomes.

Theoretical perspectives of internal human mobility from a climate related point of view

This subsection offers an original perspective of the topic by organizing the diverse assumptions about how climate and weather shocks influence human mobility into a set of common categories and how these are embedded in theoretical frameworks (Figure 3). Four categories of assumptions about the influence of climate and weather shocks on human mobility are presented here. Constructed assumptions refer to situations in which the impacts of climate hazards are subjectively and unequally interpreted, perceived or socially constructed through subjective risk perception and decision making. Deterministic assumptions treat hazards as direct drivers of movement, leaving little space for agency. Contingent assumptions view hazards as altering opportunity structures, so responses depend on available alternatives, livelihood linkages, and cost—benefit logics (e.g., heat stress reduces farm income, making wage migration more attractive). Mediated assumptions are described as those in which hazard effects are filtered through institutions, policies, and social systems. Overall, the study finds that climate-mobility linkages are most often framed as contingent (44%) meaning hazards are seen as altering opportunity structures rather than purely hazard-driven (deterministic) (11%). Mediated (28%) and constructed (17%) assumptions are instead less explored reflecting the necessity to further investigate compounding risks and tipping points, as well as institutional and governance factors.

The following paragraph examines whether the above assumptions are reflected in the main theoretical approaches emerging from the reviewed studies and identifies the common conceptual traits underlying these frameworks. Rather than providing a comprehensive historical reconstruction of climate migration theories, the aim is to highlight the main shared patterns and conceptual orientations emerging in the corpus of studies considered. Readers interested in a more detailed discussion of the historical evolution and theoretical foundations of the approaches mentioned here are encouraged to consult prior works (De Sherbinin et al., 2022; Flavell et al., 2020; Hunter et al., 2015; Neumann and Hilderink, 2015; Turner et al., 2025; Zickgraf, 2021a). A first group interprets mobility as a goal-oriented and adaptive decision-making process, in which individuals or households evaluate risks, costs, opportunities, livelihood constraints, and potential benefits associated with movement. This perspective, which characterizes theories such as Push-Pull theory (Ash and Obradovich, 2020; Baez et al., 2017; Barassi et al., 2018; Dets, 2020; Epstein et al., 2022; Hassan and Tularam, 2018; Hassani-Mahmooei and Parris, 2012; Ingram and Schollmeyer, 2021; Joseph and Wodon, 2013; Khamis and Li, 2020; Khavarian-Garmsir et al., 2019; Koubi et al., 2021; Koubi et al., 2016; Lohano, 2018; Mallick and Vogt, 2014; Mastrorillo et al., 2016; Mueller et al., 2020; Mustafa et al., 2023; Nawrotzki et al., 2017; Osbaldiston et al., 2020; Reimann et al., 2021; Smirnov et al., 2022; Thorn et al., 2023; Wang and Taylor, 2016; Zouabi, 2021), Micro-Neoclassical approaches (Bertoli et al., 2022; Burzyński et al., 2021, 2022; Hirvonen, 2016; Loughran and Elliott, 2019; Oliveira and Pereda, 2020; Pavel et al., 2023; Renas and Kumar, 1978; Tierolf et al., 2023; Junhe, 2018), The Sustainable Livelihoods Framework (Ajaero et al., 2018; Amuakwa-Mensah et al., 2019; Carrico and Donato, 2019; Etana et al., 2022; Gray et al., 2020; Gray and Bilsborrow, 2013; Gray, 2009; Leyk et al., 2017; Mallick et al., 2020; Roland and Curtis, 2020; Weinreb et al., 2020), New-Economics of Labor Migration theory (Abu et al., 2014; Defrance et al., 2022; Dillon et al., 2011; Gröger and Zylberberg, 2016; Jha et al., 2018; Kubik and Maurel, 2016; Nawrotzki et al., 2016; Sarkar et al., 2022; Ukaro and Davina, 2022), the Theory of Planned Behavior (Adams and Kay, 2019; Kniveton et al., 2011; Kniveton et al., 2012; Speelman et al., 2017; Zander et al., 2020), aligns primarily with contingent assumptions, where climatic hazards are understood as altering opportunity structures and interacting with existing socioeconomic conditions rather than directly determining migration outcomes. A second group of theories emphasizes the role of structural, and institutional processes in shaping mobility dynamics. These approaches, including Neoclassical macro perspectives (Berlemann and Tran, 2020, 2021; Cameron, 2018; Dallmann and Millock, 2017; Deuster, 2021; Graves, 1979, 1980; Renas and Kumar, 1983), Migration systems theory (Chen and Lee, 2022; Curtis, 2015; De Lellis et al., 2021; Fussell et al., 2014; Fussell et al., 2022; Martín et al., 2020; Robalino et al., 2015; Winkler and Rouleau, 2021), stress how governance systems, labor markets, social inequalities, institutional capacities, and broader political and economic structures mediate the effects of climate and weather shocks on mobility outcomes. A third group highlights the importance of relational and cognitive processes in shaping mobility dynamics. These approaches, including Forced Migration Theory (Amin et al., 2021; Bhusal et al., 2021; Blondin, 2021; Paul, 2005; Tripathy Furlong et al., 2022; Upadhyay et al., 2023), Aspiration and Capability Framework (Karutz and Kabisch, 2023; Liu et al., 2017; Suckall et al., 2017) and normative-cultural frameworks (e.g., Place Attachment Theory; Bukvic et al., 2022; Landry et al., 2007) draw attention on the effects of structural barriers (e.g., land tenure, conflict, gender norms, legal status) as well as socio-psychological factors (e.g., risk perceptions and attitudes) for mobility. Such perspectives are particularly relevant for understanding differentiated vulnerability, evacuation behavior, adaptive responses, and the socially constructed dimensions of climate-related mobility.

