Climate change education is increasingly being woven into science, technology, engineering and mathematics, but a new systematic review suggests that the research base remains geographically uneven and often stops short of the broader creative and social approaches promised by STEAM education.
Researchers reviewing 167 peer-reviewed studies found a clear rise in work connecting climate change education with STEM and STEAM between 2012 and the first half of 2025. Yet North America and Europe accounted for 100 of the reviewed studies, while Africa accounted for only six. The authors argue that this imbalance matters because the places most exposed to climate risks may have the least locally grounded evidence about how climate change should be taught.
A growing field with an uneven foundation
Climate change education is intended to do more than transfer scientific facts. It can help learners interpret evidence, understand uncertainty, connect environmental change with society and develop the capacity to act. STEM education offers obvious tools for that task, including scientific inquiry, mathematical modelling, engineering design and technology. STEAM extends the framework by adding the arts, opening space for communication, creativity, culture, ethics and emotional engagement.
The review, published in Climatic Change on 3 October 2026, examined how these educational traditions have actually been combined with climate change education and where important gaps remain. Rather than limiting the search to papers explicitly labelled STEM or STEAM, the researchers also included individual disciplines such as science, technology, engineering, mathematics and arts education when they were connected to climate change education.
How the review was conducted
The researchers used a PRISMA-based systematic review process. Literature was retrieved through ProQuest collections covering environmental science and education. Searches combined terms for STEM, STEAM and their individual disciplines with climate change education. The team screened titles and abstracts before conducting full-text eligibility checks.
Only peer-reviewed research papers in English were retained. Review articles, editorials and theses were excluded. The final corpus contained 167 studies. Each study was coded for publication year, geographic setting, participants, educational level, research method, thematic emphasis, pedagogical approach and the use of STEAM-related tools.
All authors independently coded the full set. They agreed on 135 of the 167 studies, producing an overall agreement rate of 80.8%. The 32 studies involving disagreement were re-examined and discussed. This gives the review a transparent consistency check, although the authors acknowledge that percentage agreement does not correct for agreement that could occur by chance.
Research accelerated after a slow start
The temporal pattern shows how quickly the field has developed. The review identified only one STEM and climate change education study in 2012. The annual count rose to eight in 2016 and reached 21 in 2024. Another 21 studies had already appeared during the first half of 2025, the point at which the literature search ended.
STEAM-specific climate education appeared later. The earliest study in the review was published in 2017. By 2024, 11 reviewed studies used a STEAM orientation, accounting for 29% of the relevant publications that year. The pattern suggests growing interest in combining scientific understanding with creative and human-centred approaches, although STEM remained dominant overall.
That distinction is important. Adding the arts is not simply a matter of placing a creative activity beside a science lesson. The review describes STEAM as an opportunity to connect analytical reasoning with communication, narratives, cultural interpretation and creative problem-solving. Those dimensions can be especially relevant to climate change because public responses depend not only on understanding physical processes, but also on values, lived experience and the ability to communicate possible solutions.
Most evidence came from wealthier regions
The clearest structural imbalance was geographic. North America contributed 64 of the reviewed studies and Europe 36. Asia accounted for 22, the Middle East 10, South America seven and Africa just six. Only eight studies addressed more than one region.
This concentration creates more than a simple publication-count disparity. Climate impacts, educational resources, curricula, local knowledge and adaptation priorities differ substantially between regions. A teaching approach developed around the concerns of a high-income urban school system may not translate directly to communities dealing with water scarcity, desertification, agricultural vulnerability or limited digital infrastructure.
The authors therefore warn that the international literature can overrepresent the educational priorities of developed regions. Africa’s six studies are particularly striking given the continent’s exposure to climate-related risks. The review points to indigenous water conservation and locally adapted environmental knowledge as examples of areas that may receive too little attention when research frameworks are imported rather than built around local contexts.
Students and teachers dominate the evidence
The imbalance also extends to who is studied. Students appeared in 99 studies, while 55 examined teachers’ practices or lesson planning. Pre-service teachers appeared in 24 studies, society or community participants in 12, institutions in 10 and parents in only three.
