The Spanish population's interest in climate change based on Internet searches

Scientific evidence and international policy processes (e.g., IPCC reports; the 2015 Paris Agreement) underscore the urgency of climate change (CC) mitigation. Public pressure and awareness are essential to enable ambitious policy action. Because the internet and social networks are major sources of environmental information for Europeans, internet search behavior can serve as a proxy for public interest and concern. Prior surveys (Eurobarometer, Pew, Yale Program on Climate Change Communication) show high and rising levels of concern, including in Spain. This study investigates the Spanish population’s interest in CC by analyzing Google search behavior and assessing how external factors may shape that interest. The objectives are to: (1) identify, analyze, and classify CC-related search descriptors used in Spain (Apr 2020–Mar 2021); (2) analyze monthly search trends for these descriptors (Apr 2020–Mar 2021); (3) compare annual search interest for four core CC terms across five years (Apr 2016–Mar 2021); and (4) test associations between search interest and (a) media coverage volume, (b) extreme weather events, and (c) major CC-related events (e.g., COPs, strikes, IPCC reports). The Spanish context is salient due to intense CC debate in 2019 (youth-led movements, emergency declarations, COP25 in Madrid) and ongoing national and regional climate legislation.

Internet search metrics, especially Google Trends, are increasingly used to gauge public interest and issue salience due to their large scale, low cost, anonymity, and temporal coverage. Prior work links spikes in CC-related searches to media coverage, international climate conferences (COPs), celebrity influence, documentaries, and climate protests (e.g., Fridays for Future). Some studies find that extreme weather events can increase CC information-seeking, though results vary. In Spain, previous analyses examined alignment between media coverage and searches or evaluated the quality and framing of CC information in media rather than the public’s search interest per se. Misconceptions conflating climate change, the greenhouse effect, and the ozone hole are well documented and persist in public discourse.

Design and data sources: A two-dataset approach was employed. Dataset 1 used SEMrush’s Keyword Magic Tool to identify and quantify CC-related Google organic search keywords in Spain. Starting from four seed descriptors ("climate change", "global warming", "climate emergency", "greenhouse effect"), an initial list of 55,280 keywords was retrieved. A cut-off of average monthly search volume ≥100 (Apr 2020–Mar 2021) was applied, capturing approximately 87% of total searches for the seed keywords in the period. After removing duplicates and terms not clearly related to CC information-seeking (e.g., person- or event-specific terms), a final sample of 200 keywords was retained. These were classified into eight categories based on search intent via an author-led coding and triangulation process: Generic searches; News and resources; Causes; Consequences; Solutions; Legislation, strategies and plans; Events; Denialism. Dataset 2 used Google Trends weekly relative search interest (scaled 0–100) in Spain for the four core terms ("climate change", "global warming", "climate emergency", "greenhouse effect") across Apr 2016–Mar 2021. Explanatory/context variables were compiled weekly: (a) media coverage (daily number of CC-related news items in RTVE newscasts via Civio’s Verba tool), dichotomized as below/above the overall average; (b) extreme weather events (from AEMET monthly climatological summaries, aggregated to weeks); (c) CC-related events (weeks including climate demonstrations/strikes by Fridays for Future–España, UNFCCC climate summits/COPs, or releases of IPCC reports). Measures: SEMrush provided (1) average monthly search volume per keyword (Apr 2020–Mar 2021) and (2) a 12-month keyword trend index (0–1, rescaled to 0–100). Google Trends supplied weekly relative interest values. Analysis: Using SPSS v22, the study conducted frequency analyses, Student’s t-tests to compare means (e.g., weeks with vs without events; above- vs below-average media coverage; weeks with vs without extreme weather), and one-way ANOVA to test annual differences in search interest for each keyword across 2016–2021.

Search volumes and categories (Apr 2020–Mar 2021):