Delving into channels of climate impact transmission and pathways of human (im) mobility

This section outlines the main transmission channels and vulnerability factors used to investigate how climate hazards shape mobility and immobility outcomes. It highlights the principal mechanisms through which rapid-onset shocks and slow-onset climatic changes influence mobility decisions (see also Supplementary Appendix 5.1.2), by providing a discussion for category and type of hazard. It also provides an overview of the multiple and interrelated dimensions through which climate hazards interact and the broad set of interconnected factors that often reinforce or counteract one another, generating diverse mobility and immobility outcomes across different contexts.

The first group of hazards taken into consideration relates to short-term weather shocks, including temperature spikes, rainfall anomalies, floods, drought episodes of limited duration, and tropical storms. These shocks often operate through immediate impacts on agricultural productivity and rural livelihoods (Dillon et al., 2011; Junhe, 2018; Lohano, 2018; Mastrorillo et al., 2016), particularly in contexts where households rely heavily on climate-sensitive production systems (Afifi et al., 2014; Chegere and Mrosso, 2022; Dinc and Eklund, 2023; Haug, 2002; Iqbal and Roy, 2015; Kubik and Maurel, 2016; Lohano, 2018; van Duijne et al., 2024; Viswanathan and Kumar, 2015). Short-term temperature and precipitation variability is frequently associated with reductions in crop yields and income instability (Bilar et al., 2016; Dillon et al., 2011; Junhe, 2018; Leyk et al., 2017; Lohano, 2018; Luong et al., 2023; Mastrorillo et al., 2016; C. D. Smith, 2014; Yuen et al., 2022), which may stimulate short-distance or temporary mobility as households seek to compensate for sudden losses or disruptions. Positive anomalies, such as unusually favorable rainfall or warmer temperatures, can in some contexts encourage longer-term out-migration by easing liquidity constraints and enabling households to finance mobility decisions (Baronchelli and Ricciuti, 2022; Mastrorillo et al., 2016; Sedova and Kalkuhl, 2020), whereas sustained negative anomalies tend to induce short-term adaptive movements (Abel et al., 2020; Bhattacharya and Innes, 2008; Chegere and Mrosso, 2022; De Longueville et al., 2020; Kumar and Viswanathan, 2013; Sherchan, 2019; Ukaro and Davina, 2022; Viswanathan and Kumar, 2015; Xiao et al., 2022; Zouabi, 2021). Similarly, drought events are also frequently discussed in the context of land degradation, water availability, and agricultural and livestock production (Ezra, 2001; Henry et al., 2003; Hien et al., 2023; Hoffmann et al., 2024; Ingram and Schollmeyer, 2021; Neumann et al., 2015; Raha and Gayen, 2021). These events are often discussed in the context of the most susceptible people groups, such as farmers and pastoralists (Afifi et al., 2014; Dinc and Eklund, 2023; Haug, 2002; van Duijne et al., 2024), who often rely on short-term movements to cover immediate needs like food shortages and losses of livelihoods (Grolle, 