Joseph Okotto-Okotto of VIRED International has published a paper that quantifies inter-observer agreement in hazard observations in and around water sources. His analysis finds only modest agreement between different observers, with a more experienced observer finding more contamination hazards at water sources than the rest of the team. In using hazard observation protocols, the survey team also struggled to classify some types of water source using standard typologies, particularly those that had been adapted by users to cope with water insecurity. For example, there were water tanks adapted to store both piped and rainwater that proved difficult to classify. This suggests both the water source typology and hazard observation protocols we used may need some revision, so that they can be used to identify and then address contamination hazards.