Changes in the way people live and move in cities is driving large investments in public transport infrastructure and services. Understanding long term evolution of demand is important for maximising the benefits of these investments. This short paper introduces an approach for highlighting changes in demand over time periods of several years. The main idea is to discover and explain distinctive contexts. The input data are trip logs from transport smart card tickets and a calendar feature database sourced from local web sources. Contexts comprise arrival counts over a set of days for a particular spatial region of the network and social type of traveller. Prose and visual representations of context pairs are used to explain how demand has evolved. Ground truth data of real-world events sourced from online reports is used to demonstrate that our approach accurately highlights and gives plausible explanations for changes in public transport demand.