Abstract
Extreme weather can lead to weather-induced disasters. These have a profound impact on communities worldwide, causing loss of life, damage to properties and infrastructure, and disruption of daily activities. In alignment with the United Nations Sustainable Development Goals, addressing the increasing frequency and severity of these events, exacerbated by climate change, is imperative. Exploring public perception and responses to weather disasters becomes crucial for policymakers to formulate effective strategies that not only mitigate the impacts but also contribute to the goal of ensuring sustainable and resilient communities. Social media, as a pervasive and real-time communication platform, has gathered a large amount of public opinion. In this work, we analyze public perception towards weather disasters based on tweets and metaphors. Metaphor, as a linguistic device, plays a pivotal role in unraveling cognitive processes and understanding how individuals perceive and make sense of concepts. We focus on tweets related to four distinct types of weather disasters i.e., floods, hurricanes, tornadoes, and wildfires, aiming to extract nuanced insights regarding public perceptions, concerns, and attitudes towards these specific events. We also deliver constructive recommendations, based on the insights.
Introduction
As climate change increases the frequency and severity of extreme weather, effective risk communication and disaster preparedness become matters of life and death. However, policymakers often struggle to understand the "deep" public perception—the underlying mental models and emotional constructs—that drive how people respond to these events.
Traditional sentiment analysis often misses the nuance of human thought. This research addresses this gap by focusing on metaphors. In cognitive linguistics, metaphors are fundamental tools used to conceptualize abstract or complex ideas (the "target") by likening them to familiar, tangible concepts (the "source"). By mapping these metaphors, we can uncover the foundational cognitive frameworks that shape collective public understanding.
Methodology
The authors employ a novel analytical pipeline centered on MetaPro, an end-to-end expert system for metaphor processing. The methodology follows three main steps:
- Metaphor Identification: Detecting whether a word in a tweet is being used metaphorically (e.g., "the flood raged").
- Metaphor Interpretation: Paraphrasing the metaphor into its literal counterpart to understand the underlying meaning.
- Concept Mapping Generation: Abstracting the relationship into a "Target IS Source" format. For example, "the flood thundered" is parsed into the mapping ACT IS NOISE.
Dataset
The study utilized a large human-annotated corpus originally developed by Wiegmann et al. (2020), consisting of 18,521 tweets related to 22 distinct weather events between 2012 and 2017. The dataset covers four disaster types: 9 floods, 9 hurricanes, 1 tornado event, and 3 wildfires. After parsing with MetaPro, the researchers identified 8,548 tweets containing 12,744 metaphorical concept mappings.
Analysis of Metaphorical Patterns
The analysis revealed several key cognitive frameworks used by the public to process disaster events:
- Disaster as a Strategic Battle: Across almost all disasters, the mapping ACTIVITY IS FIGHT was pervasive, suggesting a universal tendency to conceptualize disaster management as a strategic, adversarial battle requiring mental preparedness.
- The Urgency of Motion: The mapping EXPANSION IS MOTION highlights that the public perceives the spread of a disaster as a dynamic, escalating force demanding immediate action.
- Subjective Scale: Through the mapping SIZE IS IMPORTANCE, people use the physical magnitude of a disaster as a proxy for their own stress and nervousness.
- Disaster-Specific Signatures: Floods are perceived through lenses of collaboration (SHARING); hurricanes through inherent attributes (PROPERTY); tornadoes through fundamental cognitive processes (BASIC COGNITIVE PROCESS); and wildfires as hostile, sustained events (CONTINUANCE IS ACTION).
Results and Validation
To visualize these perceptions, the team used t-SNE (t-distributed Stochastic Neighbor Embedding) to map disaster events into a 2D space based on their averaged concept mapping representations.
The visualization successfully showed that events of the same disaster type tend to cluster together, validating that MetaPro can capture a consistent "public perception" specific to different categories of calamities. Human evaluation by psychology experts found that 85.8% of the metaphors identified by MetaPro were accurate, with a Fleiss’ kappa of 0.72, indicating substantial agreement.
Conclusion and Recommendations
The paper concludes with specific policy advice based on these cognitive insights. For instance, using "Expansion is Motion" metaphors in warnings can emphasize urgency, while focusing on "Sharing" and "Resilience" metaphors can aid flood-related community mobilization. While the study provides a strong reference point, future work should address platform bias and distinguish between direct victims and observers to further refine these models for government policy integration.
Summary updated