• Total average monthly searches for the 200 terms: 146,830 (~>1,750,000 annualized). The top terms by average monthly volume were: "climate change" (33,100), "greenhouse effect" (18,100), "global warming" (12,100), with "climate emergency" ranking 11th (1,600). Among the top 20, informational queries (definitions, causes, consequences) and child-oriented resources were prominent.
• Category distribution: Highest volumes targeted generic information (e.g., "what is climate change", definitions, Spain-specific info), followed by news/resources. Causes and consequences were mid-ranked; denialism and events had the lowest volumes. Searches on legislation/strategies rose due to interest in Spain’s climate law. Monthly trends within 2020–2021 (SEMrush trend index):
• Highest overall search interest months: April, June, and November 2020; lowest: August–September 2020.
• Notable spikes: Legislation/strategies surged from ~34.1 in May 2020 to ~86.6 in June 2020, coinciding with the Council of Ministers’ approval of the CC and energy transition draft law on May 19, 2020. Events rose from ~15.8 in October to ~84.7 in November 2020, aligned with the International Day against CC (Oct 24). Denialism increased in parallel (to ~88.3 in Nov 2020) and again modestly in Jan–Feb 2021. Five-year trends (Google Trends, Apr 2016–Mar 2021):
• One-way ANOVA showed significant annual differences in relative search interest for: "climate change" (p=0.000), "global warming" (p=0.011), and "climate emergency" (p=0.006); not significant for "greenhouse effect" at p<0.05 threshold in Table 2 (reported p=0.01 without asterisk). Interest generally increased from 2016, peaking in 2019, with lower values in 2020 amid the COVID-19 pandemic and partial recovery in 2021. Associations with media coverage, extreme events, and CC-related events:
• Media coverage (RTVE): Weeks with above-average CC news had higher mean search interest for all keywords: climate change 17.80 vs 8.66 (p=0.000); global warming 13.23 vs 2.49 (p=0.000); climate emergency 31.54 vs 22.18 (p=0.000); greenhouse effect 41.00 vs 33.74 (p=0.018). Total sum index 103.33 vs 67.06 (p=0.000).
• Extreme weather events (AEMET): Significant association only for "greenhouse effect" (weeks with events 38.33 vs without 29.51; p=0.008). No significant differences for climate change (p=0.500), global warming (p=0.380), climate emergency (p=0.550); total p=0.084.
• CC-related events (COPs, strikes, IPCC reports): Significant increases during event weeks for climate change 20.15 vs 10.41 (p=0.000); global warming 20.71 vs 3.60 (p=0.000); climate emergency 31.12 vs 24.51 (p=0.023); not significant for greenhouse effect 37.32 vs 36.35 (p=0.820). Total 109.29 vs 74.78 (p=0.000).

The findings indicate that Spanish public interest in CC, as proxied by Google search behavior, rose steadily from 2016 and peaked in 2019, aligning with heightened social mobilization (Fridays for Future strikes) and the Madrid-hosted COP25. Interest dipped during the COVID-19 pandemic in 2020 as public attention and media agendas shifted, with partial recovery in 2021. Media coverage appears to be a strong driver of search interest across all four keywords, reinforcing the agenda-setting role of news in shaping information-seeking. Major CC-related events (COPs, protests, IPCC reports) also stimulate search activity for climate change, global warming, and climate emergency, but not for greenhouse effect, suggesting that event-related discourse uses more accurate CC terminology and may foster better climate literacy. In contrast to some prior studies, extreme weather events did not generally increase CC-related searching, except for the term "greenhouse effect." This exception highlights persistent public misconceptions conflating CC and the greenhouse effect, a pattern widely noted in the literature. The prominence of generic informational queries and the relatively low volume of denialist searches suggest broad interest in understanding CC, though interest does not equate to support for climate action, which is shaped by values and political orientations. Overall, the results address the study’s objectives by identifying key search descriptors, temporal patterns, and contextual drivers of CC information-seeking in Spain.

This study shows that internet searches about climate change in Spain increased over recent years, peaking in 2019, and are significantly influenced by media coverage and CC-related events, including social movements. The work contributes a structured, category-based view of CC search behavior (200 keywords across eight intents), quantifies monthly and annual trends, and demonstrates statistically significant associations with exogenous factors. The persistence of misconceptions (e.g., use of "greenhouse effect" as a CC synonym) underscores the need to improve climate literacy. Implications include the importance of strategic, evidence-based communication: disseminating IPCC findings, clarifying concepts and terminology, and making information accessible via dedicated websites (as envisaged in Spain’s Law 7/2021). Collaboration with environmental journalists, communicators, and educators can help translate scientific results effectively. Future research could expand to other platforms and regions, integrate additional media sources, model causal dynamics between coverage, events, and searches, and assess how improved communication affects public understanding and support for climate policies.

The results are not fully generalizable beyond Spain and the specific study period. The analysis relies on national data sources for media coverage (RTVE) and extreme weather (AEMET), which may not capture the full media ecosystem or all relevant events. Internet search data reflect behavior in a context with very high household internet access in Spain and may not represent offline information-seeking or populations with limited connectivity. Keyword selection and categorization, while systematic, depend on cut-offs (≥100 monthly searches) and coding decisions that may omit low-frequency but relevant queries.