**3. Ancient attempts to forecast and measure earthquakes**

#### Ambraseys (2009) notes that

4 Earthquake Research and Analysis – Seismology, Seismotectonic and Earthquake Geology

indignation"), which all describe the fury in the wrath of God. Some grave events in the region include: 526 AD in Antioch, 551 in Beirut, 749 in the Levant, 847 in Damascus, 856 in Qumis, 893 in Armenia and in Iran the same year, 1042 in Syria, 1138 in Aleppo, 1157 in Syria, 1170 in Aleppo, 1202 in Syria, 1268 in Cilicia, and 1509 in Istanbul, 1667 in Shamakhi,

Fig. 2. The interior of the earth according to the perception of Athanasius Kirchner in the

Seneca (4 BC - 65 AD) assumed movements in the air inside the Earth as reasons for earthquakes. More severe events in Italy took place in the second millennium AD, e.g. 1169 in Sicily, 1348 in Friuli, 1693 in Sicily, 1694 in Irpinia, 1783 in Calabria, and 1857 close to Naples. The latter event led the British scientist Robert Mallet to lay down the foundations of modern seismology and to begin with the compilation of an earthquake catalogue, which in the end listed 6800 events with location and effects, just a few years after the first earthquake statistics put together by K.E.A. Hoff in 1840 (Bolt et al., 1995). Other European events of that time include the 1356 Basel event and the earthquakes in Lisbon in 1531 and 1755. It was especially the 18th century that brought a large number of new scientific

year 1678 with subterranean lakes, rivers and pools of fire.

1727 in Tabriz, and 1780 in Iran.

*historical evidence shows that earthquake prediction was a serious preoccupation for the early soothsayer, astrologer, or prophet, and there are many recorded instances in history of earthquakes having been forecast (Ambraseys, 2009).* 

The earliest known seismoscope, which was designed to substitute sheer soothsaying, was made in China by Chang Heng in 132 AD. Its mechanism remains unknown, although it is said that his device actually measured some earthquake four hundred miles away which was otherwise unnoticed at the site of the seismoscope. It took science a very long way until Jean de Hautefeuille led the European movement to invent a device that would measure earthquakes in 1703. Nicholas Cirrillo used pendulums to measure a series of earthquakes in Naples in the year 1731. James Forbes was probably the first scientist to attempt explicitly to give a seismological instrument a "long" period, something which was finally achieved by Emil Wiechert at the turn of the century, who invented probably the oldest type of seismograph still in use today (Dewey & Byerly, 1969). But the direction, intensity and duration of seismic waves had already been recorded by Luigi Palmieri in as early as 1856 (Bolt et al., 1995).

The forecast of earthquakes has always been highly desired. From the year 373 BC in Greece stories exist that describe the reaction of rats and centipedes that flee from a severe earthquake just hours before the event. Modern studies concentrate more on electrical resistivity of rocks, electrotelluric fields, electrochemical potential or electrical conductivity of water (Sidorin, 2002). Fossil quakes may help with the forecast if a region is earthquake prone, archaeoseismology plays an important role in this context. Tree ring analyses may result in the reconstruction of historical events, just as sediments in lakes might prove ground liquefaction (Bolt et al., 1995), and there are many more fields in which palaeoseismology may contribute historical and prehistorical records. Reicherter et al. (2009) sum up the current knowledge on earthquake ground effects for hazard assessment.

#### **4. Earthquakes in mediaeval and modern Europe**

Mediaeval earthquakes were seen as a continuation of ancient natural conditions and have found their way even into literary works of the Renaissance, like those by Shakespeare, who mentions earthquakes for example in Henry IV, Part I ("I say the earth did shake when I was born").

