A behaviour change strategy to reduce greenhouse gas emissions from international scientific conferences and meetings

The study addresses how to reduce greenhouse gas (GHG) emissions associated with international academic conferences and meetings while maintaining their social, educational, and networking benefits. Set against a backdrop of rising global emissions and the need for rapid mitigation, the paper highlights academia’s substantial contribution to travel-related emissions and the cultural entrenchment of face-to-face conferences. The authors argue that while virtual or hybrid formats can drastically reduce emissions, many academics believe in-person interactions remain essential for achieving conference objectives. Consequently, the research aims to quantify emissions and costs of two exemplar in-person events and to design behaviour change strategies using the COM-B framework to reduce these emissions without abandoning face-to-face formats. The study explicitly asks: (1) what are the estimated GHG emissions (CO2e) attributable to two differently located European summer schools for PhD students; (2) what factors are the main contributors to these emissions; (3) what are the differences in financial costs to student and faculty participants of attending differently located events; and (4) what strategies from the COM-B behaviour change framework can be mapped against sources of GHG emissions from academic events to potentially reduce such emissions?

The paper situates academic travel within broader climate mitigation literature, noting academia’s hypermobility and substantial travel-related emissions, including evidence that conference attendance can account for a large share of researchers’ carbon footprints. Prior work shows online, hybrid, and zonal events can dramatically cut emissions, but there is skepticism regarding virtual formats’ ability to meet networking and educational objectives. Mode shifting from air to rail is often preferred by academics but is financially disincentivized because rail can be significantly more expensive than flying on many European routes. Academics recognize the need for climate action and credibility risks of high-carbon behaviors, yet find it difficult to change entrenched practices. Prior recommendations (e.g., Cercedilla Manifesto; decarbonization analyses) emphasize organizer-level strategies but less on individual participant behavior change. The COM-B framework, successful in health behavior change and proposed to address motivation–impact gaps for pro-environmental behaviors, has not been systematically applied to academic conference travel; this study fills that gap by mapping emissions sources to COM-B-aligned interventions.

Design: Retrospective analysis of GHG emissions and estimated financial costs for two in-person European Academy of Nursing Science (EANS) summer schools for PhD students (Ghent, Belgium, 2018; Lisbon, Portugal, 2019). The study includes participant travel, accommodation, food consumption, venue energy, and minimal computing-related emissions. Data source: An anonymized organizer dataset with participants’ affiliated university city, event and hotel details, individual attendance days, and social event details. Participants/setting: ~100+ PhD students and ~30 faculty per 2-week summer school; students attended 5–12 days. Travel distances and modes: Assumptions—participants crossing national borders flew; those traveling within the host country used rail (exceptions for very short cross-border cases to Ghent). For Lisbon, cross-border rail infeasibility supports the flying assumption. Distances: • Flights—return great-circle distances between nearest international home airport and destination airport using Great Circle Map. Direct economy flights identified via Skyscanner; if none, connections within 1 day were used. • Rail (and coach where needed)—return distances from home to departure airport and from arrival airport to venue, and for in-country travelers from home to venue, using Google road map distances as proxy for rail. GHG emissions estimation: UK Government GHG Conversion Factors for Company Reporting 2022. • Flights (short-haul economy): 0.15102 kg CO2e per passenger-km. • Rail (national rail): 0.03549 kg CO2e per passenger-km. • Coach: 0.02733 kg CO2e per passenger-km. Venue-related emissions: Hotel rooms (bed and breakfast), lunches, social events, teaching rooms/venue electricity; evening meals excluded (not provided by site and would occur regardless). Hotel/meeting room emissions via the Cornell Hotel Sustainability Benchmarking-based tool (hotelfootprints.org). Meal emissions based on the European region’s diet profiles; assumed 10% vegetarian uptake; caloric assumptions: lunches/socials 700 kcal per person; gala dinner 1200 kcal per person. Computing emissions: 8 h/day at 0.48 kg CO2e per day plus wireless internet 0.10 kg CO2e, multiplied by days attended. Financial costs: 2023 economy fares/prices used as proxies for similar dates, due to lack of historical fare data. • Flights via Skyscanner; rail/coach via national rail/company websites; hotel costs used actual venue costs multiplied by individual attendance days. Analysis: For each participant and event, calculated total km, kg CO2e and financial costs by mode and on-site components; summarized totals and per-person means with SD; quantified contributions of travel modes and on-site elements to overall emissions and costs.