Among studies involving students and in-service teachers, 58% concentrated on learning and 22% on teaching methods. The review also identified 53 studies centred mainly on policy documents for curriculum development. These patterns show that researchers have built a substantial base around classroom learning and curriculum design, but evidence involving families, institutions and wider communities is much thinner.
That gap may limit the reach of climate education research. Climate action takes place outside classrooms as well as inside them. Household behaviour, community priorities, institutional capacity and public policy can determine whether knowledge developed at school translates into meaningful action.
Why the arts remain difficult to integrate
The review identifies a persistent tension between structured STEM approaches and the more interpretive character of arts education. Science and engineering courses often prioritise measurable outcomes, technical accuracy and hypothesis-driven investigation. Arts-based approaches may instead emphasise expression, narrative, interpretation and multiple valid perspectives.
Bringing these traditions together requires more than relabelling an interdisciplinary lesson as STEAM. Teachers need pedagogies that preserve scientific accuracy while allowing creative exploration. The authors note that approaches such as project-based learning, digital storytelling and other experiential methods may help bridge this divide, but the evidence base on effective integration remains underdeveloped.
The review also argues that climate change itself makes this integration worthwhile. Scientific literacy can help learners understand emissions, feedback systems and climate projections, while artistic and cultural approaches can make abstract risks more personally meaningful. Together, these tools may support the transition from knowing about climate change to discussing, imagining and designing responses to it.
What the findings mean for education
For educators and curriculum designers, the review suggests that the next stage of climate education should focus less on whether climate change belongs in STEM and more on how interdisciplinary learning can be designed effectively. That includes testing which combinations of scientific inquiry, engineering design, data analysis, storytelling, visual communication and creative work actually improve understanding, engagement and action.
The geographic findings also support greater investment in locally led research. Expanding the number of studies in Africa, South America, the Middle East and other underrepresented settings would not simply make the literature more diverse. It could reveal pedagogies and knowledge systems that are currently missing from internationally visible research.
Cross-regional research is another opportunity. With only eight studies spanning more than one region, researchers have relatively little comparative evidence about which approaches travel well across contexts and which need substantial adaptation. Climate change is transboundary, but the education research mapped by this review remains overwhelmingly local.
Important limitations
The findings should be interpreted as a map of the published research landscape rather than a direct test of whether STEAM climate education improves student outcomes. The review synthesised studies with different designs, populations and objectives, so its publication counts cannot establish the effectiveness of any single teaching method.
The search was conducted through selected ProQuest environmental science and education collections, meaning relevant research indexed elsewhere may have been missed. Ten non-English studies were excluded, which the authors acknowledge can introduce linguistic and cultural bias. This is especially relevant when one of the review’s central findings is the underrepresentation of developing regions.
The literature extraction also ended in July 2025. The review can therefore describe the trajectory through the first half of 2025, but it does not capture studies published later that year or during 2026. Finally, the reported 80.8% intercoder agreement is useful evidence of coding consistency, but percentage agreement is less rigorous than reliability measures that account for chance agreement.
A broader challenge for climate learning
The review portrays a field that is expanding quickly but has not yet matched the global and interdisciplinary nature of the problem it is trying to teach. Climate education research increasingly recognises the value of connecting science and technology with creativity and human experience. At the same time, the evidence remains concentrated in a relatively small group of regions and stakeholders.
Closing those gaps could change not only where climate education is studied, but what counts as climate knowledge in the first place. More locally grounded and genuinely interdisciplinary research may help schools move beyond teaching climate change as a technical topic and towards helping learners understand how environmental change intersects with culture, communities and practical choices.
Source Information
Study: The Development and Gaps of Integrating Climate Change Education with STEAM Education
Authors: Sung Ho Cheuk, Jacky Chun Kwan Chu, Ying Fun Chan and colleagues
Journal: Climatic Change
Published: 3 October 2026
DOI: 10.1007/s10584-026-04305-2
