- Start
- End
- Types of event
- Lecture
- Venue
-
Rosensäle
Fürstengraben 27
07743 Jena
Google Maps site planExternal link - Speaker
- Prof. Dr. Nina Kukowski
- Prof. Dr. Joachim Denzler
- Organized by
-
Digital Academy
- Further link
- More information de
- Language of the event
- English
- Event website
- Learn more de
- Barrier-free access
- No
- Public
- Yes
- Registration required
- Yes
Digital Collections – When Does a Series of Measurements Become a Collection?
At the heart of the talk is the question of how long-term measurement data can be understood as scientific collections, digitally made accessible, and made available for new research questions. It also explores how digital methods of data analysis can open up new perspectives on complex scientific datasets.
Versuchsaufbau mit Laserkanonen für ein Strainmeter-Array
Image: Thomas Jahr
Overview
Event details
Export this event in ICS format
Registration
You can register via the following linkExternal link.
Abstracts
Geophysical Collections: From Analogue Archives to AI-Driven Data Analysis
Vortragende: Valentin Kasburg, Nina Kukowski, Ulrich Wegler
Geophysical collections at Friedrich Schiller University Jena (FSUJ) comprise historical instruments, a large number of historical seismograms, and the Sieberg Archive, which includes scientific publications, correspondence, photographs and drawings documenting the history of geosciences. On the other hand, Moxa Geodynamic Observatory (MGO) and the Thuringian Seismic Network (TSN) collect various types of geophysical data sets, leading to long time series and maintain digital archives of those data sets that are used for ongoing geophysical research, both by ourselves, but also international research groups. Operated by FSUJ, an instrumental emphasis of MGO is to measure changes in gravity and ground deformation, together with various environmental and climate parameters. TSN is jointly operated by FSUJ and the Thuringian “Landeserdbebendienst” and monitors seismic activity in Thuringia and surrounding regions. The long-term digital records collected by both facilities offer valuable opportunities for data-driven research, including the application of deep learning methods, which has become an ongoing focus of our research in the last few years.
Historical analogue seismograms could further extend these analyses by providing access to observations that predate modern digital recording. However, making these records suitable for quantitative analysis involves more than simply scanning the original documents. Missing metadata, the need for manual annotation, overlapping traces, and noise or physical damage can all complicate the digitisation and interpretation of the records. An important aim of digitalization of historical records on paper is therefore to make historical seismic observations usable for quantitative research and to explore how they can be integrated into modern approaches to geophysical data analysis.