MetaClimage: A novel database of visual metaphors related to Climate Change, with costs and benefits analysis

2025-07-12

AI Summary

Current climate change visual metaphors (e.g., “melting ice grenade”) lack systematic databases and multidimensional evaluation frameworks, hindering rigorous investigation of their communicative efficacy. To address this, we introduce MetaClimage—the first open-source database of paired metaphorical and literal climate imagery—annotated with crowdsourced human ratings (cognitive load, aesthetic quality, emotional valence) and NLP-based semantic analyses (topic modeling, sentiment polarity, conceptual abstraction). Quantitative comparisons reveal that metaphorical images incur higher cognitive processing costs yet significantly enhance aesthetic appreciation, abstract conceptual association, and positive affective arousal—particularly among individuals with high cognitive engagement. This work pioneers cross-dimensional annotation and computational analysis of climate visual metaphors, establishing an empirical foundation and methodological paradigm for evidence-based environmental communication design.

Technology Category

Cognitive Modeling & Cognitive Systems: AnalogyComputer Vision: Language and VisionSearch and Optimization: Metareasoning and Metaheuristics

Application Category

Web Mining and Content Analysis: Web data visualizationSearch and Retrieval-Augmented AI: Web evaluation methodologies and metricsEconomics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Visual metaphors of climate change (e.g., melting glaciers depicted as a melting ice grenade) are regarded as valuable tools for addressing the complexity of environmental challenges. However, few studies have examined their impact on communication, also due to scattered availability of material. Here, we present a novel database of Metaphors of Climate Change in Images (MetaClimage) https://doi.org/10.5281/zenodo.15861012, paired with literal images and enriched with human ratings. For each image, we collected values of difficulty, efficacy, artistic quality, and emotional arousal from human rating, as well as number of tags generated by participants to summarize the message. Semantic and emotion variables were further derived from the tags via Natural Language Processing. Visual metaphors were rated as more difficult to understand, yet more aesthetically pleasant than literal images, but did not differ in efficacy and arousal. The latter for visual metaphors, however, was higher in participants with higher Need For Cognition. Furthermore, visual metaphors received more tags, often referring to entities not depicted in the image, and elicited words with more positive valence and greater dominance than literal images. These results evidence the greater cognitive load of visual metaphors, which nevertheless might induce positive effects such as deeper cognitive elaboration and abstraction compared to literal stimuli. Furthermore, while they are not deemed as more effective and arousing, visual metaphors seem to generate superior aesthetic appreciation and a more positively valenced experience. Overall, this study contributes to understanding the impact of visual metaphors of climate change both by offering a database for future research and by elucidating a cost-benefit trade-off to take into account when shaping environmental communication.

Problem

• Analyzing impact of climate change visual metaphors on communication
• Addressing scarcity of available visual metaphor research materials
• Evaluating cognitive and emotional effects of metaphor vs literal images

Innovation

• Database of climate change visual metaphors
• Human ratings for image analysis
• Natural Language Processing for tags