The COVID-19 pandemic has markedly reshaped people's language behaviour on social media and may beget varying interactive possibilities in the online environment. From a perspective of the virtual linguistic landscape, this enquiry looks into the dynamic and multilingual nature of the geotagged social media content by situating this issue in the context of the pandemic. An additional aim is to explore in what way the pandemic has influenced the form of this linguistic landscape. To achieve these goals, the enquiry collects and analyses over 90,000 posts on Instagram from three sites in Dublin, Ireland, between 2019 and 2021, using a set of established computational approaches from the field of natural language processing as well as diversity indices from ecology. The results picture a dynamic mechanism of the virtual linguistic landscape in Dublin over the COVID-19 pandemic, in which linguistic richness and diversity were revealed. This richness and diversity, according to our analysis, were beset by the pandemic associated with a decrease in online activities at the research sites.