We were very pleased to launch TEMPEST – our database of historical weather events – at this year’s RGS-IBG Annual Conference. With the support of the Geo team we organised a panel discussion and a small display of original and facsimile archive materials. Both were connected to a recent paper in Geo – ‘Dealing with the deluge of historical weather data: the example of the TEMPEST database’ – the journal’s first ‘data paper’.
Following an introductory post by the journal’s editors, in this contribution we wanted to reflect on our motivations for writing the paper, and creating TEMPEST, particularly in designing it as a freely accessible online resource.
Interest in historical weather is far from a new area of investigation. A number of well-known chronologies of British weather have been published and over the past 20-30 years, attempts have been made to produce searchable databases of historical weather information (instrumental data, proxy data and narrative descriptions of particular phenomena). It is widely recognised that these compilations of data or datasets have utility for the scientific study of climate, as well as satisfying the simple desire that many people have to know more about past meteorological events and their impacts on particular people and in specific places. However, in spite of rapid advances in technology, the growing amount of data (generated by labour intensive means) and the popularity of such resources, and the definite benefit that could come from uniting them, efforts largely remain separate. They are divided because they are technologically incompatible (the relevant data comes in many different formats covering instrumental observations to lengthy descriptive accounts in different languages, and database systems are constantly changing), or because they are funded only for finite periods. They can quickly become forgotten when new projects take priority or face obsolescence and lie in need of maintenance. They may also remain little known or largely indiscoverable, can be difficult to get to grips with or inaccessible to the general user.
As a research team we had some difficult conversations regarding the format, availability and deposit of our research data. It was a significant time investment to input the data into TEMPEST, time that could have been spent writing papers or our currently unfinished project book. However, we persevered and it now contains c. 18,000 event records – and we have already experienced the rewards. TEMPEST makes it possible to quickly see where we have gathered multiple narratives detailing the same event (creating a picture of the geographical extent of impact), and to piece together particular seasons or the weather of particular years or groups of years. Without TEMPEST these tasks would have required another significant time investment, and would have been reliant on the quality of our memory of the research data. Full recollection would have been an impossible challenge given the sheer quantity of data we have collected.
Although the creation of a freely available online resource was detailed in our original funding application to AHRC, as the project progressed and the volume (and quality) of our research data surpassed our expectations, team members were understandably reluctant to have our research data freely available before we had completed writing it up. However, the desire for others to use it, and our belief in its utility and popularity won over. Yet, even with an obligation to the AHRC to make our data available, but no dedicated arts and humanities data repository in the UK, it took some time to explore the various options that existed for depositing our dataset. We have just completed depositing our research data with CEDA (Centre for Environmental Data Analysis) where is it available for registered users to download as .csv files and analyse within Excel or other statistical software. A reference and DOI is provided for the dataset, alongside guidance notes relating to the data format, collection method and quality.
The database is also now ‘live’, though we may still change the url as a result of institutional moves and the conclusion of the funded period of the project.
Putting our own research data ‘out there’ is not enough. Few people are likely to find it unless we engage in targeted publicity and promotion, and it remains the case that significant time investment is required to properly come to ‘know’ the data, and use it to its potential – it is quite difficult to just ‘dip in’. We hope to use some of the time and finances allocated by a AHRC ‘Follow on Funding’ project to produce some sample ‘database stories’, promote the resource, and to embed and reconnect it with the archival repositories from which we have drawn data. We will also circulate our Geo paper to researchers involved in connected initiatives throughout Europe and further explore how it might be informally ‘joined up’. We also hope that we’ll be able to trace usage of our research data, whether it be by other academics wanting to contextualise their own research, by climate scientists developing computer models, by members of the public interested in the weather history of the place where they live, or by archive professionals interested in linking with other archives through documentary connections. As publications relating to the project are completed, where funds can be secured we are publishing them through the gold Open Access route, and we have definitely received wider readership and more interest in our work as a result – we can now also include reference to our research data and encourage its use.
Lucy Veale is a Research Associate in the Department of History, University of Liverpool, Georgina Endfield is Professor of Environmental History at the University of Liverpool, and Sarah Davies is a Reader in the Department of Geography and Earth Sciences at Aberystwyth University.