Ambraseys (2009) in his exhausting catalogue of earthquake events for the Mediterranean and the Middle East lists archaeological data, epigraphs and inscriptions as important sources for the analysis of past earthquake events. But he prefers the literary sources and states:

*It is symptomatic of cultural changes since the First World War that, as instrumental, electronic or other mechanical reporting of events has grown, and news is increasingly disseminated by radio and television, a parallel decline is visible in both the volume and the quality of documentary and descriptive accounts of earthquakes in the twentieth century (Ambraseys, 2009).* 

Kozák & Ebel (1996), in addition to that, suggest not to forget pictorial sources when assessing earthquake data (confer fig. 3):

*The depictions of historic earthquakes provide some macroseismic information for reevaluating the intensities of the portrayed events. Furthermore, the depictions may be used to infer other macroseismic information, such as ground acceleration levels, soil amplification or liquefaction, and the amount of tsunami damage. In some cases, an analysis of the depictions could indicate the need to reclassify the sizes or locations of some historic earthquakes (Kózak & Ebel, 1996).* 

For his catalogue Ambraseys writes about the European/Occidental sources that "for Classical, Roman and Byzantine times almost all the sources are well known, and they are relatively limited in number and mostly published" (Ambraseys, 2009). He writes that on top of that Arabic sources "have generally been identified and published", most of them being narrative histories. However, he claims, "little or no archival material survives from this early period". Ottoman sources are often connected to the cost for the repair or reconstruction of structures affected by the shock. Venetian sources, on the other hand, have a long period of observations, chiefly from coastal regions of the north-eastern part of the Mediterranean, thus adding reliable historical written material to earthquake catalogues.

During the second part of the twentieth century several scientists in developed countries realised that a new approach with respect to historical earthquakes had to be tried. In France, for example, it was Jean Vogt, who understood that historical seismicity needed a revision through a return *ad fontes* (cf. Vogt, 1979). He began to collect the original sources as far as available: periodicals, newspapers, administrative, notarial and family archives, and he did not stop short of libraries and archives of neighbouring countries. With regards to the need to identify smaller earthquake events, especially in regions with a low seismicity, Fréchet (2008) remarks:

*Often, seismic catalogues concentrate only on the largest damaging earthquakes in a region, neglecting valuable information on foreshocks and aftershocks and on smaller events.* 

And he carries on by asking for more enthusiasm from the people involved in hazard mitigation also with regards to the financial aspect:

*For each event, it is necessary to make exhaustive use of all existing catalogues in order to identify the least trace of earthquake, aftershock, and background seismicity. Once an event date and location is known approximately, it is usually straightforward to search for original descriptions in newspapers, periodicals, etc., for the last three centuries at last (Fréchet, 2008).* 

mentions earthquakes for example in Henry IV, Part I ("I say the earth did shake when I was

Ambraseys (2009) in his exhausting catalogue of earthquake events for the Mediterranean and the Middle East lists archaeological data, epigraphs and inscriptions as important sources for the analysis of past earthquake events. But he prefers the literary sources and

*It is symptomatic of cultural changes since the First World War that, as instrumental, electronic or other mechanical reporting of events has grown, and news is increasingly disseminated by radio and television, a parallel decline is visible in both the volume and the quality of documentary and descriptive accounts of earthquakes in the twentieth century (Ambraseys,* 

Kozák & Ebel (1996), in addition to that, suggest not to forget pictorial sources when

For his catalogue Ambraseys writes about the European/Occidental sources that "for Classical, Roman and Byzantine times almost all the sources are well known, and they are relatively limited in number and mostly published" (Ambraseys, 2009). He writes that on top of that Arabic sources "have generally been identified and published", most of them being narrative histories. However, he claims, "little or no archival material survives from this early period". Ottoman sources are often connected to the cost for the repair or reconstruction of structures affected by the shock. Venetian sources, on the other hand, have a long period of observations, chiefly from coastal regions of the north-eastern part of the Mediterranean, thus adding reliable historical written material to earthquake

During the second part of the twentieth century several scientists in developed countries realised that a new approach with respect to historical earthquakes had to be tried. In France, for example, it was Jean Vogt, who understood that historical seismicity needed a revision through a return *ad fontes* (cf. Vogt, 1979). He began to collect the original sources as far as available: periodicals, newspapers, administrative, notarial and family archives, and he did not stop short of libraries and archives of neighbouring countries. With regards to the need to identify smaller earthquake events, especially in regions with a low seismicity,

