The historical weather records are essential for improving our understanding of past climate variability and extreme weather events. These records, often stored in archives, in the form of ship logs, reports, personal diaries, and other documents, offer invaluable insights into weather patterns prior to modern observational networks. However, much of this data remains fragmented, undigitized and inaccessible, limiting its potential to inform long-term climate analyses. The data-rescue process—digitizing and transcribing these records—plays a crucial role in filling the gaps in historical weather datasets, enabling a deeper understanding of long-term climate trends, rainfall variability, and the progression of climate change.
This session will explore the exciting field of data-rescue and the focus will be on how these rescued datasets are critical in reconstructing and understanding weather events from the past, providing invaluable insights for historical reanalyses. It will explore innovative techniques for recovering and digitizing historical weather observations, focusing on extreme events such as droughts, floods and storms.
We encourage submission of talks that address these key topics (not exhaustive):
- Methodologies for identifying and analyzing extreme weather events in the pre-modern era.
- Best practices for rescuing, digitizing, and integrating historical weather data.
- The role of historical observations in extending and improving climate reanalyses.
- Applications of data-rescued observations in reconstructing past climates and validating models.
- Collaborative efforts in data-rescue: success stories, challenges, and future directions.
- Experiences with emerging technologies like artificial intelligence (AI) and machine learning (ML), to automate extraction of weather data from complex archival sources.
The integration of rescued historical observations with modern datasets underpins many fields of climate research – from estimating pre-industrial baseline from which current climate is compared, providing boundary conditions for number of climate variables to correctly run GCMs, accurately reconstructing past climate, extreme events by being assimilating into long-term reanalyses. Data-rescue thus serves as a crucial bridge between historical observations and modern climate science, enabling researchers to reconstruct and reanalyse the Earth's climate system with greater certainty.