# The linguistic and emotional effects of weather on UK social media users

## TL;DR

This study explores how different weather conditions affect the mood of UK social media users by analyzing the language and emojis they use.

## Contribution

The paper introduces a novel context-sensitive sentiment metric to analyze emotional responses to weather on social media.

## Key findings

- Emotional responses to weather are influenced by combinations of weather variables and regional language differences.
- Language used in weather discussions predicts the severity of each weather condition and varies across different weather combinations.

## Abstract

Weather significantly impacts mood and happiness, yet observing this at scale and differentiating across weather types is challenging. This study examines the variation in public sentiment related to different weather conditions, as reflected in the vocabulary used in UK-based social media (Twitter) content. We introduce a novel context-sensitive sentiment metric to construct scales that rank words and emojis by both weather severity and emotional intensity, controlling for linguistic variations that naturally occur in different discussion topics. Our findings reveal that emotional responses to weather are complex, influenced by combinations of weather variables and regional language differences. For five weather conditions (temperature, precipitation, humidity, wind speed and barometric pressure) we first identify the sentiment and weather severity associated with words commonly used to discuss them, highlighting the distinct vocabulary used to express positive and negative emotions for each weather type. Next, we demonstrate that language used in weather discussions predicts the severity of each condition and varies across different weather combinations. These findings highlight the importance of context-sensitive sentiment methods for better understanding public mood in response to weather. This approach reveals systematic relationships between weather conditions and public mood, offering insights for impact-based weather forecasting and risk communication.