2015; Hermans and Garbe, 2019; Murali and Afifi, 2014). Beyond the agricultural channel, short-term shocks also influence human mobility through social and behavioral mechanisms, especially in the context of rapid-onset hazards such as floods and tropical storms. Here, factors such as risk perception, early warning systems (Addai et al., 2025; Borowski and Stathopoulos, 2023; Ha et al., 2025; Sreejith and Sinimole, 2022; Tanaka et al., 2026), prior hazard experience (Kakimoto and Yoshida, 2022; Sohn and Kotval-Karamchandani, 2023; Yin et al., 2022), household composition (Berlemann and Tran, 2021; Sohn and Kotval-Karamchandani, 2023; Sreejith and Sinimole, 2022), and the strength of local or distant social networks shape the likelihood of evacuation (Lu et al., 2016; Sohn and Kotval-Karamchandani, 2023; Yin et al., 2022), temporary displacement, or permanent relocation (Entwisle et al., 2020; Giannelli and Canessa, 2022; Joarder and Miller, 2013). Social connections can either facilitate or restrict movement (Brower et al., 2023; Das, 2019; Fussell et al., 2014; Smith and Mccarty, 2009; Thiede and Brown, 2013) by influencing access to information support during evacuation, or post-disaster recovery (Ajaero et al., 2018; Borowski and Stathopoulos, 2023; Mallick, 2023; Nahin et al., 2023; Remegio et al., 2026; Sreejith and Sinimole, 2022).

The second group of hazards relates to long-term climatic changes, including gradual shifts in temperature and precipitation and sea-level rise. These changes generally unfold through economic and environmental pathways, influencing mobility by progressively undermining agricultural productivity, water availability, and rural habitability. Many studies conceptualize long-term climatic changes as a factor that constrains livelihood opportunities (Adama et al., 2026; Amorim et al., 2014; Barassi et al., 2018; Carrico and Donato, 2019; Cattaneo and Massetti, 2019; Dehcheshmeh and Ghaedi, 2020; Deuster, 2021; Etana et al., 2022; Gray et al., 2020; Jennings and Gray, 2015; Thiede et al., 2022), thereby increasing mobility pressures, particularly toward urban areas (Barbieri et al., 2016; Barbieri and Confalonieri, 2011; Barrios et al., 2006; De Longueville et al., 2019; Henry et al., 2004a; Henry et al., 2004b; Kniveton et al., 2012; Oliveira and Pereda, 2020; Szymanowski and Latocha, 2021; Thiede et al., 2016; Weinreb et al., 2020) or regions perceived as offering better employment prospects, water security, or more stable environmental conditions (Barrios et al., 2006; Chen et al., 2025; Gutmann et al., 2005; Marchiori et al., 2012; Sherchan, 2019; Shiva and Molana, 2025). However, the evidence also underscores the presence of immobility traps (Brunel and Liu, 2025; Henry et al., 2004a; Henry et al., 2004b; Kniveton et al., 2011), where limited financial resources, restricted access to credit, and strong place attachment prevent households from relocating even when environmental degradation is severe (Brandt et al., 2016; Cattaneo and Massetti, 2019; De Longueville et al., 2019; Henry et al., 2004b; Jennings and Gray, 2015; Marchiori et al., 2012).

