1. Goals
  The main scope of the project "A Basic European Earthquake Catalogue and a Database for the evaluation of long-term seismicity and seismic hazard" (BEECD) was to establish and test a procedure for compiling and upgrading a parametric earthquake catalogue of Europe and a database of primary data, to serve both as a tool for understanding the long-term seismicity and as a reliable input for seismic hazard evaluation.

  By "European" it was meant the territories of the countries belonging to EC and neighbouring areas, West of 31° East. Actually, for the assessment of seismic hazard of Europe it is necessary to take into consideration also the seismic sources which lie outside EC and can affect its territory. Therefore, the project covered also areas which do not fall within EC frontiers but the earthquakes which originate there may significantly affect EC countries.

  By "long-term" it was meant as far back as data allow. The data potential for the Middle Ages and before is not uniform throughout Europe: therefore, a comprehensive review of this time-window would have required heavy historical investigation, was likely to be extremely time consuming and to produce a badly incomplete set of data. The time-window of the XX century presents special problems due to the combination of macroseismic and instrumental data; as other research was in progress on this topic (e.g. GSHAP), only limited efforts have been devoted to it. Finally, the time-window 1400-1899 was adopted.

  As the compilation of a complete European parametric earthquake catalogue (PEC) may involve the analysis of several tens of thousands of entries, which was far from the real possibilities, the project mostly concentrated on damaging earthquakes and, among them, on a basic set of strong earthquakes, essential for long-term seismicity and hazard assessment.

2. Results and deliverables
  The main results and deliverables are:

1. the qualified BEECD working file 1400-1899 (WF): a file where the main European parametric earthquake catalogues, frozen as they were in the most recent public version, are compiled together according to a standard format. In this file the supporting datasets (roots) have been qualified by means of an ad-hoc classification tool (root class), developed in the frame of the project and then divided into three levels (acceptable, average, low). The extension of the WF to the time-windows before 1400 and after 1900 was also initiated;
   
2. the inventory, qualification and formatting of more than 1000 roots of higher root class with respect to the corresponding ones, not fully used, yet, for upgrading the national catalogues;
   
3. more than 600 earthquake studies, devoted to increase the quality of the supporting data sets for allowing better parameters determination. This research allowed the identification of some tens of so far unknown earthquakes;
   
4. a file of about 150 fake quakes, that is, a file of entries which were inserted in the current PEC on the basis of the conclusions of previous studies (in general earthquake compilations) which were proved unconsistent by recent, reliable historical investigation;
   
5. the start of the construction of the European Earthquake Roots Database - including the European intensity database - with special reference to a basic set of 383 strong earthquakes, selected among the largest earthquakes with respect to 4 European sub-areas;
   
6. a survey of the procedures of earthquake parameters determination used by the current catalogues and/or proposed more recently, including some tests and proposals toward uniform procedures;
7. a simple, robust and standard procedure for implementing hereafter the dataset and the related parametric earthquake catalogue;
   
8. the application of the procedure to start the compilation of the parametric European Catalogue of Damaging Earthquakes 1400-1899 (EuCaDE, parametric catalogue of the events above the damage threshold, Io/Ix >= 5/6, M >= 4.0).

3. How the results were achieved

  3a - The qualified BEECD Working File
  The BEECD Working File (WF) is a file where:

• several input PEC are compiled together in a single file where each entry clearly shows its origin (that is, the file where it comes from);

• the parameters of each input PEC are translated into the parameters adopted for the comprehensive format;

• entries considered by the compiler as multiple determinations of the same event are flagged;

• in case of multiple determinations, one "preference" entry is flagged.

  The input PEC to be used for compiling the WF were selected among those published and carrying references. This means that partners agreed that BEECD would adopt the most recent, published version of the input PEC, but not necessarily the very last one, and that one goal of BEECD would also be to make transparent the changes made to that version afterwards.
  Parameters and format of the BEECD WF were determined by considering the format of the input PEC, the GSHAP format and some special needs of BEECD. The adopted size threshold was Io/Ix >= 4/5 or M >= 3.5; accordingly, BEECD WF 1400-1899 contains 13.209 entries; BEECD WF <1400 contains 655 entries; BEECD WF 1900-1963 contains 16465 entries.
  Multiple determinations, that are entries from different PEC which refer to the same event, providing coincident or conflicting set of parameters, were identified in a preliminary way by experienced compilers and flagged by means of ad-hoc codes.
  The next step was to evaluate the data sets supporting each entry of the WF. Supporting data sets (hereafter defined as roots) show a great variety of type: they range from a few lines, compiled by someone with the aim of summarising the earthquake knowledge, to papers or monographs reporting comprehensive investigation on specific earthquakes. A root scale based on type (earthquake studies with datapoints, without datapoints, etc.) and quality (roots based on primary sources, primary sources and compilation, etc.), consisting of 13 classes, was created, tested, upgraded and then applied to all entries of the WF. The root classes were finally grouped into 3 levels, to establish upgrading priorities.

  3b & 3c - Available studies inventory and earthquake investigation
  In the first phase of the project partners inventoried earthquakes studies published in the last fifteen years and found out which ones of them provided "new, available roots" of better quality than the roots supporting the corresponding entries in the BEECD WF (more than 1000 roots).
  After inventorying the new available roots, partners started investigating earthquakes, the parameters of which rely on low quality roots and, additionally, key-events for some areas, for which an improvement of the present knowledge was considered necessary. The investigation was performed according to two approaches:

• standard, basic investigation, consisting in retrieval and reappraisal of the information supplied by the bibliographical references, tracing back as many primary sources as possible and processing ex-novo the whole set of information.

