This poster presents an analysis of drought policies, sociotechnical choices, and stakeholders’ standpoints through the regional press and Twitter in Pima County, Arizona. This project is part of the transverse program HYDECO, which analyzes the processes of connection and disconnection in the functioning of socio-hydrosystems. The main hypothesis is that socio-ecological reconnections could be better promoted and integrated if there were all made equally visible to water managers and users. Bringing together, in an interdisciplinary approach, communication and media studies, environmental geography, and instrumented analysis of textual corpus as well as analysis of discourse, this research analyzes how the media report on drought issues. The analysis of visibility-invisibility processes is conducted in 5 steps: 1) Conducting an empirical field survey; 2) Building a network of stakeholders (relationships, strategies, initiatives); 3) Analyzing the media coverage of current events (press); 4) Analyzing the communication of stakeholders on digital media (Twitter); 5) Interpretive synthesis on the visibility-invisibility of observed data and stakeholders by case study and between case studies. This mix-method combines qualitative methods (case studies in social sciences and discourse analysis) with unsupervised learning by Probabilistic Latent Dirichlet Allocation (Topic Modeling-LDA), NLP tools, and snowball sampling. Visual representations are used as a heuristic method to analyze data. The methodology presented could be useful for understanding and comparing how diverse socio-hydrosystems, studied in different OHM, are facing drought and climate change issues.In the Pima County case, the analysis of a corpus of 1,095 press articles and 58, 082 tweets compares the distribution of topics according to sources (stakeholders) in the press and interactions of stakeholders on Twitter over time (1999-2022). The results show that drought contingency plans, climate change, and urban adaptation to water scarcity are increasingly visible topics in local media. In this context, dominant stakeholders (federal and state water agencies) tend to keep their dominant position while local watershed associations struggle to have access to the conversation on drought policies and sociotechnical choices.