The influence of both slow- and fast-onset climate hazards on human mobility is filtered through multiple dimensions (micro, meso and macro) and is shaped by a broad set of factors (Socioeconomic, Demographic, Agricultural and Livestock related, Migration related, Geographical, Environmental, Political and Other data) often reinforcing, mediating or counteracting one another (Figure 4). Structural and agency-related dimensions (individual, household, community and society) of human mobility emerge as key factors of “vulnerability” and “resilience” to understand the often complex dynamics in place (Amuakwa-Mensah et al., 2019; Black et al., 2011b; Joseph and Wodon, 2013; Winkler and Rouleau, 2021). Following the IPCC framework, vulnerability is understood as the propensity or predisposition to be adversely affected, and it is shaped by exposure, sensitivity, and adaptive capacity (Pörtner et al., 2022). In the context of human mobility, vulnerability arises from both the inherent characteristics of individuals or groups and the influence of past, present, and future conditions, including exposure to previous climatic hazards and lived experiences (Gilodi et al., 2022). At individual level, a central element of vulnerability and resilience are demographic characteristics such as age, gender (sex), racial and/or ethnic identity, sexual orientation, but also social and cognitive conditions as the status in society and education (Carrico and Donato, 2019; Mastrorillo et al., 2016; Ocello et al., 2015). Secondly, household and family factors, such as experiences, values, roles, attitudes and information available, are considered key factors (Abu et al., 2014; Koubi et al., 2016; Koubi et al., 2016; Shi et al., 2019) in determining vulnerabilities and choices to move. Lastly, community and institutional opportunities, and constraints appear to play an important role in migration flows since they might influence the degree of sensitivity and resilience to climate shocks (Bodine et al., 2022; Chen and Lee, 2022; Esnard et al., 2011; Smith, 2014).

These vulnerability dynamics are closely linked to the concept of system resilience and to the impacts of climatic and non-climatic hazards (Figure 5), ultimately shaping a diverse range of mobility and immobility outcomes across different contexts (Berlemann and Steinhardt, 2017; Borderon et al., 2019; Flavell et al., 2020; Ghosh and Orchiston, 2022; Piguet et al., 2018). System resilience refers to the capacity of socio-ecological, economic, and institutional systems to absorb, adapt to, and recover from external shocks and stresses while maintaining their core functions and adaptive capacities (Preston et al., 2019). Within the climate–mobility literature, increasing attention has recently been devoted to the cumulative and compounding effects of interacting hazards and stressors over time. Slow-onset processes such as sea-level rise, together with associated impacts including soil salinization and coastal erosion, may compound the effects of extreme events such as floods, cyclones, and storm surges (Hauer et al., 2016; Kim et al., 2014; Robinson et al., 2019, 2020; Shukla et al., 2021), intensifying risks to livelihoods, infrastructure, and health (Bera et al., 2021; Chen and Mueller, 2018; Reimann et al., 2021; Saroar and Routray, 2013). At the same time, compounding stressors, such as conflicts and health crisis, have been found to amplify or exacerbate climate impacts for example by increasing competition for scarce resources and exacerbating humanitarian conditions (Ali et al., 2018; Bohnet et al., 2021; Brunel and Liu, 2025; Bustami et al., 2026; Ghimire et al., 2015; Liu et al., 2025; Petrova, 2021; Smirnov et al., 2022; Thalheimer et al., 2022; Xiao et al., 2018), but also heightening anxiety or reduce willingness to evacuate (Borowski and Stathopoulos, 2023; Li et al., 2026; Sohn and Kotval-Karamchandani, 2023). These interacting risks can progressively erode system resilience, shifting impacts from reversible to irreversible, and leading to a wide spectrum of outcomes ranging from voluntary migration and adaptation in place to displacement, involuntary immobility, and trapped populations. Nevertheless, adaptive capacity and development interventions (such as irrigation expansion) can moderate these trends, that are influenced also by risk perceptions, emotional and cultural ties to place, expectations about future opportunities, and social capital, all of which contribute to aspirations, motivations, and the perceived feasibility of migration (Chen and Mueller, 2018; Codjoe et al., 2017; Speelman et al., 2017, 2021).