*Often, seismic catalogues concentrate only on the largest damaging earthquakes in a region,* 

*For each event, it is necessary to make exhaustive use of all existing catalogues in order to identify the least trace of earthquake, aftershock, and background seismicity. Once an event date and location is known approximately, it is usually straightforward to search for original descriptions in newspapers, periodicals, etc., for the last three centuries at last (Fréchet,* 

*neglecting valuable information on foreshocks and aftershocks and on smaller events.*  And he carries on by asking for more enthusiasm from the people involved in hazard

*The depictions of historic earthquakes provide some macroseismic information for reevaluating the intensities of the portrayed events. Furthermore, the depictions may be used to infer other macroseismic information, such as ground acceleration levels, soil amplification or liquefaction, and the amount of tsunami damage. In some cases, an analysis of the depictions could indicate the need to reclassify the sizes or locations of some historic earthquakes (Kózak* 

born").

states:

*2009).* 

*& Ebel, 1996).* 

catalogues.

Fréchet (2008) remarks:

*2008).* 

mitigation also with regards to the financial aspect:

assessing earthquake data (confer fig. 3):

Fig. 3. Earthquake events in Rossanna and Constantinople, according to Hermann Gall, in the year 1556.

#### **5. Criticism of the sources**

But of course we are well aware of the fact that the quoting of "original" texts is often unprecise or even faulty. Many of the written sources have already been shortened, translated, paraphrased or compiled, thus distorting the information of the underlying primary sources. On top of that, some compilations may include mistakes like wrong datings. Sometimes compilations are a mixture of reliable and questionable data (e.g. Pfister, 1988; Glade et al., 2001 on that topic), as is especially the case with the "long" German catalogues of Central European earthquakes collected by Reindl (1903a, 1903b; Gießberger, 1922; Sieberg, 1940, see also Fig. 4).

An illustrative collection of problems when dealing with historical written sources is given by Kárník (1988):

*In some cases storm effects, landslides or subsidences are reported as seismic phenomena. Another source of error is the wrong transcription of names of localities, or the case of some localities having identical or very similar names; as a consequence, earthquake epicentres have been moved to wrong places. 'New' earthquakes can originate simply by listing twice an event reported with the date given in different sources according to the Julian or the Gregorian calendar. Another source of similar manufacturing of earthquakes are errors in transcribing the dates, e.g. Jan.-June, VI-XI, etc. An opposite phenomenon may occur because of a long period of war, foreign occupation of a country, plague, or other reason for which the records either were not made or were destroyed, which results in an artificial interval of quiescence. It is imperative to work with the original reports as much as possible, but this is not easy because some old sources are not accessible to an investigator or have been destroyed or require special knowledge of language.* 

The plausibility of the data can only be elicited by a thorough assessment of the data itself and a comparison of the data in question with established catalogues. In general, the cooperation of historians and geoscientists proves very valuable for the assessment of historical hazard data (Alexandre, 1990; Coeur et al., 1998), even if the majority of the historical documents only provide binominal data compared to the more valuable censored data of mostly younger documents (cf. Stedinger & Cohn, 1986).

As usual with historical data, several aspects regarding the quality of written sources need of course to be kept in mind (cf. e.g. Pfister & Hächler, 1991; Coeur et al., 1998; Fliri, 1998; Pfister, 1999; Glaser et al., 2002). The documents may differ greatly with regard to terminology, detail, educational background of the author etc. Many sources are not capable of providing all the data expected and needed for a thorough assessment of earthquakes. But the more different sources (or -later- archives) can be included, the more precise earthquake data will become overall. Excellent examples of the analysis of historical data are the publications of Galadini et al. (2001) about the Veronese earthquake of 1117, the work of Mucciarelli & Stucchi (2001) about disaster scenarios (or, for the general approach, e.g. Hammerl & Lenhardt, 1997 and Gisler et al., 2004). An up-to-date overview on German historical approaches is for example given by Grünthal (2004).

