Understanding water scarcity through sentiment analysis

Water scarcity in Brazil’s semiarid cities disrupts daily life and magnifies inequalities in essential services. This study examines the predominant emotions and sentiment polarity expressed within televised news narratives about water shortages in the city of Campina Grande, Brazil. The analysis comprises eight news reports available on YouTube dated from February 2019 to October 2021. Transcripts were produced with TurboScribe and processed in R. Emotion classification applied the NRC Emotion Lexicon via the R package tidytext. Sentence-level polarity estimation used the R package syuzhet. Across items, anticipation and trust were consistently the most frequent emotions, followed by sadness and fear, while anger and disgust appeared at lower levels. The polarity series showed positive peaks up to 24.6% and negative peaks up to 23.3%. Coverage linked prolonged interruptions to failures at pumping and treatment stations, uneven distribution in higher-elevation and peripheral neighborhoods, and limited institutional communication. Reported coping strategies included reliance on water trucks, resort to unsafe alternative sources, long-distance water hauling, and the temporary closure of healthcare facilities. Newscasts also documented community solidarity through well-water sharing, which mitigated burdens for some households while exposing unequal adaptive capacity. Overall, the narratives indicate cyclical infrastructural vulnerability, uneven risk, and contingent resilience. The study concludes that equitable, topology-aware operations, more transparent crisis communication, and targeted support to underserved neighborhoods are urgent. Methodologically, coupling emotion and polarity analytics with qualitative reading offers a reproducible lens to monitor social signals in urban water crises and to inform more just water governance.