The field has been receiving growing attention, drawing on a range of disciplines and methodological approaches. However, progress in understanding this complex topic has been hindered by inconsistencies in definitions, conceptual and theoretical frameworks, and by strong geographical clustering within the literature. The following subsections highlights the main gaps and limitations, with the aim of informing future avenues for research (see also Supplementary Appendix 5.2). The discussion is organized first by climate-hazard categories and subsequently turns to definitions and the conceptual and theoretical frameworks underpinning this body of work.

The most evident links between environmental change and human mobility are observed for short term climatic events such as floods, storms, droughts, and short-term temperature and precipitation variations. Flood-related human mobility underscores the need for better understanding of how risk perception and policy interventions influence movement patterns, as well as the social vulnerabilities contributing to forced displacement. Regarding tropical storms, there is a need to explore spillover effects on mobility and further investigate the role of human and social networks in influencing decisions of mobility and immobility. Both these two types of extreme events have widely been investigated in association with either voluntary and involuntary forms of mobility by stressing barriers, capacities and motivations to movements such as dependency ratio and poverty, perception of current and future climate disaster risks, evacuation attitudes and behaviors (Bodine et al., 2022; Borowski and Stathopoulos, 2023; Giannelli and Canessa, 2022; Kakimoto and Yoshida, 2022; Liu et al., 2017; Loughran and Elliott, 2019; Nahin et al., 2023; Pavel et al., 2023; Yin et al., 2022). These events have also been investigated in relation to which population groups have managed to or not return to their pre-disaster recovery (Curtis et al., 2020; Fussell et al., 2014; Stringfield, 2010), as well as the role that societal institutions play in supporting this recovery process. Assessing the impacts of short-term temperature and precipitation fluctuations on human mobility requires a comprehensive understanding of the relationship between international and domestic human mobility within the context of climate change. This understanding can be achieved through geographical comparative analyses, as well as through a further integration of political, institutional, and socio-demographic factors. For droughts, there is an increasing necessity to uncover gender dynamics among emigrants and non-emigrants, address compounding risks, and ultimately consider non-environmental factors.

Long-term climatic changes, expressed in terms of extended periods of temperature and precipitation variations, and SLR may also have disproportionate impacts. These can leave people more vulnerable to other extreme hazards and ultimately spur new mobility even if these relationships are currently not clear and in need of further investigation. SLR-human mobility models require better incorporation of hazards like salt-water intrusion and coastal erosion into models, along with a thorough understanding of the network connections and place attachment ties (Adams and Kay, 2019; Buchori et al., 2021; Bukvic et al., 2022; Codjoe et al., 2017; Speelman et al., 2017; Tierolf et al., 2023). Important research gaps remain in accounting for cascading, interacting, and compound hazards, including multi-hazard dependencies, hazard clustering, and systemic risks, as well as testing temporal sequencing and lag structures (Li et al., 2026). Addressing perceptions of climate change and specific vulnerabilities among different population groups is also essential for adaptation measures and constructing mobility scenarios considering adaptive capacity and population vulnerability in various aspects such as water supply, livelihoods, and public health (Adams and Kay, 2019; Buchori et al., 2021; Burzyński et al., 2022; Hauer et al., 2016; Reimann et al., 2023; Speelman et al., 2021; Suckall et al., 2017; Tierolf et al., 2023). Concerning long-term precipitation and temperature, it is also crucial to conduct research targeting specific demographic groups, such as the elderly, children, women, and low-income populations (Cattaneo and Massetti, 2019; Gray et al., 2020; Jennings and Gray, 2015; Thiede et al., 2016). As with sea-level rise, little is known about how people perceive long-term climatic changes or how these perceptions influence mobility intentions, and existing projections rarely account for adaptive capacity and vulnerability. A few studies have also included in their analysis matters related to financial recovery, governance assistance and economic credits with regards to whether or not they were associated with mobility decisions (Joarder and Miller, 2013; Kabir et al., 2018; Petrova, 2021), even if these aspects need to be better integrated in the research.