The past few hundred years have seen quite a large number of descriptive and parametric catalogues of historical earthquakes as a result of long-lasting archive work with written sources. Among the larger and exhaustive catalogues are the writings of Bonito (1691), Coronelli (1693), Seyfart (1756), whose catalogue is considered the first 'modern' catalogue (Fréchet 2008), Hoff (1840), and Milne (1911), as Ambraseys (2009) points out. In France for example major catalogue projects were undertaken from the 19th c. onwards, e.g. by Alexis Perrey (cf. Perrey, 1841), later came the global catalogue of Ferdinand Montessus de Ballore,

But of course we are well aware of the fact that the quoting of "original" texts is often unprecise or even faulty. Many of the written sources have already been shortened, translated, paraphrased or compiled, thus distorting the information of the underlying primary sources. On top of that, some compilations may include mistakes like wrong datings. Sometimes compilations are a mixture of reliable and questionable data (e.g. Pfister, 1988; Glade et al., 2001 on that topic), as is especially the case with the "long" German catalogues of Central European earthquakes collected by Reindl (1903a, 1903b; Gießberger,

An illustrative collection of problems when dealing with historical written sources is given

The plausibility of the data can only be elicited by a thorough assessment of the data itself and a comparison of the data in question with established catalogues. In general, the cooperation of historians and geoscientists proves very valuable for the assessment of historical hazard data (Alexandre, 1990; Coeur et al., 1998), even if the majority of the historical documents only provide binominal data compared to the more valuable censored

As usual with historical data, several aspects regarding the quality of written sources need of course to be kept in mind (cf. e.g. Pfister & Hächler, 1991; Coeur et al., 1998; Fliri, 1998; Pfister, 1999; Glaser et al., 2002). The documents may differ greatly with regard to terminology, detail, educational background of the author etc. Many sources are not capable of providing all the data expected and needed for a thorough assessment of earthquakes. But the more different sources (or -later- archives) can be included, the more precise earthquake data will become overall. Excellent examples of the analysis of historical data are the publications of Galadini et al. (2001) about the Veronese earthquake of 1117, the work of Mucciarelli & Stucchi (2001) about disaster scenarios (or, for the general approach, e.g. Hammerl & Lenhardt, 1997 and Gisler et al., 2004). An up-to-date overview on German

The past few hundred years have seen quite a large number of descriptive and parametric catalogues of historical earthquakes as a result of long-lasting archive work with written sources. Among the larger and exhaustive catalogues are the writings of Bonito (1691), Coronelli (1693), Seyfart (1756), whose catalogue is considered the first 'modern' catalogue (Fréchet 2008), Hoff (1840), and Milne (1911), as Ambraseys (2009) points out. In France for example major catalogue projects were undertaken from the 19th c. onwards, e.g. by Alexis Perrey (cf. Perrey, 1841), later came the global catalogue of Ferdinand Montessus de Ballore,

data of mostly younger documents (cf. Stedinger & Cohn, 1986).

historical approaches is for example given by Grünthal (2004).

*In some cases storm effects, landslides or subsidences are reported as seismic phenomena. Another source of error is the wrong transcription of names of localities, or the case of some localities having identical or very similar names; as a consequence, earthquake epicentres have been moved to wrong places. 'New' earthquakes can originate simply by listing twice an event reported with the date given in different sources according to the Julian or the Gregorian calendar. Another source of similar manufacturing of earthquakes are errors in transcribing the dates, e.g. Jan.-June, VI-XI, etc. An opposite phenomenon may occur because of a long period of war, foreign occupation of a country, plague, or other reason for which the records either were not made or were destroyed, which results in an artificial interval of quiescence. It is imperative to work with the original reports as much as possible, but this is not easy because some old sources are not accessible to an investigator or have been destroyed or require special knowledge* 

**5. Criticism of the sources** 

1922; Sieberg, 1940, see also Fig. 4).

by Kárník (1988):

*of language.* 

completed in 1907 (Montessus de Ballore, 1904-1907), a process which is typical for most of the developed countries in Europe.

Fig. 4. Earthquakes in central Europe in the 15th century, according to Sieberg (1940).