• ad-hoc investigation, mostly performed for the earthquakes of the Balkan area, the knowledge of which is generally of low level; it consisted in detailed search in many archives concerning the region.

  More than 600 new roots have been produced in the frame of BEECD. Some tens out of them refer to previously unknown damaging earthquakes. They were reported according to a standard format, then asssigned a root class according to criteria discussed in Task 2, and filed as new entries in the BEECD WF.

  3d - Fake quakes
  Special attention was paid to the problem of the so called "fake quakes", that are, earthquakes which were inserted in the catalogues on the basis of uncareful conclusions which are afterwards proved false on the basis of rigorous historical investigation.
  It seemed opportune that the investigation concerning earthquakes which proves an earthquake to be fake is reported in a similar way as for the other events and that a root is filed in the BEECD WF in the same way as for the true earthquakes. Such roots have been marked with Z ("zomby" entries).
  About 150 Z studies have been inventoried or produced in the frame of BEECD. To avoid that events proven as fake are simply cancelled from files without leaving trace (and to avoid future re-apparition because someone might considered them as forgotten), these 150 Z studies will form the core of a "Z catalogue" which, by the way, needs to be expanded, with priority to the period before 1400, where fake quakes are more frequent.

  3e - The European Earthquake Root Archive
  The root archive is composed by roots of varied type and quality. For the next steps, and for allowing a uniform, transparent procedure to be followed in the future, it is important that they are compiled according to standard format; it is not strictly necessary that they are all resident in the same place.
  This apply strictly for roots of type 1 (studies with intensity datapoints), for which some hints from the most advanced experiences have been proposed. A set of about 20.000 intensity datapoints exist, in principle, for the European earthquakes (time-window 1400-1899).
  For the roots of type 2 and 3 (earthquake studies, finished or in progress), a standard format has been proposed.
  The situation of the data with reference to these criteria has been checked for a set of 383 "strong" earthquakes, selected among the most significant in varied European regions. As the earthquake size distribution is really uneven in Europe, the selection was performed adopting varied Ix/Io and M threshold in 4 sub-areas.

  3f - A survey of the current procedures for earthquake parameters determination
  and some tests on recently proposed procedures
  Earthquake parameters are derived from the supporting datasets according to varied procedures, which mostly depend on the personal choice of the compiler. Moreover, even when the same procedure is used, coefficients are often determined in varied ways. The analysis performed shows that in some cases the parameters found in the input catalogues are really derived form the procedures declared, while in some others they do not fully fit in.
  Recently some formalised procedures have been proposed, aiming to allow that the same procedure, applied to the same dataset, gives the same result independently from the person who performs the determination. However, these procedures still need tests and, when adopted, need also some calibration performed at regional scale. Two tests were performed on a limited dataset of strong earthquakes.

  3g - A procedure for implementing the dataset and the related earthquake catalogue
  The upgrading of parametric catalogues was often performed simply modifying entries of the file according to: new research output, changes of parameterisation criteria, changes of compiler's mind, etc., often without leaving any trace of the change itself.
  The procedure assessed by the project is simple and transparent: an entry of the catalogue can be changed if a root of better quality is available and/or if the existing root has to be parameterised according to new criteria. This process must be recorded and made available for future upgradings.
  The procedure was tested for 383 "strong" earthquakes; it works and it can be used from now on. It is important to stress that, following this scheme, also the changes performed "unofficially" to the input PEC after their publication might be traced back and fully explained.

  3h - Towards the parametric European Catalogue of Damaging Earthquakes (EuCaDE)
  The procedure was then applied to the whole set of damaging earthquakes (Ix/Io >= 5/6 or M >= 4.0), starting from the time-window 1400-1499. Multiple determinations of the same event were sorted out, selecting one out of each group of them by means of the root class; that is, in case of several multiple determinations the preference was given to the entry carrying the best Rc.
  The determination of the new parameters was performed only for a few cases; it can easily be completed as soon as a general agreement is found.
  EuCaDE can be implemented in the time-windows before 1400 and after 1900. It is recommendable that at least the time-windows 1200-1399 and 1900-1963 are implemented with priority.

4. Dissemination

  BEECD web site
  Most of the material mentioned above is on the way of being made available to public on the server "Emidius" (http://emidius.irrs.mi.cnr.it/BEECD), which is run as a joint venture between Istituto di Ricerca sul Rischio Sismico and Istituto per le Tecnologie Informatiche Multimediali, CNR, Italy. The implementation of the parametric European Catalogue of Damaging Earthquakes will be performed in an interactive way, searching the consensus of the partners and of the national compilers by means of this web site.
  In the frame of the project also the two volumes which represent the output of the previous, precursor EC project "Review of Historical Seismicity in Europe" (RHISE) have been made available on the same server (http://emidius.irrs.mi.cnr.it/RHISE).

  Publication
  As this report shows, the project achieved many results. However, as BEECD was firstly devoted to evaluate and make available the existing material, then to perform investigation aiming to fill "holes" of knowledge, in order to avoid effort duplications, most results are not worth to be published on their own but rather only in a more comprehensive frame.
  Therefore, it has been planned to publish, as a joint output of BEECD and other autonomous initiatives, the following items:

• the "European Catalogue of Strong Earthquake (EuCaSE)", referring to the above mentioned 383 strong events and earthquakes, and may be more (CD ROM and/or website);

• the reports of more than 500 earthquake investigation performed in the frame of BEECD (BEECD new roots), besides the 83 included in the previous item (CD ROM and/or website);

• the final version of the parametric "European Catalogue of Damaging Earthquakes (EuCaDE) 1400-1899", completed according to the described procedure and, possibly, comparing the BEECD dataset with the GSHAP files.