Participants and travel: • Ghent 2018: n=122; 90.1% (111) estimated to fly; average return flight distance 1,565.85 km (SD 1,087.37); average total rail 347.34 km (SD 235.75). • Lisbon 2019: n=137; 94.9% (130) estimated to fly; average return flight distance 3,904.88 km (SD 1,693.27); average total rail 117.79 km (SD 159.31). Emissions: • Mean travel emissions per participant: Ghent 248.80 kg CO2e (SD 166.16); Lisbon 593.90 kg (SD 255.62). Flights contributed the vast majority of travel emissions: 95.05% (Ghent) and 99.30% (Lisbon). • Mean on-site emissions per participant: Ghent 85.34 kg CO2e (SD 39.49); Lisbon 129.80 kg (SD 64.69). On-site share of total event emissions: 25.55% (Ghent) and 17.94% (Lisbon). • Total event emissions: Ghent 40.77 tonnes CO2e; Lisbon 99.15 tonnes CO2e. • Total per-participant emissions: Ghent 334.14 kg CO2e (SD 173.34); Lisbon 723.70 kg (SD 262.49). Perspective against EU per-capita: Daily per-person footprint for attendance: 0.04 t (Ghent) and 0.11 t (Lisbon) versus EU 2021 daily average of ~0.02 t; i.e., 2–5.5× higher. Relative to EU 2030 target of 2.3 t CO2 per person per year, these represent 5.7× and 16× per person per day. Costs: • Mean combined travel + accommodation cost per participant: Ghent €825.01 (SD €303.08); Lisbon €761.07 (SD €286.14). • Flights were costlier to Lisbon, but Ghent had higher hotel and airport–venue rail costs; net difference ~€60 per person. Drivers and sensitivity: Distance traveled was the largest driver of emissions; Lisbon’s peripheral location increased average flight distance ~2.5× over Ghent, dominating totals despite higher on-site emissions profiles in Lisbon.

The findings show that aviation distance is the dominant contributor to conference-related emissions; on-site decarbonization (e.g., renewable power, efficient venues) can reduce only a minority share (≤26%) of total emissions in these cases. Given limited near-term aviation decarbonization, the most effective mitigation lever for in-person events is reducing flight kilometers—primarily through venue choice closer to participants’ home bases and locations with strong rail connectivity. Using the COM-B framework, the authors argue that organizers must first create structural Opportunity: centrally located venues; accessibility via electrified high-speed rail and efficient transfers; countries/venues with higher renewable energy shares; low-emission hotels; plant-based, seasonal menus; and contracting sustainable travel agencies. Opportunity alone is insufficient; boosting Capability (clear travel-planning resources, tools such as seat61.com and booking support through sustainable agencies) and strengthening Motivation (impact information, comparative feedback on emissions by travel mode, embedding carbon calculators in applications, and financial incentives or subsidies for low-carbon travel) are essential to overcome habitual flying and cost barriers. The proposed multi-component intervention portfolio includes environmental change, enablement, education, incentivisation, and persuasion. Universities’ HR and promotion systems should also align incentives with low-carbon academic practices to reinforce and normalize behavior change.

The study quantifies emissions and costs from two European summer schools and demonstrates that venue location and associated flight distances dominate total emissions. Centrally located venues with strong non-aviation transport, plus sustainable energy, accommodation, and food policies, substantially lower the emissions footprint. The authors contribute a COM-B–based behavior change framework integrating organizer-driven structural changes with participant agency to reduce emissions while retaining face-to-face benefits. They provide an example sustainability policy and strategy set, and plan prospective testing of the framework in 2024–2025 by implementing it in events and collecting organizer- and participant-level data to evaluate emissions outcomes. They invite other organizations to adopt and evaluate the framework to accelerate emissions reductions from academic meetings.

• Data source and assumptions: Based on secondary organizer data without individual surveys of travel or diet; assumed cross-border air travel and in-country rail. Some attendees (especially to Ghent) may have used ground modes internationally, potentially leading to overestimation of flight emissions for Ghent; this conservative stance likely reduces the observed between-event difference. • Missing first/last-mile data: Home-to-airport/station distances not included; likely small and similar across events. • Computing offset not applied: Did not subtract at-home computing emissions; computing accounted for <0.02% of event emissions. • No fully online comparator: While virtual or nearly carbon-neutral formats would minimize emissions, entrenched preferences for in-person meetings led the authors to focus on decarbonizing in-person behaviors. • Regional scope: Analysis limited to European events with European participants; intercontinental participation would markedly increase emissions and possibly eliminate location-based differentials; Europe’s relatively strong ground transport limits generalizability elsewhere. • Financial estimates: Travel fares derived from 2023 prices as proxies due to lack of historical fare data.