The literature has been characterized by the multiplicity of definitions, conceptual and theoretical frameworks adopted to frame internal human mobility in the context of environmental change. The involuntary nature of human (im)mobility is often overlooked due to the lack of agreement on definitions making it difficult to differentiate among the various forms of human movements (voluntary migration, displacement, or planned relocation) and their potential interlinkages. Internal human mobility is also prevalently studied from an adaptive and coping perspective despite significant gaps remaining in exploring it through the lens of a survival, risk reduction and resilience perspective. The literature also lacks a thorough and holistic integration of the dynamics, tipping points and non-linear processes between climate, environment, and human society, as well as the inclusion of societal intervening obstacles and facilitators characterizing human mobility. Scholars have tried to overcome these structural divisions by combining aspects of different theories into new conceptual frameworks (Baronchelli and Ricciuti, 2022; Call and Gray, 2020; Cattaneo and Massetti, 2019; De Longueville et al., 2019, 2020; Duijndam et al., 2023; Ezra and Kiros, 2001; Fussell et al., 2010; Graif, 2016; Gray and Mueller, 2012; Guo and Kapucu, 2018; Maharjan et al., 2021; Mallick, 2023; Neumann et al., 2015; Speelman et al., 2021; van der Land and Hummel, 2013) emphasizing the ability to anticipate, cope with, resist, and recover from such events (Barbieri et al., 2016; Barbieri and Confalonieri, 2011; Buchori et al., 2021; Chumky et al., 2023; Liu and Balk, 2020; Saldaña-Zorrilla and Sandberg, 2009a, 2009b; Stringfield, 2010). These attempts particularly stress the relevance of macro and the micro level factors, mainly identifying the economic considerations associated such as wage or income differentials, employment opportunities and access to assets, over meso level factors such as technology, social network, and kinship ties. However, they often remain limited to specific case study-sensitive approaches. Future work is called to further integrate people’s needs and capabilities, as well as aspirations and motivations, and translate them into theoretical and conceptual frameworks.

This article has reviewed the literature on the relationship between environmental change and internal human mobility by focusing on quantitative empirical evidence. Research on mobility responses to hazards is geographically clustered and predominantly micro-level, with much stronger evidence for short-term weather shocks than for long-term climatic change. There is much less empirical work on the Middle East and North Africa, large parts of Latin America and the Caribbean, as well as much of Europe and North America. Studies on sea level rise and long term temperature and precipitation variability are very geographically clustered and in need of further investigation, as well as their compounding and interacting impacts with short-term weather shocks. Existing conceptualizations of human mobility lack a widely accepted common ground of definitions, resulting in varied perspectives and classifications of movements that do not deliver a clear overview of the whole spectrum of voluntary and involuntary mobilities. The analysis shows that research has largely framed mobility as a proactive response shaped by shifting opportunity structures, while reactive and administered forms of mobility and immobility remain marginal. Climate - mobility linkages are most commonly framed as contingent, with hazards viewed as reshaping opportunity structures rather than from a constructed, deterministic and mediated perspectives, highlighting the need for greater attention to compounding risks, tipping points, non-linearities. This emphasis is also mirrored in theoretical frameworks that conceptualize mobility as a rational economic choice, often neglecting aspirations, capabilities, and motivations, as well as institutional and governance dynamics. The literature suggests that the impacts of climate hazards are transmitted through different channels, among which the more investigated are agriculture, asset exposure and social networks, that interact with a broad range of factors across scales to shape people’s capacity and decisions to move or stay. Nevertheless, the present review is subject to several limitations that should be acknowledged when interpreting the findings. First, despite adopting a systematic search strategy, the review may not have captured the full breadth of the rapidly expanding climate–mobility literature since it only focuses on peer-reviewed literature, potentially introducing publication and language biases and limiting the geographical and epistemological diversity of the reviewed evidence. Second, although qualitative studies provide critical insights to the topic into analysis, especially related to behavioral dynamics, and decision-making processes, they have been excluded from the final analysis in order to maintain a clearer focus on quantitative and modeling-oriented approaches. Third, the reviewed studies exhibit substantial heterogeneity in methodological approaches, spatial and temporal scales, definitions of mobility, climate hazards, vulnerability dimensions and indicators, theoretical assumptions, limiting the direct comparability of results across studies and contributing to analytical fragmentation within the field.