2025
Ref: THESIS-2025-0007

Creation and Validation of the Belgian Macroseismic Database and its Importance for Engineering Seismology

Neefs, Ben


PhD thesis supervised by Van Noten, Koen; Almeida, João Saraiva Esteves Pacheco de (Université catholique de Louvain)

Abstract: For the average citizen in Belgium, earthquakes are perceived as distant events, unlikely to ever affect their daily lives or warrant concern. This perception is supported by the relative absence of significant seismic activity in Belgium over the past 30 years. Such inactivity is not unusual, as Belgian intraplate seismicity is characterized by low to moderate diffuse seismicity with long periods of inactivity. Occasionally, however, large(r) earthquakes have occurred and are likely to occur again, causing damage throughout vast regions in Belgium. Since 1985, Belgium’s instrumental seismic network has been significantly expanded which improved the detection accuracy and earthquake coverage across the country. Due to Belgium’s low to moderate seismicity, however, the number of recorded earthquakes remains limited and unrepresentative of the long-term seismic activity. To achieve a more comprehensive assessment of the seismic hazard, an additional source of information on the impact of earthquakes is required. This gap can be addressed using macroseismic intensity data. Macroseismic intensity is the classification of the severity of ground shaking at a specific location. It is based only on the observed shaking effects on people and its surroundings. Since the start of the 20th century, the Royal Observatory of Belgium (ROB) has collected an extensive amount of macroseismic data on felt events, offering a rich database which summarizes the impact of seismic events on Belgium. Up until now, however, this wealth of information was not readily available, no comprehensive compilation of collected macroseismic data was provided and information on it was limited and scattered throughout various research papers and projects. Consequently, the main objectives of this PhD research are to facilitate and promote the use of Belgian macroseismic data in future applications towards seismic hazard assessments and real-time impact monitoring. This was accomplished through: 1) The publication of the Belgian Traditional Macroseismic (BTM) database, a comprehensive compilation of 20th-century macroseismic data including 23,950 intensity data points (IDPs) for 80 felt events (Neefs et al. 2024a, 2024b). 2) The publication of the Belgian Online Macroseismic (BOM) database, providing a first compilation of online collected macroseismic intensity data since 2002, with 1,220 IDPs for 39 felt events. 3) Illustrating the potential of Belgian macroseismic data, through an evaluation of the predictions provided by the most recent Belgian seismic hazard model with observed intensities, and through an evaluation of the performance of various Intensity Prediction Equations (IPEs) on the observed intensity attenuation rate. The macroseismic survey methodologies that provided the data incorporated in the BTM and BOM databases are described in detail in this dissertation: from small-scale improvised surveys to the systematic mass-distribution of communal questionnaires and the online volunteer-based “Did You Feel It?” inquiries which, in practice, have replaced traditional macroseismic surveys in Belgium. These various macroseismic methodologies and sources are subjected to a critical review of their quality, illustrating the considerable extent of their associated uncertainties. To ensure the continuation of the collection of both traditional and online macroseismic surveys and to reduce their uncertainties in the future, new survey methodologies are suggested. Both databases combined provide a detailed summary of the impact of 125 years of seismic activity on Belgium. The maximum intensity registered is intensity 7 on the European Macroseismic Scale (EMS-98), with widespread damage in the affected localities where this intensity was observed. The earthquakes with the largest impact on Belgium in this time span are the ML = 5.6 Zulzeke-Nukerke 1938 earthquake, the ML = 5.1 Liège 1983 earthquake and the recurring shallow, small to moderate and triggered earthquakes in the Hainaut coal basin throughout most of the 20th century. The information in the BTM and BOM databases allow evaluating possible limitations on the current seismic hazard assessment of Belgium. This evaluation indicates that the predicted values align with the macroseismic observations, providing a validation of the Belgian seismic hazard map. However, as only limited data on local site effects from seismic station locations and their unconstrained share in amplification effects was used, large-scale site effects mapping is needed to extrapolate these finding throughout Belgium. Finally, the performances of various selected IPEs are compared with observed intensity attenuation rates. This evaluation illustrates that earthquake source depths are the main variable that determines IPE performances. Several IPEs are suggested to model the attenuation rates of Belgian earthquakes: the IPE by Camelbeeck et al. (2022) is best fit for shallow earthquakes, with focal depths lower than 10 km, even outside the Hainaut coal basin for which this IPE was developed. For deep(er) earthquakes, with focal depths equal to or greater than 10 km, the “chi-square regression” model by Stromeyer and Grünthal (2009) and the “French stable continental region” by Bakun and Scotti (2006) provide the best attenuation rates.

Keyword(s): 20th century ; Felt earthquakes ; Macroseismic Database ; Maximum shaking intensity
Funding: ROB PhD Grant/ROB PhD Grant/ROB PhD Grant


The record appears in these collections:
Royal Observatory of Belgium > Seismology & Gravimetry
Theses



 Record created 2025-06-11, last modified 2025-06-28


Files:
Download fulltextPNG
Files:
Download fulltextPDF