Chapter 1 Introduction

The past year has seen increasing public debate about the environmental quality of water bodies in England. This debate has partly played out on the social media platform Twitter as environmental non-profits, local associations, celebrities, statutory bodies and other platform users form digital publics around a variety of issues related to the water environment.

With a geographical focus on London and England, this project explores digital water publics by bringing advanced methods of computational analysis into a critical dialogue with human geography. Spread across a series of three projects, we deploy computational statistics, natural language processing, and web mapping in order to interrogate the performance of environmental publics on a social media platform:

Modelling multiple water publics

In the first project, we use word frequency analysis and topic modelling of a large corpus of tweets in order to identify six distinct topics which indicate shared ways of problematising water: plastic, pollution, place, biophilia, water use, and water sector. Comparing these topics across six actor categories, we suggest distinguishing between two water publics characterised by place-based and professional modes of engaging with the water environment, respectively.

Sentiment analysis of EA mentions

The second project deploys frequency and sentiment analysis of tweets mentioning the Environment Agency in order to understand the relative importance and emotional charge of water-related environmental management issues such as conservation, flooding, and pollution. The results suggest that users tend to express high levels of trust in EA competence to manage environmental conservation and flood risk, but dramatically lower levels of trust in the management of pollution. Subsequent close reading of tweets in the flood and pollution sub-samples suggests the active performance of a critical water public by users particularly on the occasion of media reports about pollution issues.

Urban Water Sentiment Map

Finally, the third project builds on the previous findings to demonstrate the prototype of an Urban Water Sentiment Map. Contrasting expert measurement of ecological quality with social media sentiment about London’s water bodies, the dynamic web map performs the multiple ontologies of urban water. As a device for gathering a public, the map invites the active participation of social media users in the social scientific representation of urban water governance.