AI should enhance, not replace, scientific observation and human judgement. Our own data remains the primary source behind our work. We recognise the potential for AI to support and improve research, and we will use AI where it can meaningfully enhance our analysis, understanding and communication of environmental data, while maintaining scientific oversight and transparency.
Data analysis and visualisation
As a project, we have a growing archive of local observations that span a five year period. AI enables us to explore these data sets, perform calculations, identify patterns and produce graphs and visualisations.
AI has also supported comparing our data to wider climate datasets, such as those from the Met Office, and improve context when writing our reports.
We’ve used OpenAI’s GPT-5.6 Sol to assess data quality, compare our data to nearby stations, and reconstruct missing data using ordinary least-squares (OLS) linear regression
Automating our outputs
AI has also become a useful development tool for St Nicholas Weather, enabling us to in-source more of what we do, rather than relying on commercial companies.
It assists with writing and debugging code, developing interactive features, integrating weather-data APIs, producing data visualisations and improving the design and functionality of the website.
AI supported development of our API, live weather dashboard and the systems used to present station observations online.
Writing and communication
AI may be used to help organise information, improve explanations and make technical or statistical information easier to understand.
That said, we acknowledge that generative AI can make mistakes, misinterpret information or produce inaccurate information. Therefore, we do not automatically publish AI-generated material. Instead, we evaluate it rather than treat it as an authoritative source.
St Nicholas Weather remains under human editorial control, and responsibility for what appears on the website remains with us.